Implementation Guide

Success by Design
Microsoft Dynamics 365
Success by Design Implementation Guide

Published by Microsoft Corporation, 1 Microsoft Way, Redmond, Washington 98052.

Copyright © 2021 Microsoft Corporation.

Library of Congress Control Number: 2021918317

All rights reserved. No part of this book may be reproduced or used in any manner
without permission of the copyright owner.

Azure Cosmos DB, Bing, BizTalk, Excel, GitHub, Hololens, LinkedIn, Microsoft 365,
Microsoft Access, Microsoft Azure, Microsoft Dynamics 365, Microsoft Exchange Server,
Microsoft Teams, Office 365, Outlook, Power Automate, Power BI, Power Platform,
PowerApps, SharePoint, SQL Server, Visio and Visual Studio are trademarks of Microsoft
Corporation and its affiliated companies. All other trademarks are property of their
respective owners.

The materials contained in this book are provided “as-is” and Microsoft disclaims all
liability arising from you or your organization’s use of the book.

Information contained in this book may be updated from time to time, and you should
refer to the latest version for the most accurate description of Microsoft’s FastTrack for
Dynamics 365 program.

Case studies used in this book are fictional and for illustrative purposes only.

First Edition, v1.2

Printed in the United States of America

ii
This book is dedicated to Matthew Bogan, our colleague and friend. The way you think
and the way you approach everything you touch embodies a growth mindset—that
passion to learn and bring your best every day to make a bigger difference in the world.
Thank you for being a shining example.

iii
 What’s inside
Strategy / 1 Prepare / 513
1 Introduction to Implementation Guide 2 18 Prepare for go live 514
2 Success by Design overview 12 19 Training strategy 553
3 Implement cloud solutions 27
4 Drive app value 61
5 Implementation strategy 87 Operate / 593
20 Service the solution 594

Initiate / 108 21 Transition to support 625

6 Solution architecture design pillars 109

7 Process-focused solution 124 Conclusion / 666
8 Project governance 153
Acknowledgments 669
9 Environment strategy 192
Appendix 670

Implement / 229
10 Data management 230
11 Application lifecyle management 266
12 Security 291
13 Business intelligence, reporting, and analytics 331
14 Testing strategy 360
15 Extend your solution 394
16 Integrate with other solutions 426
17 A performing solution, beyond infrastructure 474

Implementation Guide: Success by Design iv

 About
the cover
The elements and details on the cover are a metaphor
for the foundations of the Success by Design
Implementation Guide. Like a sailing navigated by
an experienced guide, this book provides safe and
enjoyable passage to the intended destination—that
of a successful implementation.

Ripples and wake represent

Implementation Guide movement and progress
Success by Design
Implementation Guide: Success by Design
Microsoft Dynamics 365

People represent the

implementation team Waves represent
change management

Boat represents the Success

by Design framework
Color blue represents
depth, trust, loyalty,
sincerity, wisdom,
Artistic details confidence, stability,
represent the design and intelligence
of the implementation:
intricate but organized Rower represents FastTrack
for Dynamics 365 and the
Implementation Guide, navigated
Front of boat represents by Success by Design
moving in the right direction

Implementation Guide: Success by Design: About the cover v

 Foreword
Dear Reader,

This Implementation Guide comes at a pivotal time for the world, and
certainly for us at Microsoft. The pandemic has affected almost every
aspect of everyone’s lives, while also bringing about tremendous
change to business models and processes across every industry.

At Microsoft, we believe that the transformation we are witnessing in

businesses everywhere will continue to accelerate as we push past this
pandemic. Over the coming decades, the use of technology by our
customers will almost literally be turned upside down—with profound
implications for every aspect of their business, from customer experience
to core operational processes to the very nature of their products and
services. We are entering the predictive era, ushering out a decades-long
reactive past. Microsoft believes that the pervasiveness of data is the
fundamental, underlying ingredient that presents new opportunities for
businesses everywhere.

We have dedicated much of Chapter 1 of this book to providing our

perspective on this tectonic digital shift. The enormous opportunity
inherent in data is the reason the Azure Data, AI, Internet of Things
(IoT), Dynamics 365, and Power Platform teams are together under the
umbrella of the Digital Transformation Platform (DTP).

I am always amazed at the breadth and depth of change achieved

by our customers using Dynamics 365 and Power Platform. While it
is one thing to see the end result of an implementation project, the
journey must be thoughtfully planned and executed. Every step of that
journey matters—requirements analysis, development, testing, change
management, and ultimately realizing business value. This is where our
Success by Design framework comes into play.

FastTrack for Dynamics 365: Forward vi

 Our FastTrack for Dynamics 365 team has helped realize Microsoft’s
digital transformation vision by driving thousands of implementations
successfully to go live and beyond. In my years of working with this
team of talented solution architects, I’ve witnessed numerous permuta-
tions of implementations that have not gone according to plan. More
importantly, I’ve seen how applying the principles of Success by Design
developed by the FastTrack team early and consistently can eliminate
many of these issues, allowing customers to get to business value with
their Dynamics 365 deployments on time and as intended.

As such, I can think of no one more qualified to provide prescriptive

guidance for how to run a successful implementation than the
FastTrack for Dynamics 365 experts. Read on if you’d like a deeper
understanding of the thinking and methodology behind the thousands
of successful implementations Microsoft has driven with customers
across industries, and let us know what you think via the feedback
channels included in the book. Keeping our promise of updating the
Implementation Guide twice a year, we are pleased to bring you this
latest edition and we look forward to providing you ongoing innovation
in future releases of this e-book.

I’m excited to be able to share this expertise because it is also an

invitation to you, Microsoft’s community of customers and partners,
to join us in driving successful deployments of our platform.

Thank you for accepting this invitation, and happy reading!

James Phillips
President, Digital Transformation Platform

FastTrack for Dynamics 365: Forward vii

Strategy
1 Introduction to Implementation Guide
2 Success by Design overview
3 Implement cloud solutions
4 Drive app value
5 Implementation strategy

1
1 Guide
Introduction to
Implementation
Guide
Implementation Guide: Success by Design: Guide 1, Introduction to Implementation Guide 2
“Our success is dependent on our customers’
success, and we need to obsess about them—
listening and then innovating to meet their
unmet and unarticulated needs. No customer of
ours cares about our organizational boundaries,
and we need to operate as One Microsoft to
deliver the best solutions for them.”
– Satya Nadella, Chief Executive Officer of Microsoft

Overview
Microsoft believes that every business is in the
business of creating great customer experiences.

As a precursor to the Dynamics

To achieve that, business applications must do more than just run your
365 implementation resources in back office, marketing, supply chain, or even field operations as discrete
this book, this chapter will: entities. Organizations need to tap into data signals from across their
Explore the opportunities presented as business—and use them to drive ongoing improvements in the products,
organizations shift into the predictive era.
services, and experiences they intend to deliver to their customers.
Outline our vision for Dynamics 365.

Help readers understand what an

investment in Dynamics 365 invites
in terms of enabling holistic, unified
digital transformation.
Introduction
In this chapter, we explore how the shift from the reactive past to the
predictive era is currently affecting our industry—and will eventually
affect every industry. We also look at Microsoft’s vision for Dynamics
365 and how our apps allow holistic, unified digital transformation. In
our view, these topics are essential to anyone about to embark on an
implementation of Dynamics 365.

Implementation Guide: Success by Design: Guide 1, Introduction to Implementation Guide 3

 Armed with these perspectives, the rest of this book—which aims to
share the mindset and guidance needed to successfully implement
Dynamics 365—is sure to take a more meaningful shape for readers.
For Microsoft, this is especially important, considering the investment
that you are making in us.

Every industry turned

upside down
We find ourselves at a turning point: over the coming decades, the role
of technology will be turned upside down with profound implications
Fig. 1-1
for every aspect of business—from customer experiences to core
operational processes to the very nature of products and services.
Reactive past
Picture this: in the reactive past (Figure 1-1), a customer purchases a
Applications
car from a dealer. One day, while going down the road, the car stops
running. The customer calls the service center and reports a problem.
Application platform At this point, a service advisor opens an application, types a few notes
about the problem into a form, and then schedules a repair visit. It
may be the first time in years that any data related to the customer
or the vehicle has flowed—a few kilobytes trickling down, and only in
reaction to the car’s malfunction—into the dealer’s relational database.
Only now the dealer can provide service to the customer, but the car’s
Relational database manufacturer might never see a single byte of data from this event. As
this example illustrates, in the reactive past, data flows downward from
applications at only a drip.

Now consider what is possible today: the same customer purchases a

car from the dealer. With appropriate permissions granted by its new
owner, the car provides a continuous stream of data back to the dealer
This chapter’s content is adapted and the manufacturer. It may be a self-driving car, with computer
from the combined thinking of James
Phillips, President of Microsoft’s Digital vision models trained in the cloud via data captured by the entire
Transformation Platform Group; Rolf
fleet’s cameras. Every operating parameter of the engine, every press
Harms, Corporate Vice President of
Microsoft’s Cloud and AI Strategy; and of the brake, every trip route—all are signals that provide a rich basis
Muhammad Alam, Microsoft’s Corporate
Vice President of Dynamics 365.
for predictions to ensure a higher-value relationship between the
customer, the dealer, and the car’s manufacturer.

Implementation Guide: Success by Design: Guide 1, Introduction to Implementation Guide 4

 In Microsoft’s view, the possibility of a higher-value relationship is pres-
ent for every product, every service, and every business process. In the
predictive era (Figure 1-2), the paradigm is that everything is a source
of data and potential insights. As organizations embrace this shift, they
enable an upward data flow that originates from everywhere and, at
Fig. 1-2 the same time, informs their products, services, and applications.

Predictive era

Dynamics 365 apps

Power Platform

Azure
Data ingestion (Planet Scale
Foundation)

For companies that “grew up” in the predictive era, that paradigm is
second nature. But for mature enterprises, the path can be more diffi-
cult, as it requires transformation of existing capabilities and processes.

For Microsoft, this isn’t just theory or prediction. We have been pursu-
ing our own transformation and have faced many of these challenges
ourselves. At the heart of our transformation is the way we build prod-
ucts. Instead of the multiyear product cycle of traditional, on-premises
products, Dynamics 365 is delivered as a service that’s continuously
refined and always up to date. It also allows us to capture rich teleme-
try on bugs and usage to inform ongoing feature development.

Beyond its influence on the way we run our own business, this digital
transformation paradigm has affected how we think about success for
our customers. Microsoft’s vision has always been about democratizing

Implementation Guide: Success by Design: Guide 1, Introduction to Implementation Guide 5

 software and technology capabilities to empower everyone. Looking
forward, we have renewed our purpose in the context of the cloud in a
world where everything is a source of data and potential insights.

Microsoft’s mission is “to empower every person and every organi-

zation on the planet to achieve more.” Harnessing data to help you
predict the future and intelligently guide actions represents a core
capability that invites organizations to achieve more by innovating
Fig. 1-3
everywhere (Figure 1-3).

Innovate everywhere

Suppliers Assets Operations Employees Products and Customers

Create a resilient, Get the most from Optimize your Empower your services Delight your
efficient supply your assets. operations. people and teams. Connect your customers from
chain. products and first touch to
services. delivery to success.

Microsoft’s data-first
cloud strategy
The opportunity contained in the shift from reactive to predictive—
knowing that the car is on its way to failure before it fails—is hard to
overstate. For Microsoft and for our customers, every product, ser-
vice, and business process is ripe for reinvention. But the reinvention
requires a wholesale reimagination—a literal turning upside down—of
the systems that power the interconnected fabric of products, services,
and applications.

To help companies reinvent their businesses and thrive in the predic-

tive era, Microsoft has invested in applications (such as Dynamics 365
Sales), an application platform (Microsoft Power Platform), and an
infrastructure platform (Microsoft Azure) spanning from enterprise and

Implementation Guide: Success by Design: Guide 1, Introduction to Implementation Guide 6

 independent software vendor (ISV) developers to citizen developers.

The entire Microsoft cloud (Figure 1-4) comes together as a unified

digital-transformation platform with consistent security, identity, and
compliance boundaries.

With everything as a source of data and potential insight, Microsoft’s

goal is simple: help customers take the right action at the right time
with the right message via the right channel to achieve the right
Fig. 1-4 business outcome. (Throughout this book, you
Microsoft cloud may find that we refer to this concept as the digital
feedback loop, as shown in Figure 1-5. It demon-
strates how organizations tap into data signals from
Microsoft 365 Microsoft Dynamics 365 across their business, and use those signals to drive
ongoing improvements in the products, services,
and experiences delivered to their customers. The
Power Apps Power Automate Power BI Power Virtual Agents customers then respond to these improvements,
Microsoft Power Platform thus fueling a self-reinforcing cycle.)

Microsoft Azure Dynamics 365:

Identity, security, management, and compliance
Cloud-based
business apps
Fig. 1-5
for holistic,
Engage
customers
Empower
people
unified digital
Intel
l i ge nt
transformation
Data & AI
Keeping the opportunities of the predictive era top
of mind, we want to take a step back and focus on
s
sy

te
ce

ms en
s

i
an d ex per
what Microsoft got right with Dynamics 365, further
Optimize Transform explain our data-first strategy in terms of Dynamics
operations products 365, and provide advice to customers embarking on
an implementation of Dynamics 365.

Implementation Guide: Success by Design: Guide 1, Introduction to Implementation Guide 7

 Looking back, it’s clear that Microsoft’s companywide push into the
cloud set in motion the trajectory of what we know today as Dynamics
365. But also notable was Microsoft’s decision to embark on a business
applications journey into the cloud with its previously named Dynamics
CRM and Dynamics AX products.

Although they were separate applications at the time, Microsoft’s

decision to move forward with Dynamics CRM and AX foreshadowed
the importance we place on developing a cloud-based business
application solution that offers customers a holistic, unified digi-
tal-transformation platform covering the front office, the back office,
and beyond.

Fast forward to today, and few of Microsoft’s competitors can offer a

complete cloud business application with the breadth, depth, and level
of integration with existing Microsoft products that Dynamics 365 has.

Dynamics 365 is a portfolio of business applications that meets orga-

nizations where they are—and invites them to digitally transform.
Each app is designed to remove the barriers and eliminate silos within
organizations by working together with existing systems—or the entire
portfolio of Dynamics 365 apps—for outcomes you simply can’t get
unless every part of the business is connected seamlessly.

Microsoft’s data-first
Dynamics 365 strategy
Dynamics 365 is a If you’re reading this book, then your business likely has already invested
in Dynamics 365. However, understanding Microsoft’s future vision for
portfolio of business Dynamics 365 will help you take full advantage of your investment.
applications that
meets organizations Building on Microsoft’s cloud strategy (which includes Dynamics 365),
several other factors contribute to Microsoft’s data-first strategy to put
where they are—
Dynamics 365 into a category of one:
and invites them to ▪ Dynamics 365 provides front-office and back-office cloud applica-
digitally transform. tions that are consumable in a composable manner, which means

8
 that our customers can maximize their investment without having
to do an extensive rip and replace of what were historically behe-
moth customer relationship management (CRM) and enterprise
resource planning (ERP) processes.
▪ Dynamics 365 is built upon the low-code Power Platform to
enable pro and citizen developers to build solutions, automate
processes, and generate insights using Power Automate, robotic
process automation (RPA), and Power BI—which no other vendor
can offer with its core platform.
▪ All of these are natively built on Azure, the cloud with an unmatched
level of security, trust, compliance, and global availability.
▪ Add to this the assets of Office 365 (such as Microsoft Teams,
LinkedIn, and Bing), which can be used to enrich the business
application experience.

With Dynamics 365, organizations have the only portfolio of intelligent

business applications that enables holistic, unified digital transforma-
tion and empowers everyone to adapt and innovate—anywhere.

Advice to customers
embarking on an
implementation of
Dynamics 365
Microsoft’s ideas on what customers should keep doing—whether
The ability to they are embarking on an implementation of Dynamics 365 or any
start quickly in other business application—are detailed later in this book, but they
the app allows include many of the traditional activities, such as business-process

organizations to definition, quality requirements management, end-user readiness, and

change management.
bring value to
their business in However, Microsoft recommends against project teams that struc-
an accelerated ture their engagements as massive 9-month to 18-month waterfall
efforts. From a cloud services perspective, this old, on-premises
fashion.
9
 implementation style needs to be jettisoned in favor of agility.

The ability to start quickly in the app allows organizations to bring

value to their business in an accelerated fashion. The focus of project
teams should be on finding the function or area of the application that
the business can start using as quickly as possible, and then expand
and enhance from there. The application and the business are going
to change anyway, so it is in the best interest of organizations to claim
value early.

More pointedly, Microsoft’s view is that being agile does not mean
engaging in sprint cycles that still require a 9-month to 18-month team
effort to deploy part of the application to end users.

Traditional business application models can no longer keep up. Long

setup times, expensive and complicated customizations, and slow
adoption curves make it nearly impossible to meet the needs of a
changing and demanding customer base.

Think beyond the

implementation—and
beyond the technology
Organizations that The key question as you embark on an implementation of Dynamics
understand the 365: why are you doing it? Is it just a rip-and-replace effort, or does the
power of data—and implementation represent an opportunity for digital transformation,

want to harness it advance your business model, allow you to leapfrog over the competi-
tion, or something else?
so that they can
disrupt the market— In making an investment in business applications, organizations are
are the business called upon to clearly identify the value they intend to drive toward.
Will the technology get you there? Does the software or the imple-
application projects
mentation partner have the vision to get you there?
that Microsoft wants
to be a part of. An investment in business applications requires a level of organizational

10
 maturity to think about what you and your business are after. It also
requires you to think beyond just the implementation of technology.

Organizations that understand the power of data—and want to

harness it so that they can disrupt the market—are the business appli-
cation projects that Microsoft wants to be a part of. This is where we
believe the world is going, and our products reflect that.

The journey forward

We understand that every organization comes with a unique set
of challenges, opportunities, and priorities. No matter where you
are in your digital transformation, Implementation Guide offers the
guidance, perspective, questions, and resources to help you drive a
successful Dynamics 365 implementation, and harness data for insights
that lead to the right actions and the right business outcomes across
any business process.

The first section of Implementation Guide provides an overview of the

Success by Design framework, which Microsoft created to help our
customers and partners drive successful implementations of Dynamics
365. Additional chapters in the first section focus on the mindset and
thinking required to implement in the cloud.

Subsequent sections of Implementation Guide address the major top-

ics in the Dynamics 365 implementation lifecycle, including Microsoft’s
point of view on solution design, environment strategy, data, security,
business intelligence, integration, performance, user readiness, sup-
port, and much more.

Whether you’re a first-time implementer or a been-there-done-that

Dynamics 365 architect, we hope you’ll bring us along on your journey.

Implementation Guide: Success by Design: Guide 1, Introduction to Implementation Guide 11

2 Guide
Success
by Design
overview
Implementation Guide: Success by Design: Guide 2, Success by Design overview 12
 Success by
Design overview
This chapter introduces Success by Design, a
framework and practice created by Microsoft to
help project teams implement Dynamics 365.

Based on thousands of real-world customer projects, Success by Design

is the sum of our Dynamics 365 implementation experience. It offers
topic-specific reviews and prescriptive guidance (approaches and rec-
ommended practices), which provide a reliable path to project success.
Success by Design is intended to be used by Dynamics 365 system inte-

One of the primary grators, independent software vendors (ISVs), and customers as a means
to better architect, build, test, and deploy Dynamics 365 solutions.
goals of this book is
to democratize the For our customers, Microsoft recognizes that Success by Design doesn’t
practice of Success guarantee implementation outcomes, but we’re confident that it will

by Design by making help you achieve your project’s goals while enabling the desired digital
transformation for your organization.
it available to the
entire community For our partners, Microsoft is confident that Success by Design, coupled
of Dynamics 365 with your implementation methodology and skilled resources, will
increase your team’s effectiveness in delivering successful Dynamics 365
implementers.
projects to your customers.

Implementation Guide: Success by Design: Guide 2, Success by Design overview 13

 One of the primary goals of this book is to democratize the practice of
Success by Design by making it available to the entire community of
Dynamics 365 implementers.

Success by Design history

As demand for Dynamics 365 cloud services increased across the enter-
prise, Microsoft identified the clear need to change the way we thought
about evolving our services and our responsibility to customers and
partners. We recognized that it wasn’t enough to design a platform
containing a set of features. We needed to also understand what it
takes to deliver a fully functioning, end-to-end solution that runs the
mission-critical processes of our customers’ businesses. From this need,
Success by Design was born.

Microsoft believes that customer success is the precursor to every

Dynamics 365 product development decision. Because of this, questions
fundamental to product development now challenge our engineers at
every step of the process:
▪ In what way must each product feature be delivered to ensure
customer success?
▪ How do we make Dynamics 365 apps valuable, durable, reliable,
and performant?
▪ What aspects of the product should be jettisoned because they
don’t meet fundamental customer success criteria?

For the first time in the product’s history, such questions have led to
the successful transition of 100 percent of Microsoft Dynamics 365

Microsoft believes online customers to one version, to a dependable and safe deployment
process, to a steady and coherent feature release cadence (including
that customer many overdue feature deprecations), and to a reliable and performant
success is the platform that delivers customer value. But product is only one half of the
precursor to every customer success equation.

Dynamics 365 Microsoft has put equal emphasis on the need to provide you the
product development prescriptive guidance and recommended practices to ensure a smooth
decision. implementation project and a properly designed and built Dynamics 365

14
 solution that successfully transforms business operations. This push has
also resulted in the transformation of Microsoft’s FastTrack for Dynamics
365 service, which makes certain that our community of Dynamics 365
implementers, customers and partners, has access to Success by Design.

This chapter focuses on the fundamentals and practice of Success by

Design and its desired result: Dynamics 365 projects whose technical
and project risks are proactively addressed, solutions that are roadmap
aligned, and project teams that are successful in maximizing their orga-
nization’s investment in Dynamics 365 cloud services.

Make Success by Design

your own
Enabling Dynamics 365 project teams to practice Success by Design
FastTrack for Dynamics 365 is a
customer success program run by means an ongoing commitment by Microsoft to provide the community
Microsoft’s Dynamics 365 product
of implementers, customers and partners, best-in-class business applica-
engineering team that helps
customers implement Dynamics 365 tions, along with the latest in Success by Design project resources.
apps and realize business value faster.
The practice of Success by Design is
fundamental to FastTrack’s approach, It also means project team willingness to incorporate the practice of
but the Success by Design framework
is available for use on any Dynamics Success by Design into their overall project approach, regardless of
365 implementation project. implementation methodology or the Dynamics 365 product being
implemented. At its best, Success by Design fortifies a project team’s
chosen implementation methodology with a model for product-aligned
project governance.

No matter where you find yourself in your Dynamics 365 implemen-

tation, fundamental to Success by Design is the willingness of project
teams, despite the common pressures of time, budget, and resources—
to pause in order to understand and address technical and project risks
before it’s too late in the project lifecycle.

To this end, this chapter focuses on the what and the why of Success by
Design, as well as how project teams can use it to accelerate customer
success throughout the implementation of Dynamics 365.

15
 Success by Design
objectives
Success by Design is prescriptive guidance (approaches and recom-
mended practices) for successfully architecting, building, testing, and
deploying Dynamics 365 solutions. Success by Design is informed by
the experience of Microsoft’s FastTrack program, which has helped our
customers and partners deliver thousands of Microsoft’s most complex
For interested customers, Microsoft
recommends that project leaders team Dynamics 365 cloud deployments.
up with their implementation partner
to enable Success by Design within their
project. In addition to the Success by Design Success by Design reviews are exercises in reflection, discovery (obser-
guidance available in this book, it’s highly
recommended that project teams ready vation), and alignment (matching to known patterns) that project teams
themselves by enrolling in the Success by can use to assess whether their implementation project is following
Design training on Microsoft Learn.
recommended patterns and practices. Reviews also allow project teams
to identify (and address) issues and risks that may derail the project.

Success by Design should be used as an adjunct to the project team’s

chosen implementation methodology for the following benefits:
In this book, we use the term “Finance and
Operations apps” in reference to Dynamics ▪ Reduced risk due to early detection of problems
365 Finance, Dynamics 365 Supply Chain
▪ Alignment to recommended practices
Management, Dynamics 365 Commerce,
Dynamics 365 Human Resources, and
Dynamics 365 Project Operations.
The result is a roadmap-aligned solution architecture that is performant,
scalable, and future-proof. ¹
Fig. 2-1

Success
by Design
phases
Initiate Implement Prepare Operate
Success by Design maps the
Dynamic 365 implementation
lifecycle into four methodology-agnostic phases: Initiate, Implement,
Prepare, and Operate (Figure 2-1). In this section and the following
¹ Acknowledging the possibility that feature sections, we outline the Success by Design phases, their relationship to
deprecations or other changes to product roadmap
in compliance with Microsoft policy may occur. Success by Design reviews, and the desired outputs and outcomes.

Implementation Guide: Success by Design: Guide 2, Success by Design overview 16

In the Initiate phase, the project team is in discovery mode,gathering
and validating business requirements, finalizing the high-level solution
approach, making inroads to define all in-scope workstreams, and up-
dating the project plan to reflect these updates. When the project team
has produced the high-level solution design and the related project
workstreams are more or less defined, Success by Design begins with the
Solution Blueprint Review. (More on the Solution Blueprint Review and
other reviews later in this chapter.)

In the Implement phase, the project team is focused on build-

ing the solution per the agreed-upon solution design and scope.
Implementation Reviews are introduced in this phase, having been in-
formed by the findings and recommendations of the Solution Blueprint
Review. As we learn later in this chapter, Implementation Reviews are
used to more deeply address questions related to the specific aspects of
the solution design (data model, security, integration) and implemen-
tation practices (ALM, testing strategy). Implementation Reviews are
meant to fully address the risks identified during or after the Solution
Blueprint Review but before the solution build is too far along.

By the Prepare phase, the solution has been built and tested and the
project team is preparing for the final round of user acceptance testing
(UAT) and training. Additionally, all necessary customer approvals have
been granted, information security reviews completed, the cutover plan
defined (including go/no-go criteria), mock go-lives scheduled, the sup-
port model ready, and the deployment runbook completed with tasks,
owners, durations, and dependencies defined. At this point, the project
team uses the Success by Design Go-live Readiness Review to identify
any remaining gaps or issues.

In the Operate phase, the customer solution is live. The goal of this phase
is stabilization and a shift in focus towards functionality and enhance-
ments that are earmarked for the next phase of the project.

Success by Design reviews

With a high-level understanding of Success by Design phases, we now

17
 turn to Success by Design reviews, which are also sometimes referred to
as Success by Design workshops.

Each review raises questions that serve as points of reflection that project
teams can use to generate important discussion, assess risk, and confirm
that best practices are being followed.

The Solution Blueprint Review serves as the starting point of Success

by Design. We suggest that the Solution Blueprint Review be a man-
datory review for the project because findings that come from it lead
to Implementation Reviews, which offer project teams the opportunity
to drill down into topic-specific areas where deeper dives are deemed
necessary for further understanding project and solution risk. Finally, the
Later chapters of this book address Go-live Readiness Review, which we also suggest as a mandatory review,
elements of Success by Design, but for the
most comprehensive coverage of Success is the last stop for assessing any remaining risks before go-live.
by Design framework, please refer to the
Success by Design training on Microsoft
Learn. Taken together, the guidance in Figure 2-2 illustrates that Success by Design reviews are not to be
this book, used alongside the Success by
Design framework, should equip project
conducted as abstract exercises separated from the project. Rather, the
teams for better, more successful project scheduling and implementation of each review relies on the availability
outcomes.
of key project artifacts and the readiness of the project team to discuss
them. (For a more in-depth look at Figure 2-2, visit the Appendix, which
provides Microsoft’s view of the typical deliverables, activities, tasks, and
Fig. 2-2 stakeholders involved in each Success by Design phase.)

Initiate Implement Prepare Operate

Get ready to start Design and build Deploy Live

Requirement analysis Code Testing and acceptance Solution health

Project governance Configure Go-live planning Usage

Fit gap analysis ... User readiness Maintenance

Customer kick-off Data modeling Cutover planning ...

... Solution performance ... Post Go-live Readiness Review

Solution Blueprint Review Integration Go-live Readiness Review

Success by Design
review and workshops
...

Implementation Guide: Success by Design: Guide 2, Success by Design overview 18

 Success by Design outputs
With a basic understanding of Success by Design’s objectives, phases,
and review flow, it’s important to pause to understand the makeup of
review outputs and their purpose, findings and recommendations.

As described in the previous section, reviews are informed by project ar-

tifacts (project plans, requirements, fit gap, solution architecture, design
documents, and so on) produced by the project team, and other infor-
mation such as formal or informal project touchpoints. The availability
of such information points to the readiness of the team to schedule and
conduct Success by Design reviews. Additionally, project teams may rely
on Microsoft for telemetry and other tools to inform review discussions
and generate review outputs.

The primary review outputs fall into two related categories: findings and
recommendations.

Findings come in three types:

▪ Assertions  Findings that capture noteworthy aspects of the
solution or approach. Assertions highlight what the project team is
doing right, typically in line with best practices.
▪ Risks  Findings that could potentially impact the implementation
negatively if not mitigated.
▪ Issues  Findings that are currently impacting implementation
negatively or will do so if not resolved.

Findings should include as much detail as possible and be matched

Fig. 2-3 to known patterns. Findings
matched to known patterns often
Matched to Lead to
Findings known patterns recommendations yield insights that lead to recom-
mendations or actions necessary
to resolve the issues identified
(Figure 2-3).

To better illustrate this point,

consider this customer example:

19
 A global corporate travel company is implementing Dynamics 365
Customer Service to drive its call center transformation. As the project
team digs into the business’s requirements, they learn that the Dynamics
365 solution must account for integration with multiple legacy systems
(many with high transaction volumes). Additionally, the requirements
point to a business-mandated customization that the project team
agreed could not be achieved using out-of-the-box functionality.

In preparation for the Solution Blueprint Review, the project team parses
these and other details, including confirming that solution performance
testing was purposely left out of the project scope on the assumption
that Microsoft’s Dynamics 365 cloud service should be performant on
its own. (Although it’s true that Microsoft is responsible for delivering a
reliable and performant cloud service to its customers, our experience
is that solution design and the resulting configurations, customizations,
and ISVs to achieve that design may play a role in impacting overall
Dynamics 365 solution performance.)

Considering that the Dynamics 365 solution in question is projected to

support 4,000 users at scale (including the multiple integrations and a
key customization mandated by the business), the project team’s find-
ings are clear: keeping solution performance testing out of scope is a risk
that negatively impacts the project. Among other findings and recom-
mendations not covered in this example, the project team’s architect
(who led the Solution Blueprint Review) recommends that the project
Steering Committee approve adding solution performance testing into
the test cycle.

The architect’s findings are summarized as follows:

▪ The solution requires custom development to meet its business
requirements.
▪ Performance testing is not included in the test cycle.
▪ Following best practices, solution performance testing should be
In Success by Design, findings link
observations to known patterns that invite added to the test cycle (and if necessary, the solution performance
actions necessary to address project risks
workshop should be scheduled to further explore the risk and report
and issues.
any additional findings back to the project Steering Committee).

Implementation Guide: Success by Design: Guide 2, Success by Design overview 20

 Track success measures
One more step remains in the Success by Design process: tracking
success measures. After completing Success by Design training, you
may decide to create your own success measure tracking template.
Otherwise, project teams can use Success by Design tooling (as served
by FastTrack or your implementation partner) to track success measures.
But what are success measures, and why are they important?

What are success measures?

By its very nature, Success by Design is a project-specific endeavor. For
anyone practicing Success by Design across one or many Dynamics 365
projects, the question arose: How do we measure the health of just one
of those many projects? Success measures allow us to do just that.

Success by Design empowers project teams to track the health of proj-

ects across seven categories and over 30 success measures. Figure 2-4
Fig. 2-4 highlights some of these categories and related success measures.

Review Findings Recommendations Tracking success measures is

simple: After a Success by Design
review or other compelling project
event, your FastTrack Solution
Architect or partner architect will
access Microsoft’s Success by
Success measures Design tooling and update the
relevant success measures for the
project. Success measure updates
Risk Assertion Issue are either red, yellow, or green,
and include project-related details
relevant to the success measure.

Why are success measures important?

Success measures are important because they provide access to micro
and macro project health trends. Tracking success measures for a single
project allows stakeholders to assess the overall health of the project at a

Implementation Guide: Success by Design: Guide 2, Success by Design overview 21

 glance. Similarly, the benefit of tracking success measures over 10, 20, or
100 projects is that Microsoft, the partner, or the customer project team
can see patterns that may be impacting them. For example, macro-level
tracking may yield an application lifecycle management (ALM) problem.
Access to this data allows Microsoft or the partner to understand wheth-
er it’s a project or product problem, and gain insights on how to fix it.

Conclusion
Success by Design equips project teams with a model for technical and
project governance that invites questions and reflection, which leads to
critical understanding of risks that might otherwise go unnoticed until
too late in the project.

Considering the pace of cloud solutions and the investment that organi-
zations make in Dynamics 365 software and implementation costs, even
the most experienced customers and partners with the best methodolo-
gies will benefit by incorporating Success by Design into their projects.

As covered in Chapter 1, Microsoft believes that every business is in

the business of creating great customer experiences. To achieve that,
business applications must do more than just separately run your back
office, marketing, supply chain, or even field operations. They must give
If you’re using Success by Design within you the agility to remove every barrier in your way. When this happens,
your Dynamics 365 implementation
project, we want to hear from you! To
business applications become more than just operational solutions. They
share your Success by Design feedback become resilient solutions that enable the digital feedback loop and
and experiences, reach out to us at
[email protected]. adapt to customer demands, delivering stronger, more engaging experi-
ences around the products you make and services you provide.

22
 Case study
An inside look at the
evolution of Success by Design
When our Solution Architects got their start with Dynamics 365, they
found themselves hooked on the challenge of effectively aligning an
organization’s business processes and requirements with the (software)
product in hopes of delivering solutions that resulted in actual value
for users.

The content of this case study is based on Because of this, FastTrack Solution Architects also know the challenge
interviews conducted with Dynamics 365
of working with people, understanding the business, shaping software
FastTrack Solution Architects. The goal is
to provide readers with an inside look at to meet the needs of the business, and delivering a solution that is
Success by Design. We hope you enjoy
this insider view.
accepted by users. They know that projects don’t always go well and
often fail.

It’s for exactly this reason that the FastTrack for Dynamics 365 team
and Success by Design were created.

A shift to Dynamics 365 in the cloud

Dynamics 365 isn’t software that is purchased and implemented as is.
As such, implementing Dynamics 365 products rely on qualified part-
ners to get the implementation right.

Complicating matters further, the promise of the cloud has been

that you can focus on the application and forget about everything

Implementation Guide: Success by Design: Guide 2, Success by Design overview 23

 underneath. But the reality is that cloud implementations require a
comprehensive understanding of how design and build decisions may
impact Dynamics 365 cloud performance, scalability, and more.

Accordingly, the shift to implementing Dynamics 365 in the cloud

remains challenged by on-premises implementation habits in which
the mindset is to overly customize, “because you can,” without impact.

The shift to implementing Dynamics 365 in the cloud requires not

just technical capability but also a definitive understanding of how to
design and build Dynamics 365 solutions within the constructs of the
online service.

The evolution of Success by Design

FastTrack for Dynamics 365 was formed to help customers maximize
their investment in Dynamics 365 by making sure that they (and
their partners) have access to the right information, recommended
practices, and a connection to the product group throughout the
implementation lifecycle.

When the FastTrack for Dynamics 365 team was formed, these goals
were initially achieved without the use of a framework. In other words,
FastTrack Solution Architects used their experience and connection to
the product team, but this approach contained too many variables to
be consistently reliable.

Over time, Microsoft’s experience revealed that Dynamics 365 custom-

ers were running into very similar problems. For example, FastTrack
would identify that some aspect of the solution was designed and built
in a manner that wasn’t compliant with recommended Dynamics 365
implementation practices. And unfortunately, too often the Solution
Architects would catch the issue too late in the implementation lifecycle.

As a result, Microsoft began to ask itself: How can we change the

FastTrack approach so our Solution Architects can address problems
before they manifest themselves as problems? FastTrack also acknowl-
edged that its approach couldn’t be “one Solution Architect solving the

Implementation Guide: Success by Design: Guide 2, Success by Design overview 24

 problems of one customer project.” Our Solution Architects needed to
work in a way that would benefit all customers.

To identify and address problems early, our Solution Architects had

to engage with project teams on multiple levels. We needed to un-
derstand project plans, how project teams were thinking about doing
customization, data migration, and more. This need for early visibility
across all dimensions of the project meant aligning to an approach
that afforded our Solution Architects the opportunity to get involved
in these conversations from the start, to identify the problems and risks
early, and to give the right guidance to the project team before more
problems arise.

This required a period of FastTrack experimenting with the right approach.

We started a concept that we referred to as checkpoints, which involved
using surveys, but we found that this approach kept us in a reactive mode.
We kept revising our approach and eventually we got it right.

From this experimentation, Success by Design was born: a proactive

practice that invites a purpose-built dialog that touches all levels and
dimensions of the project (technical and nontechnical) so problems are
As stated earlier in this chapter, Microsoft
recommends that project leaders team identified before they impact the project.
up with their implementation partner
to enable Success by Design within their
project. In addition to the Success by Success by Design gets to the heart of how project teams are thinking
Design guidance available in this book,
it’s highly recommended that project
and what they’re doing so that any risks can be identified and mitigated.
teams ready themselves by enrolling in
the Success by Design training on
Microsoft Learn. With Success by Design, FastTrack joins the customer, our partners, and
the product team in a manner that ensures alignment between all three.

Success by Design for every Dynamics 365

customer project
At some point during the evolution toward Success by Design,
Microsoft asked the question: How can we make sure every Dynamics
365 implementation is successful from the start? The key phrase is
“every Dynamics 365 implementation.”

As this chapter highlights, the FastTrack for Dynamics 365 team doesn’t

Implementation Guide: Success by Design: Guide 2, Success by Design overview 25

 have a special lease on the practice of Success by Design. This book’s
goal is to democratize Success by Design and make it available to the
entire community of Dynamics 365 implementers.

The more project teams invest in Success by Design, the more they will
get out of it.

Implementation Guide: Success by Design: Guide 2, Success by Design overview 26

3 Guide
Implement
cloud solutions
Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 27
 Introduction
The cloud, the constellation of connected servers, data
warehouses, and associated software and services, is at
the core of a digital transformation strategy.
Organizations of all sizes recognize that the cloud is the fastest way to
modernize their operations to increase efficiency while better serving
customers. For many organizations, the shift to the cloud simply means
running their applications in a public cloud infrastructure instead of
the traditional on-premises datacenter. Other organizations employ
a software as a service (SaaS) model that uses cloud applications and
an underlying cloud platform to provide deeper transformation and
Adopt a cloud mindset greater agility to adapt to future changes.

Cloud implementation Whatever model an organization chooses, we find that the key to
success isn’t what implementation looks like, but rather how leaders,
Customize and extend
cloud applications architects, and entire companies approach the digital transformation
journey. Success in the cloud isn’t just about the technology or the
Operate in the cloud features available; it’s about the organizational mindset. For a success-
ful digital transformation, your organization must prepare for changes
Evergreen cloud
that span the entire enterprise to include organizational structure,

Upgrade from on-premises processes, people, culture, and leadership. It’s as much a leadership
to the cloud and social exercise as it is a technical exercise.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 28

 In this chapter, we provide a perspective on what to expect when imple-
menting your cloud solution using Microsoft Dynamics 365. We introduce
principles relevant to any SaaS application. Concepts introduced here are
elaborated on in subsequent chapters. We hope these principles help you
set your course and guide you in your leap into the cloud.

We start by addressing how to develop a cloud mindset in an organi-

zation, which is foundational to your digital transformation. Then we
delve into factors to consider and understand before you shift to a
shared public cloud from an on-premises setup. These include how the
Supporting research and studies shift will impact controls you have in place and how to think differently
McKinsey about scalability, performance, shared resources, and more. After
The keys to a successful digital
that, we discuss the principles around customization and extending
transformation
cloud apps. Finally, we explore the operating model, always up-to date
Smarter with Gartner
Cloud Shift Impacts All IT Markets
evergreen cloud models, and the options to migrate to the Dynamics
365 cloud from on-premises.

Adopt a cloud mindset

Adopt a cloud mindset
Cloud implementation
Embracing change is a common theme, but infusing it throughout your
Customize and extend organization is vital to your digital transformation journey. Therefore,
cloud applications we include it as a thread that runs throughout this chapter and book.
Adopting the cloud mindset is also an intentional way to help organiza-
Operate in the cloud
tions break the mold and think differently about how applications are
designed, developed, secured, and managed in the cloud.
Evergreen cloud

Upgrade from on-premises First, let’s understand the “why” behind the pervasive adoption of SaaS
to the cloud business applications by organizations across the globe. Traditional
approaches to business applications mostly involve the IT department
building a custom application or using an off-the-shelf product aug-
mented with customizations to meet the specific business requirements.
The typical approach includes phases to gather requirements, build,
go live, operate, and then follow the change management process to
request changes or add new features. The IT department has full control
of the application, change process, releases, updates, infrastructure, and
everything needed to keep the solution running for years.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 29

 On the surface, this sounds ideal. It has been the standard operating
model for deploying and operating business applications. This way of
working, however, is losing relevance in today’s fast-paced world of
shifting supply chains and evolving customer needs and expectations.

Application development and deployment was disrupted not neces-

sarily because of the cloud or new technology, the undercurrent that
made the cloud pervasive was the changing landscape of business.
Unharnessed data from ubiquitous smart and IoT devices and other
sources fueled startups that disrupted whole industries seemingly
overnight. With that, the average lifespan of companies plummeted.
This inflection made the survival of a business dependent on delivering
the best product, the best customer experiences, and constant inno-
vation to meet ever-growing expectations. This new business arena
means businesses must change more often, evolve their processes
more quickly, and become more data driven to understand customer
needs, pain points, and preferences.

The changes we’re talking about aren’t triggered by the cloud or tech-
nology. These are thrust upon us by the changing business landscape.
Adopting a cloud mindset, then, is about transforming your processes,
people, culture, and leadership in such a way that your organization can
embrace change quickly and successfully. We believe the fundamental
organizational characteristics that will determine success in this environ-
ment come down to focusing on delivering business value through a
“Every company is a technology company,
secure cloud technology platform. This includes the ability to harness the
no matter what product or service it
provides. Today, no company can make, data that is captured to generate insights and actions. This platform will
deliver, or market its product efficiently
without technology.”
also rely on automation to quickly react to changing needs.

—Forbes, “Why Every Company is a

Technology Company” In this section, we explore these organizational characteristics that help
drive the cloud-first mindset to achieve digital transformation.

Focus on business value

Technology is a ubiquitous force that influences almost every business.
Indeed, the IT department within a bank or a manufacturing com-
pany may be larger than entire tech companies. Understanding how
technology affects customers and business is important to help any

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 30

 company survive and thrive. It’s also important to not let technology
hinder growth or impede business. The role of technology is to deliver
business value by driving efficiency.

However, we often see enterprises whose technological landscape has

become so large and complex over the years that they fail to deliver the
agility demanded by the business. The IT departments are consumed
with dealing with the complexity of software compatibility, update
cycles, end of life deadlines, aging infrastructure, and antiquated security
policies with little time to focus on delivering business value.

An evergreen SaaS cloud approach to your business application,

however, takes away much of the infrastructure and operational IT
effort to keep the servers running, software patched, and other routine
tasks, and turns the focus of IT on the business. Operating in the cloud
doesn’t mean there are no operational responsibilities, environment
IT has a greater management, updates to SaaS services, feature deprecations, user
opportunity to communications, or coordinating change management for enhance-
ments and new capabilities, monitoring, and support. These functions
focus on delivering are all still required, but with a key difference: IT is now much closer to
positive business the business application layer and has a greater opportunity to focus
outcomes. their energy on delivering positive business outcomes.

Traditionally, business applications depict a set of key business processes

that allows the users to capture, track, and report on the process and
its progress. These business processes usually take the form of having
a user interface (UI) built on top of an underlying data store to capture
or display the information from the database. These “forms over data”
software applications can help organizations keep track of information
and make it easily available through reports. However, from an end
user’s perspective, they’re a data capture tool that doesn’t really help
them do their job any faster or better. This design of business applica-
tions hasn’t changed in decades. It’s an outdated model that doesn’t
bring any additional value to the business or to the end user, and doesn’t
differentiate a business’s offerings and customer experience from their
competition, which is likely following a similar outdated approach.

Additionally, on an individual level, we want to automate routine,

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 31

 boring tasks and focus on more creative, challenging work.

Automate the For example, take a point-of-sale application that helps you address
your customer by first name and provides them personalized relevant
routine tasks so offers based on their loyalty, or recommends products based on their
you can focus on purchase history and Artificial Intelligence (AI). This will be better
creative, more received by your employees and customers than an application that

challenging work. can barely determine if a product is in stock or not. Simply offering the
best deal or best quality product isn’t enough in today’s hyper-com-
petitive environment. To win, you need to differentiate between your
customers, respond to their unique behaviors, and react to fluctuating
market demands.

Businesses likely already have the data to do this, but may lack the
technology to turn data into practical insight. For example, Dynamics
365 Commerce taps your systems to gather customer intelligence from
a myriad of sources (such as social media activity, online browsing hab-
its, and in-store purchases) and presents it so that you can shape your
service accordingly. You can then manage product recommendations,
unique promotions, and tailored communications, and distribute them
across all channels via the platform.

Dynamics 365 applications are focused on delivering value to the busi-

ness and empowering end users, changing the business application
industry paradigm by forcing every business to innovate faster.

The “If it ain’t broke, don’t fix it” mentality common in large organiza-
tions, where business applications can remain unchanged for up to 10
years, puts the enterprise at risk. Unless your business is a monopoly
and can sustain itself without investing in technology, the hard reality
is most businesses today won’t survive without strategic plans for
digital transformation.

You need to ask yourself some questions:

▪ Has the rest of your industry transformed and embraced technology?
▪ Do your customers demand a better and faster process?
▪ Is your competition growing their market share by eating into yours?
▪ Does your competition have a better Net Promoter Score (NPS)?

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 32

 If you answered yes to most or if you don’t know the answers, your
business might be failing to adapt and react to changes in customer
behavior and expectations. Businesses that don’t invest in technology
will face challenging disruptive competitors, either global behemoths
or innovative startups. You need digital transformation technology
that gives you the command of three essentials: deep, cross-channel
customer insight; synchronized operations from front end to back end;
and scalability to drive rich experiences as needed.

Decision-making Decision-making around such technology now carries existential

consequences. Lack of strategic visionary thinking, long-term planning,
around digital and matching investments can prove to be catastrophic.
transformation now
carries existential The critical point is your business application shouldn’t be designed

consequences. and deployed with an expectation that it will remain unchanged for
long periods of time. Instead, the core design principle for business ap-
plications should be ease of change. You should take advantage of the
wonders of automation to deliver changes as often as daily and weekly,
at the whim of business. This continuous development philosophy is
core to a Dynamics 365 SaaS application with weekly maintenance
updates and multiple waves of new features delivered every year.
Realizing the value of these cloud applications, therefore, is about
adopting the latest best-in-class capabilities, continuously delivered.
With this new mindset, SaaS services aren’t just used to meet business
requirements, they’re helping drive business value and adoption.

Increasingly, technology is becoming so powerful and democratized

that non-developers and people with no formal training in software
development are empowered to build applications, create workflows
that automate their work, and use machine learning with ready-made
AI models to analyze data. This concept of citizen application devel-
opment has drastically transformed businesses, where end users are
developing solutions to solve business problems and sharing these
applications with colleagues. This organic approach to app devel-
opment has short-circuited the application development lifecycle,
helping address the perennial issue of software adoption. IT depart-
ments are now responsible for the governance of these applications,
securing the access to data and environment, and focusing on building

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 33

 the underlying services and connector interfaces that the business can
consume in their applications.

These scenarios all help illustrate how a transition to the cloud can play
an instrumental role in refocusing technology to deliver business value.

Driven by data
This disruptive innovation we’re seeing across several industries is
fueled by data. Almost everything around us generates data: appli-
ances, machines in factories, cars, apps, websites, social media (Figure
3-1). What differentiates one company from another is the ability to
harness this data and successfully interpret the information to generate
meaningful signals that can be used to improve products, processes,
and customer experiences.

Organizations recognize data is an asset, and becoming data driven

is a key motivator for digital transformation. Still, the accounting
rigor that we see in finance is seldom observed when it comes to
managing organizational data. If data is the new currency, then we
should treat it like that. Get an understanding of your organizational
data estate, the databases, data warehouses, and data lakes, that help
Fig.
3-1 your company manage its cor-
porate data. Develop a holistic
Engage Empower view of customer data across
customers people departments, functions, and
applications, while acknowledg-
l i ge nt
Intel ing the regulatory boundaries
on data usage. These steps are
practically a prerequisite to digital
Data & AI transformation.
s
sy

te
ce

ms n
ie
s

an d ex per Depending on the volume of

data, its age, and organizational
Optimize Transform silos, the deduplication, correla-
operations products
tion, and conflation of data from
different systems could become a
complex project on its own.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 34

However, modern AI-powered customer data platform (CDP) technol-
ogies could come to the rescue, for example Dynamics 365 Customer
Insights. A CDP tool creates a persistent, unified customer database
that is accessible to other systems. Data is pulled from multiple sources,
cleaned, and combined to create a single customer profile.

It’s also important to define your organizational data strategy for

business applications, develop guidelines on how new applications
store their data, what schema they use, and how other applications in
the organization can understand and use this data. You need to use the
most optimal datastore for your use case. For example, keeping his-
torical credit card transactions in Dataverse (a repository to store and
manage operational data used by business applications) is suboptimal;
a data lake might be a better choice.

Being data driven is about better understanding your data’s quality

and relevance, and being able to generate valuable insights that are
actionable and can drive meaningful changes to your processes. Going
back to the point of embracing change, your data strategy should
reflect how you design your process, develop your applications, and
operate. If it takes several months to deploy a change request or years
to roll out an improved process, there is little point.

A data-driven business application cloud platform like Dynamics 365

enables you to:
▪ Create a data estate that is intelligent and AI-ready
▪ Connect and consume data from legacy on-premises systems,
other line of business applications, and third-party applications
▪ Have the necessary controls to secure and meet data compliance
requirements
▪ Build smart applications that are continuously improving
▪ Infuse data intelligence directly into the application, empowering
users to make effective and smart decisions
▪ Have agility so applications can quickly adapt to changes in your
process

Overall, the intelligence and insights you generate from your data
will be proportional to the quality and structure of your data. You

35
 can explore this topic more in Chapter 10, “Data management,” and
Chapter 13, “Business intelligence, reporting, and analytics.”

Think platform
An organization can take several approaches towards digital trans-
formation. In many cases, you start with a single application being
deployed to a SaaS cloud. A key responsibility of IT decision-makers
and enterprise architects is to deliver a cloud platform for their or-
ganization’s digital transformation. Individual applications and their
app feature sets are important, but you should also look at the larger
picture of what the cloud platform offers and what its future roadmap
looks like. This will help you understand if the platform can meet the
short-term and long-term objectives of your digital transformation.

You’re investing Thinking about the platform versus a single app offers clear benefits.
You avoid reinventing the wheel with each additional application,
in the platform, and you instead deliver a foundation approved by governing bodies,
not just the with clear patterns and practices. This approach limits risk and brings
application. a built-in structure that enables reuse. Platform thinking doesn’t nec-
essarily mean that other cloud platforms or legacy applications have
to be rebuilt and replaced. Instead, you can incorporate them as part
of an all-encompassing platform for your organization, with well-de-
fined processes and patterns that enable integration and flow of data
between applications. Bringing this “systems thinking” to deliver a
platform for business applications can drastically reduce the amount
of time lost getting security and design approvals for individual apps,
thereby improving your agility and time to value.

Your business applications platform should already provide the following:

▪ Necessary security and design approvals for processing data
in specific categories based on its sensitivity and regulatory
requirements
▪ Clear guidelines on extending and customizing the platform to
ensure supportability and agility
▪ Established governance process to manage access and service
operation
▪ Approved integration patterns for communication with other

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 36

 platforms and on-premises systems, with a clear process for ap-
proving deviations
▪ Data storage guidelines for apps and patterns for sharing
▪ Clear guidelines for data ownership and access
▪ License entitlements that are available for individual apps
▪ Documented service protection limits and best practices that your
applications should comply with
▪ Resources for additional learning and guidance with access to experts

The due diligence and foundational work you do when choosing a

platform in coordination with IT and business decision-makers can save
weeks, if not months, of effort downstream when developing individual
apps. This thorough evaluation also helps drive predictable success.

Embrace DevOps and automation

Due diligence
Traditionally, the process to manually release a change to applications
now helps drive was convoluted, involving several teams, change boards, manual
predictable success regression tests, and approvals. It took weeks. This complexity makes
in the future. teams averse to change. Enhancements and updates are then con-
sistently deferred, which affects the adaptability of your applications
to shifting business and technology priorities. Automating builds
and releases and building in testing automation is crucial to allow for
continuous software updates.

The biggest selling point of the cloud is to stay current and quickly
deploy software. You can’t rely on teams to manually test and deploy
each and every change to achieve that vision.

Traditionally, a time lag existed between code and test and deploy.
A bug in production meant repeating the lengthy cycle. The fear of
breaking code meant that teams tended to delay updates for as long
as possible. With automation, you deploy fast and potentially fail fast.
This is where the culture of fail fast comes in. Automated processes
help companies empower their teams to take risks. You fail but quickly
release a fix. Over time, failures decrease and success rates improve,
instilling confidence in teams to deploy and innovate, and deliver-
ing value to the business faster. At the core of the cloud mindset is

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 37

 understanding and embracing DevOps. DevOps is the union of people,
process, and technology to continually provide value to customers.
DevOps is a cultural change in which we break down silos between devel-
opers and administrators, and create shared responsibility for the quality
of released software. The DevOps process automates software delivery
and speeds up the technology deployment lifecycle (Figure 3-2).

Fig.
3-2

Plan

Monitor Code

DevOps
Operate Build
processes

Deploy Test

Release

The idea is to enable development, IT, and quality and security teams
to collaborate to produce more reliable and quality products with the
ability to innovate and respond more quickly to the changing needs of
business. Implementing DevOps inherently depends on automation:
automating the build process to enable continuous deployment,
regression tests to improve speed and reliability, and administrative
actions like user and license management.

At this point, we should also explore continuous integration (CI) and

continuous delivery (CD) in the context of the cloud. The idea is each

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 38

 code check-in from the developer should go into the main build and
get tested early. This takes CI to the next level by allowing for automat-
ed testing and continuous, automated code deployment.

If a change includes a bug, the build fails, and the cycle repeats. The
automation allows for changes to happen quickly without unneces-
sary overhead, a team can focus on the business logic and not the
infrastructure. Therefore, with CD we always have the most up-to-date
working software.

CI and CD bring agility to cloud solutions by allowing for the latest

changes and bug fixes to move to production code without delay. This
is a paradigm shift from the world of on-premises software, in which
each change had to wait until all user stories were completed, the full
package was deployed, and a testing team tested the change. With
automation, we’re reducing time and improving quality.

But this isn’t just a technological change. DevOps requires close collab-
oration between development, testing, and deployment teams. This is a
cultural and mindset change in which the whole team trusts each other,
feels empowered, and is collectively responsible for product quality.

Preparing your organization for the fast-paced world of the cloud

means that automation isn’t optional anymore. Instead, investments
in automation will deliver significant efficiencies and help improve the
overall reliability of your applications.

Adopt a cloud mindset

Cloud implementation
Cloud implementation
Customize and extend considerations
cloud applications
Implementing solutions from cloud SaaS products can reduce man-
Operate in the cloud
agement overhead and allow you to deliver business value more
quickly by focusing on the application layer. Fundamental application
Evergreen cloud
considerations, however, still apply to SaaS cloud applications. Security,
Upgrade from on-premises scalability, performance, data isolation, limits, and capacity are still
to the cloud critical, but could have a different approach when compared to an

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 39

 application deployed on-premises. This section focuses on some of
these considerations for SaaS cloud apps.

Security readiness
The responsibility for security in SaaS cloud applications is shared by
both the service provider and the customer. That means your existing
security policies might not be suitable to meet the security require-
ment in the cloud. The SaaS service provider processes customer data
and is responsible for aspects such as securing the backend infrastruc-
ture and data encryption in transit and at rest. As a data controller, the
customer is still responsible for securing access to environments and
application data.

The IT information security team should clearly understand the bound-

Dynamics 365 offers deep integration aries of these shared responsibilities to ensure the following:
with Azure DevOps with build tools to
▪ The SaaS service provider meets the organizational security,
help automate your deployment pipeline
and effectively manage the application privacy, and compliance requirements. This is usually done in the
lifecycle (Figure 3-3).
beginning to approve the platform for use supported by a regular
review or audit process to ensure continued compliance.
Fig.
3-3

ALM Powered by Azure DevOps

Initiate Build Release
Getting started, faster Build and walk away Automated, predictive, repeatable

Pack source Run

Provision Deploy Import Run solution Export Run solution Import as Increment Export
code from Run unit test Unpack repo Run unit test Pack solution integration
environment dependencies solution checker solution checker unmanaged version managed
repo test

Initial build pipeline instantiates pristine Building pipeline automates manual steps. No more Automated release pipeline removes manual steps. Weekly, daily, or hourly releases
development environment daily. need to upload to solution checker and manually become the new standard.
export solution, unpack, and push to repo.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 40

 ▪ The security team is aware of the controls and configurations
required to govern access to data and fulfill the customer se-
curity responsibility. These could include defining the data loss
prevention policies, access control policies, and environment man-
agement policy. They usually have a default setup with a process
to manage deviations for specific apps.
▪ The governance process makes sure the application-level security
requirements are met. This is primarily managed by the appli-
cation teams and driven by business requirements for each app
deployment. You might have additional checks before deployment
to ensure compliance.

A good example of this is General Data Protection Regulation (GDPR)

in the EU, in which the SaaS service itself might be fully GDPR com-
pliant from a data processor standpoint but the customer is still
responsible for implementing data controller processes like a “forget
me” request or managing marketing consent for contacts.

Organizations need to review their information security policies and

ensure that they’re cloud ready. Doing this exercise earlier, before the
app deployment cycle, will help avoid delays. It’s important to closely
work with your cloud SaaS service provider to ensure all your security
requirements are met. Microsoft publishes the details of how this
shared responsibility of securing data is fulfilled from a data processor
perspective on the Microsoft Trust Center, which provides detailed in-
formation on our security and privacy policies as well as the certificate
of compliance for various regulatory norms and internal standards.

Scalability
The scalability of the SaaS platform is a key consideration for business
applications. Being able to scale out and scale up to support seasonal
workloads or spikes in user activity will impact the overall user experi-
ence (both staff and customers) and effectiveness of business processes.

The cloud scalability parameters available in SaaS differ from a

traditional on-premises system. Instead of adding more servers
or increasing the power of machines, these parameters could be

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 41

translated as available API capacity. You also shouldn’t assume that
cloud means infinite scale and computing power that can process
everything thrown at it. The good old coding and design best practices
are still relevant, but might need to be adapted. Although managing
and scaling cloud services is complex, increasing or decreasing your
data storage, processing power, and networking requirements can be
done seamlessly. In many cases, you can do so automatically or with
simple configuration changes. This microservices architecture, includ-
ing capacity-based routing to resources and storage management, is
transparent to Microsoft customers.

To summarize, the SaaS cloud offers the flexibility to scale horizontally

and vertically to support thousands of concurrent users. Embracing the
capacity-based model for resource consumption in the cloud helps not
only build optimized applications but also a better plan for operating
cost post deployment.

Latency and performance

Performance is a key consideration for business applications, it not only
impacts the end user experience and adoption, but can directly impact
business goals and key performance indicators (KPIs) of success.

In on-premises deployments, enterprises had complete ownership

and control of the infrastructure, and could guarantee and monitor
applications’ latency and performance. In the cloud world, it’s not as
straightforward.

Research shows that a few milliseconds of latency lead to a big per-

centage drop in page load times. For e-commerce companies, this
could mean a sizable drop in user attention and therefore sales reve-
nue. Low latency is not a good to have, but a critical deciding factor in
an enterprise’s brand and reputation. The same applies to back office
business applications operating in the cloud, user experience and pro-
ductivity can be significantly impacted in a high-latency environment.

Several contributing factors can impact application performance,

including network latency, user device and browser, application

42
 design, and customizations. SaaS services are inherently scalable, have
vast compute power, and are available from multiple locations around
the world, but that doesn’t necessarily guarantee performance if the
solution is poorly designed, or if users are accessing the service from
environments with high network latency.

Your cloud applications are powered by hundreds of pieces of virtu-

alized hardware running on Azure datacenters sprinkled around the
world. A crucial decision you must make, then, is choosing the data-
center to deploy your application so that all users of the application
get an acceptable level of latency.

Performance also can suffer because of a poorly designed application.

If you’re changing standard code in SaaS or deploying custom applica-
tions in platform as a service (PaaS) or infrastructure as a service (IaaS),
your team must design a high-performing application and carry out
proper performance tests, and only then do you deploy to production.
Now, most cloud providers work on a shared network infrastructure in
which your requests go through the internet and are used and shared
by everyone.

Network latency is a crucial factor to consider alongside other architec-

tural and design decisions that impact performance.

One option for businesses looking for more reliability and security is
to use a private connection. Cloud providers offer dedicated channels,
You can test latency using the Azure for example Azure ExpressRoute. These connections can offer more
Latency Test and Azure Speed Test 2.0 for
each datacenter.
reliability, consistent latencies, and higher security than typical connec-
tions over the internet.

Isolation in the shared cloud

The cloud operates on a shared common infrastructure across different
businesses, which leads to economies of scale. But although the infra-
structure and processing may be shared in public clouds, you have data
isolation, which means one customer’s data isn’t shared with others.
Security and data use policies are taken very seriously by Microsoft; it’s
imperative that we maintain the highest data and security standards.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 43

 These standards and practices are audited and certified on a regular
basis by trusted third-party organizations. Security isn’t an afterthought,
but ingrained into how we develop applications, so that the computing
infrastructure is shared and other security related aspects are isolated.

Private connections The other aspect of sharing resources and infrastructure in the cloud
is the possibility of monopolizing resources or becoming a noisy
offer reliability, neighbor. While rare, these cases are more often caused by poorly de-
consistent latencies, veloped runaway code than a legitimate business use case. Automated
and higher security monitoring and protection throttles are built into the Dynamics 365
platform to prevent such situations, so it’s important to understand
and comply with these service boundaries and throttles when design-
ing for cloud platforms.

Service protection
and a capacity-based model
Service protection and limits are used in cloud services to ensure
consistent availability and performance for users. The thresholds don’t
impact the normal user operations; they’re designed to protect from
random and unexpected surges in request volumes that threaten the
end user experience, and the availability and performance characteris-
tics of the platform or solution. It’s crucial to understand the protection
limits and design for them with the appropriate patterns, especially
around high-volume workloads like integrations and batch processing.

The cloud provides us the scalability to deal with large workloads, but
Plan for the daily peak and the monthly
maximum order transaction volumes a shared public cloud doesn’t mean you have unlimited capacity or
expected to ensure that the service is
licensed to satisfactorily support the peak
computing power. In many cases, these capacity limits are enforced
as well as the total maximum expected via licensing. You have to embrace this capacity-based model of cloud
volumes. Also, plan and prioritize for
integrations and Open Data Protocol infrastructure and plan to operate within the entitlements, taking
(ODATA) requests based on the volumes,
into account usage profile, data volumes, and integration patterns.
so they’re honored and not throttled
due to request limits. These checks and Understanding these protections and service boundaries that apply
balances prevent overutilizing resources,
preserve the system’s responsiveness,
to a specific service helps bring clarity on available resources. It also
and ensure consistent availability and drives the right behavior when it comes to design and development so
performance for environments running
Dynamics 365 apps. the solution is optimized to operate within the allocated capacity, or
additional capacity is procured to meet the requirements in advance.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 44

 Check out Chapter 20, “Service the solution,” for more information.

Integration with on-premises systems

The SaaS cloud might be the preferred approach for new applications,
but in most enterprise solutions, you could still have on-premises
elements. Therefore, it’s common to build integrations with these
on-premises systems. Establishing these integration patterns in
compliance with your internal security policies while allowing cloud
applications to authenticate and connect to on-premises services is
critical. This could involve allow listing URLs and IP ranges on your
firewall, or using technologies like the Azure AD Application Proxy or
an on-premises gateway that allows such integrations without having
to open inbound connections into your customer’s corporate network.
Patterns that
Even from a design standpoint, patterns that worked on-premises
worked on-premises could be challenging in the cloud, due to latency and additional secu-
could be challenging rity restrictions imposed. An example is the encryption of traffic leaving
in the cloud. a customer network or routing via a peering network.

Cloud and edge

The public cloud is the default and recommended choice of deployment
for the majority of Dynamics 365 customers, but you may want to extend
the cloud to unlock important capabilities for key scenarios where con-
nectivity is an issue. Your industry may need a solution on the ground,
such as brick and mortar commerce, field service, a manufacturing
execution system (MES), warehouse operations, and project operations.

These implementations are called cloud and edge, where cloud is the
system of intelligence and edge is the system of record.

For example, Dynamics 365 Commerce customers use the cloud and
edge model, in which the store operations run on-premises (edge)
and the main instance handles centralized back office functions like
finance, data management (such as products and customers), and
analytics (cloud).

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 45

 Cloud and edge deployments provide business continuity for
mission-critical functions like point of sales, warehousing, and manufac-
turing. We want you to be able to continue running operations when
disconnected from the cloud. In case of a network outage, Dynamics 365
Point of Sales can sustain itself with basic operations and keep data gen-
erated locally. When connectivity resumes, data is synchronized to the
cloud. Also, whenever throughput is high and more heavy processes are
run in parallel, user productivity may be impacted. We want to enable
you to scale out and run manufacturing and warehouse processes in
isolation so high user productivity is always supported.

Another reason for choosing an on-premises deployment model is

that these businesses have made significant investments in local IT
infrastructure. We want to ensure you receive a meaningful return on
investment (ROI) before taking the full cloud route.

Another reason is internet connectivity. Regions with poor internet

connectivity see more occurrences of on-premises deployments.
Organizations operating in such regions go with an on-premises model
to mitigate the risk to their business from poor connectivity. They also
want to maintain a stable user experience for their staff.

Adopt a cloud mindset

Customize and extend
Cloud implementation
cloud applications
Customize and extend
cloud applications A SaaS application with rich features also provides powerful capa-
bilities to customize and extend the app to meet specific business
Operate in the cloud
requirements. Customizations to the SaaS application can be at several
levels. Many involve little code (embracing the low-code/no-code
Evergreen cloud
philosophy) and offer the flexibility to tap into the underlying PaaS
Upgrade from on-premises layer. This allows you to take advantage of professional software devel-
to the cloud opment techniques that include custom code.

Chapter 15, “Extend your solution” covers the details of customization

and app extensions, considerations, and scenarios. In this chapter, we
focus on the principles to keep in mind when making decisions about

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 46

 customizing the application or extending platform capabilities. While
these principles broadly apply, in some cases a tradeoff is necessary,
in which the approach depends on specific situations, in-house skills,
timelines, budget, user expectations, processes, and other factors.

Know the platform

When using a SaaS cloud platform like Dynamics 365, it’s important
to know that this isn’t a one-time software purchase. It’s an ongoing
partnership between the customer and service provider, with an
operating contract that governs the service-level agreement (SLA),
security, ongoing updates, and more. You need to understand how the
service runs and how updates and changes impact your solution (and
therefore business), and employ the supported way of extending and
customizing the application in harmony with the service contract. With
this knowledge and compliance, you can ensure continued support
and avoid unexpected issues.

Using a SaaS cloud Even when using supported extension techniques, you need to adhere
to best practices. A good example is that while the platform might
platform is an allow you to run synchronous code for up to two minutes when creat-
ongoing partnership ing a record, running to the time limit would block a user’s UI thread
between customer for two minutes when they save a record. Is that acceptable? Similarly,

and service provider. you could use custom code to limit access to records and implement
a custom security requirement, but certain access mechanisms could
bypass your custom code, leading to exposure. The point is to carefully
design your customizations and assess their impact, particularly when
those customizations affect the end user’s experience or deviate from
the security control constructs provided by the platform.

In general, the low-code/no-code techniques that rely on configu-

ration rather than custom code are optimized by the platform, and
should be a preferred approach. However, custom coding may be
required in some use cases. Dynamics 365 has a no-cliffs philosophy
with limitless possibilities to extend the application using the underly-
ing platform. The ability to address this by using the underlying PaaS is
a superpower in the world of SaaS applications.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 47

Be aware that custom coding and PaaS extensions bring additional
maintenance responsibilities and are best left to professional devel-
opment. Even when using PaaS to extend your solution, serverless
technologies are usually better suited to this approach.

Other factors include solution capacity planning, which should con-

sider the impact of your customizations, and the implications of using
certain PaaS technologies that use a pay-as-you-go model. We recom-
mend your organization establish guidelines for customizations that
take into consideration the platform, user experience, security, and
approved patterns for integrations. In addition, enforce best practices
with automated checks wherever possible in the build pipeline. Power
Platform includes an app checker and app monitor that can help
identify unsupported techniques and anti-patterns that can lead to
performance and scalability issues.

Don’t mimic your existing system

A common misstep is trying to mimic a legacy system or the features of
a different SaaS platform when implementing a solution. In other words,
lifting and shifting the current on-premise application experiences into
the cloud. This mindset limits transformation to the technical aspects
of how the application is operated. Attempts to recreate on-premise
apps in a SaaS service can lead to excessive customization, prevent you
from realizing the true potential of the platform, and limit your access to
future innovation in the standard service.

To focus on business value, you should adopt a business process–

oriented approach. In this approach the implementation focuses on
supporting business processes rather than providing specific technology
features. It’s important that business requirements aren’t muddled with
implementation details or dictate a specific approach, but focus on a
business outcome instead. The final transformed process is facilitated
through leveraging platform capabilities instead of replicating the
existing solution. This puts focus on the core process and is future-proof,
less costly, and easier to maintain than simply lifting and shifting your
existing experiences to the cloud. You also avoid technical debt from ex-
cessive customizations. Aligning to the platform allows you to maximize

48
 its capabilities so you get an evergreen, always up-to-date solution.

Future-proofing
One of the key principles we have established in this chapter and through-
out this book is getting comfortable with change. SaaS enables you to
adopt new features to maintain a competitive edge. These features and
capabilities are built on top of an existing baseline of features and tables.
Although repurposing or using a custom table may not be unsupported,
deviations can severely impact your ability to take advantage of future
capabilities or can affect the interoperability of your solution.

In addition to the business process-oriented approach to implemen-

tation, we also recommend that you review your existing business
processes and compare that with what the standard application has to
offer. You can benefit from adopting standardized processes backed
by research and feedback from leading customers in their industries,
because you benefit from all that expertise. Using standard processes
also makes your business less dependent on expensive and hard-to-
maintain customizations.

The Common Data Model (CDM) is an open standard schema for

business data created in collaboration with SAP and Adobe, and is
used across Dynamics 365 applications. Aligning to the CDM can help
future-proof your solution. It can also make sure you can take advan-
tage of enhancements as well as other business applications across the
industry that honor the schema and can easily interpret your data.

Another way to future-proof is to understand your roadmap, in-

vestment direction, and trends so you can better align to the future
direction of the platform. This helps reduce technical debt and rework.

Understand the tradeoff

Expect tradeoff decisions to be made when implementing your
solution, particularly when it comes to the user experience (UX). In
Dynamics 365, the platform provides several out-of-the-box controls
and UX patterns that can be reused in a model-driven paradigm to

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 49

 quickly assemble applications. These address common usage patterns
and also meet accessibility standards. It’s common for a business to
make specific requests around application navigation, behavior of a
specific control, or even colors to highlight specific controls. From a
technical perspective, most of these UX needs can be met either by using
custom code to build custom controls or by using a different app para-
digm,in which you start with a blank canvas to develop a pixel-perfect UI
that you could use in combination with model-driven controls.

Several factors could guide these decisions, including user adoption,

but you should also assess the cost-benefit ratio of customization.
In some cases, building and maintaining custom controls can be
time-consuming and require specialized skills. If you can easily address
some scenarios with user training, you can save hundreds of hours of
development and test effort. In addition, specialized UX requests are
often subconsciously driven by the need to replicate an existing system
or just a familiar pattern, even though the requests defy modern UX
best practices and accessibility needs.

Independent software vendors

Independent software vendors (ISVs) are third-party organizations that
exclusively develop software or software solutions that are compatible
with the service to deliver additional capabilities.

An effective way to approach extending your solution is to use an ISV

solution from the app marketplace for your platform. This can save
time and effort in development and testing, and provide you with a
tried and tested solution used by peers in your industry.

Several industry-focused ISVs provide solutions to fill unique product

gaps and address specific business needs in industries such as fashion
and hospitality. Make sure to perform due diligence when looking for
an ISV. For example, an ISV needs to make sure their solution is avail-
able for the version you’re targeting for go live. Otherwise, deployment
timelines can be affected. It’s also important that you’re aware of the
support model for an ISV, their commitment to keep it up to date, and
the protection clause if the solution provider goes bankrupt or is sold.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 50

 You will also want to know how the ISV handles deprecation notices for
phased out software and update cycles.

Chapter 15, “Extend your solution,” provides a deeper dive into making
a solution work for you.

Adopt a cloud mindset

Operate in the cloud
Cloud implementation
Successfully operating in the cloud will drive business value for your orga-
Customize and extend nization. Having the right skills and the change management processes
cloud applications aligned to your platform will ensure you are set up to operate smoothly
and derive continued value from service updates and enhancements.
Operate in the cloud

Evergreen cloud
Center of Excellence
Upgrade from on-premises You might have a partner and specialized development team assist
to the cloud with the initial deployment, but once in “business as usual” operations
mode, does your team have the skillset to manage the platform, main-
tain the custom components, and assess and implement appropriate
new features? Not having the correct level of expertise and guidance
onboard after you go live can inhibit solution advancements and
undermine your ability to take full advantage of the platform. Creating
a Center of Excellence (CoE) within the organization and investing in
developing the skills to operate on the platform is extremely valuable.
Organizations that embrace organic application development by end
users also need a well-established CoE. This center can administer,
nurture, govern, and guide users to adopt the citizen development
approach to business applications, empowering everyone to create
innovative apps with high standards by using a consistent application
lifecycle management (ALM) process.

Microsoft provides a CoE starter toolkit that includes prebuilt processes:

▪ Administer  Gain insights into your Power Platform adoption
▪ Govern  Establish audit, compliance, and archive processes
▪ Nurture  Accelerate your adoption with a thriving community
▪ Drive innovation  Pick the most impactful and valuable scenari-
os for development

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 51

 ▪ Consistent ALM  Apply source control strategies using GitHub
▪ Standards and theming  Create, manage, and share
enterprise-branded themes as well as maintain and follow organi-
AppSource offers a variety of ISV solutions
to suit your needs.
zational best practices and standards

Usage and monitoring

Running your business on the cloud means you want the cloud to be
You can read more about CoE at
Establishing a Microsoft Power
always available, responsive, and scalable. You also want to be alerted
Platform Center of Excellence. to unexpected issues in the applications and infrastructure so you can
diagnose and fix them. Additionally, you want to comply with SLAs you
have in place from other businesses. Having a monitoring capability
can prepare your business to meet these objectives.

What to monitor?
Different personas in your company will want to monitor for different
aspects of the system. End users may be more concerned with respon-
siveness, admins may be looking at changes, and the business may be
looking at KPIs like the time taken to place an order.

When running a SaaS application, you want to focus on the business,

while a service vendor assumes the role of maintaining the solution.
As a customer, you still have a level of control over the solution. This is
especially important when you have customizations in place to fine-
tune the business processes for your organization. If you introduce a
new feature in production and it causes unexpected issues, you want to
detect it proactively and resolve it before it becomes a problem. Using
features and flags to control new features can serve you well in these
situations because you can quickly turn off a faulty feature.

A proactive and reactive monitoring strategy that empowers your staff

to detect and resolve issues should be part of every customer’s arsenal.

Best practices for monitoring

Keep in mind the following best practices:
▪ Decide what exactly to monitor, especially the KPIs, and whether
the service is meeting them.
▪ Monitor your consumption, subscription, and services fees. The

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 52

 more cloud services you use, the more they will cost.
▪ Use automation to automatically trigger actions and to alert
people when thresholds are reached.
For example, you can configure Dynamics
365 and Azure Application Insights to
▪ Track user activities for accountability (including the applications,
access telemetry and diagnostics data screens, and features they use), the responsiveness, and the
and set up alerts and dashboards to
proactively monitor your Dynamics 365 frequency of use.
environments.

Adopt and align to a product roadmap

Your cloud solution will continuously evolve and deliver on the prom-
ise of digital transformation by delivering new capabilities that help
you stay competitive in your industry. Cloud vendors do their own
research and listen to customer feedback to improve the product by
adding new features. Being actively engaged in this process by giving
feedback, staying abreast of the latest roadmap, and planning for
upcoming changes is an effective way to operate in cloud. Also, the
key stakeholders from the business need to remain engaged and stay
plugged in to new features development, investments, and product
direction. We recommend developing your own roadmap for business
applications and aligning this to the product direction to get the most
out of your investments.

Dynamics 365 has two release waves per year in which several incre-
mental enhancements and new capabilities are made available to
customers. Adopting these mandatory changes and new features,
many of which are included in existing license plans, are a fundamental
aspect of operating in cloud.

Stay engaged
Implementing the system is one thing, but getting continuous return
on investment (ROI) from your cloud solution requires engagement.
Companies who stay passionate and curious about the solution and
You can choose from application-specific adopt the latest features are the ones who enjoy most success. We
forums like the Microsoft Dynamics
Community Help Wiki, Yammer groups,
urge all customers to maintain engagement through conferences
blogs, events, and how-to videos where and events throughout the year, and stay connected by using com-
you can discuss ideas, learn features, learn
roadmaps, and ask questions. munity groups. Several official and unofficial communities have been
formed by customers, partners, and Microsoft engineering. Take every

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 53

 opportunity to influence the product through active engagement with
product leaders, advisory boards, table talk with product managers,
preview programs, and collaborative development opportunities.

Adopt a cloud mindset

Evergreen cloud
Cloud implementation
One of the benefits of a modern SaaS solution is that every custom-
Customize and extend er runs on the latest version of the service. It’s the only model that
cloud applications scales. With this approach, every security patch, bug fix, performance
enhancement, and functional improvement accrues to all implemen-
Operate in the cloud
tations across the globe. Microsoft assumes responsibility for keeping
the platform current. This means that you no longer have to pull large
Evergreen cloud
teams of people together every few years to perform a traditional
Upgrade from on-premises upgrade over many weeks or months with limited added value for the
to the cloud business. The evergreen approach and the model of continuous up-
dates gives your business access to the latest capabilities to stay head
of the competition and meet changing customer expectations.

Dynamics 365 has embraced this approach of operating on one

version in a way that delivers benefits without disruption. The improve-
ments and enhancements embrace the CI/CD model, with service
updates delivered on a weekly basis, but these changes aren’t targeted
at the end user experience. You can take advantage of release waves to
preview upcoming changes and test and prepare your users before the
changes are deployed.

Achieving the right balance of delivering improvements and business

capabilities in a safe manner without disrupting the user experience is
paramount to our evergreen cloud approach.

Chapter 20, “Service the solution,” delves further into the updates
wave approach.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 54

 Adopt a cloud mindset
Upgrade from
Cloud implementation
on-premises to the cloud
Customize and extend
cloud applications Organizations over the years might have invested in on-premises

Operate in the cloud

deployment of business applications. These applications could serve
the users and business well in their current state, but keeping them
Evergreen cloud on-premises over time could limit your opportunity for expansion. In
these cases, you should explore upgrading or migrating an existing
Upgrade from on-premises business application to the cloud and bringing along exist-
on-premises to ing business processes and data to speed up your digital transformation.
the cloud
Dynamics 365 provides the tooling and support to help on-premises
deployment upgrades or migrations to the cloud (subject to compat-
ibility and security checks). This option to upgrade to the cloud also
helps organizations take advantage of their existing investments in
implementation, as long as the supported implementation techniques
are used and compatible with the cloud. Still, everything we discuss
in this chapter on the cloud mindset, considerations, extension, and
operations is fully applicable.

When upgrading or migrating to the cloud, consider the following:

▪ The most critical factor in deciding on the upgrade-to-cloud ap-
proach is to assess if the current application being upgraded meets
the requirement of the business and is well adopted by users.
▪ The data model, design, and data quality should ensure that the
application being upgraded becomes a stepping stone to acceler-
ate digital transformation. The upgraded application should serve
as a foundation for further growth and adoption of new capabili-
ties in the cloud. It shouldn’t stifle cloud adoption by carrying over
poor design and data.
▪ The cloud platform should be approved by security to store and
process data in accordance with the organization’s security policies
and regulations. The patterns used for authentication and integra-
tion also need to be revisited to make sure they’re cloud ready.
▪ Understanding the data category, flow, transformations, location,
and encryption during and post migration will help you navigate

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 55

 organizational security policies and plan your cutover.
▪ The upgrade to the cloud can be a multi-staged process that involves
stringent compatibility and validation checks, so it’s critical to under-
stand the process and review the prerequisites early in the process.
▪ Some components in the on-premises application may need to
be remediated or replaced to ensure cloud readiness. Examples
to look at include deprecation, unsupported customizations, and
non–cloud ready patterns that pose a security or performance risk.
▪ Existing integration and peripheral applications will require chang-
The Dynamics 365 Migration Program can
help you take the first step toward cloud es to ensure they work with the cloud applications.
success by migrating your on-premises
▪ Any impact on process, performance, and usability due to latency,
solution with expert guidance from
Microsoft. For more information, refer to service protection limits, or license capacity should be considered.
the overview of our program or perform a
standard migration assessment.
The ability to bring existing Dynamics 365 on-premises applications
You can also move from Dynamics AX to
Dynamics 365 in the cloud. into the cloud with proper due diligence can be a game-changer for
your digital transformation.

Conclusion
Embracing SaaS applications to run your business can significantly
accelerate your digital transformation, but it’s also important to rec-
ognize that organizational cloud maturity will play a significant role in
your strategy’s long-term success.

Organizations working on their first major implementation in the SaaS

cloud should expect some level of disruption to the existing ways of
deploying and managing applications. We hope this chapter serves as a
primer for Success by Design thinking by helping set the context for the
changes you can expect and driving you towards greater cloud maturity.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 56

 Design the solution to meet the nonfunctional re-

Checklist quirements, considering factors like network latency,

end-user environment, and devices.

Ensure that when upgrading existing on-premises solu-

Adopt a cloud mindset tions to cloud, the data model, design, and data quality
make certain that the application becomes a stepping
Look to transform business processes by using SaaS-
stone and doesn’t stifle cloud adoption by carrying over
based applications and capabilities that provide value to
poor design and data.
the business process immediately and provide a path for
continuous improvement.
Customize and extend
Avoid implementing cloud-based processes that are
designed based on the recreation of legacy on-premise Understand the expected value of each application
user experiences. extension and quantifiable impact on aspects of the
business process such as data driven decision making,
Have a shared understanding at every level, from the
efficiency, automation, and ease of adoption.
executive sponsor to the developers, of the business
impact being delivered. Create guidelines for when to customize and extend
out-of-the-box apps and only adopt documented
Ensure the organization has done its due diligence
extension techniques.
organizing the data estate and understands the impact
of the new solution on the data estate. Avoid deviations and repurposing the out-of-the-box ta-
bles and data models (Common Data Model) because this
Gain approval for the cloud platform for use with the
inhibits future adoption of new features and capabilities.
appropriate application data category from information
security and compliance.
Operation
Ensure the respective teams understand all administra-
tive, operational, support, and monitoring aspects of the Ensure that the necessary expertise is available via the
platform, and the organization policies, processes, and partner (system integrator) or internal teams post-de-
patterns conform to the SaaS cloud. ployment to support and evolve the solution.

Implement DevOps and CI/CD pipelines to support Engage with the community to keep up with the latest
automation for build, testing, and deployment. innovation and help influence the product direction
using channels, events, blogs, and other mediums.
Design the solution according to service boundaries and
available licensing capacity, and ensure the controls to
further expand and scale the solution are understood.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 57

 Case study
Rev up for speedier service
A family-owned business based in South Australia is one of the largest
retailers in the region, with ventures that span fuel, convenience, quick-
serve restaurants, and real estate.

One of its key businesses is a chain of retail stores that serves busy
guests across Australia.

Their focus on customer service has led the company to reimagine

the way its store systems are designed and connected. As their retail
chain has grown, they have faced common challenges: outdated
technologies, changes in customer behavior, and the pace of change
with existing suppliers. The company was ready for an overhaul and an
infrastructure solution that could serve its goals now and in the future.

Given the high volume of traffic that the retail chain processes and the
diversity of its retail store locations, the company was seeking a hybrid
solution with reliable, flexible connectivity.

Scalability and speed are critical for their retail business, and they
needed an infrastructure design that optimizes for both.

“We serve millions of guests each month,” says the company’s store
systems specialist. “To achieve speed of service, we know that we need

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 58

 to have something on-premises to retain our workloads in the store.
And, at the same time, we need to have that cloud connectivity to the
back office.”

A cloud engineer at the company adds, “We have stores in rural areas
where they don’t have high internet speed connections. We wanted
to have something that is both in-store and in the cloud that would
synchronize when it comes to failovers and redundancy, so that we can
have the best of both worlds.”

When it came to renewing store hardware and to keep up with the

pace of innovation, the company chose to go with a hybrid model that
integrates Azure, Azure Stack Edge, and Dynamics 365 Commerce.

The new design impacts the stores’ day-to-day connectivity. “What

we are planning to do is a direct connection to Dynamics 365, and the
failover needs to happen in Azure Stack Edge,” one engineer said. “We
expect to have a minimum of 99 percent of uptime.”

One of the team’s other top considerations was how easy the infra-
structure would be to manage, so they factored in what the company
was already using across its retail business.

“We have already heavily invested in Microsoft products. We use

Dynamics 365, Microsoft 365, Power BI, and multiple other products,”
explains one engineer. “We wanted to have a single pane of glass
where we could administer and monitor all of the systems.”

The hybrid deployment model using Azure Stack Edge and Dynamics
365 Modern Point of Sale provides store operations redundancy in case
of network outage.

The new infrastructure design provides the connectivity, consistency,

and efficiency that the company needs. Over time, the company ex-
pects that it will help them avoid store downtime, maintain real-time
stock levels, simplify its infrastructure management processes, and
reduce maintenance and compute costs.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 59

 The new infrastructure design will impact the customer experience at
the store level. “We offer a comprehensive omnichannel experience to
guests walking into our stores,” says the store systems specialist. “We
enable our guests to pay at the pump without having to come into the
store by using the [retail chain] app. We also enable our guests to click
and collect their food, coffee, or supermarket needs. To enable that, we
need real-time connectivity to our head office systems, that is key for
us. With Azure Stack Edge and an in-store database, we are hoping to
get that real-time connectivity.”

This orchestration ensures connectivity and consistency, as a cloud

engineer explains. “We deployed three virtual machines. One is used
to monitor the in-store cameras, and the other is used to store a
Dynamics 365 database backup. If the store loses its internet connec-
tion to Dynamics 365, it would be able to operate normally while that
virtual machine continues with the operations. Then, once the connec-
tion is restored, it will be able to synchronize back to Dynamics 365.”

As the company continues to roll out its new infrastructure, it’s piloting
a solution to recognize vehicles’ number plates based on real-time
CCTV feeds. One benefit of this is loss prevention, in situations where a
number plate is tied to a known offender in a published database.

They also plan to use the in-store cameras to do machine learning and
AI on Azure Stack Edge virtual machines to predict stock levels and
alert the staff, all locally, without the cloud.

Implementation Guide: Success by Design: Guide 3, Implement cloud solutions 60

4 Guide
Drive app
value
Implementation Guide: Success by Design: Guide 4, Drive app value 61
 Introduction
In the constantly evolving world of Dynamics 365,
software as a service (SaaS) cloud applications, and
changing business needs, it’s important that your
business applications evolve and are consistently
delivering value.
This chapter outlines the approach to digital transformation and dis-
cusses strategies on how to shorten the time to value, cut through the
noise, and focus on business outcomes.

The digital transformation journey of an organization usually involves

various phases and levels of maturity. A green field implementation could
start with a pilot or even a minimum viable product (MVP) targeted at
Key topics in this chapter: a specific business process, then incrementally expand the solution with
more capabilities for the users and onboard new business processes.
Think like a startup: start lean, build a
successful MVP, evolve quickly

Alternatively, replacing an existing system could mean that a significant

Phases and incremental strategy
amount of functionality needs to be made available at the start in order
Manage the change stream: business,
user, and product
to have parity. No matter where you start, realizing the long-term goals
of a SaaS-based digital transformation requires effective planning.
Achieve hyper-scale with expansion
approaches and enable organic growth
In this chapter, we explore how to create a digital transformation road-
Better together: satellite applications
and aggregations map for your organization that can help accelerate the time to value
while maintaining relevance in a constantly changing technology and

Implementation Guide: Success by Design: Guide 4, Drive app value 62

 business landscape. This guide can help you navigate the challenges
of expanding your solution and drive continuous, long-term business
value through digital transformation.

Approach to digital
transformation
Throughout this book, we discuss several concepts related to Success
by Design, including how to initiate, implement, and deploy a project,
but the scope of Success by Design isn’t limited to a successful go live
or operational excellence. The long-term goal of Success by Design is to
create the right foundation for an organization to evolve their business
application landscape and expand their digital transformation footprint.

Success by Design Every implementation related to business applications doesn’t nec-

essarily qualify as a digital transformation program aligned to the
creates the business; several systems are migrated to the cloud in their current
foundation for form as a part of infrastructure modernization or to ensure sup-
an organization portability. In some cases, this is just the first step towards enabling

to evolve their digital transformation; opportunities could arise to take it further and
deliver transformational value to the business rather than looking at it
business application as just a tactical upgrade to projects run by IT with little involvement
landscape. from the business.

The discussion in this chapter targets business-focused digital trans-

formation programs and their expansion. Before we go into the details
of expanding a solution to further drive value, let’s discuss the funda-
mental approach to transformation: depicting your current business
model, the factors that affect your business model, how they can
trigger transformation, and defining your goals and scope of business
processes for a transformation roadmap.

Represent the business model

Before you define a digital transformation roadmap, a good strategic
exercise is to holistically develop an understanding of your business

Implementation Guide: Success by Design: Guide 4, Drive app value 63

 model. You can do this at an organization, department, business unit,
or team level to understand the value that you currently deliver, who
your customer is, and how to deliver that value and the key activities
and resources involved.

You could use several different techniques or frameworks to describe

an organization’s business model and how they create value. One
framework that we discuss in more detail is the Business Model Canvas,
developed by Alexander Osterwalder. It offers a simple, holistic, and
easy-to-understand representation of a business model. This model
isn’t officially endorsed by Microsoft or part of Success by Design, but
you can use the model to not only map your existing business model
but also highlight what aspects of your business model are changing,
which your digital transformation strategy will help you adapt to.

The business model represents the “why” and what it takes to deliver
the value to customer; how we go about doing it is the business
Disruptions in the process definition.
business model
create opportunities Triggers
for transformation. Disruptions and inflections in the business model create big opportu-
nities for transformation. In a world that’s changing at a pace greater
than ever, businesses have to reinvent themselves more often, improve
customer experiences, and attract and retain talent. The opportuni-
ties for impact are plentiful (for a deeper discussion, see Chapter 1,
“Introduction to Implementation Guides”).

Let’s explore some examples of how changes in the business model,

disruption in an industry, new regulations, or even external factors like
global events can be a catalyst that triggers changes to business pro-
cesses and applications.
▪ Changes in customer segment  As businesses evolve their prod-
ucts and expand their target customer base and demographics, they
often see that their approach to nurturing their potential customer
base also needs to evolve. Targeted messaging that resonates and is
delivered through relevant channels becomes extremely critical. Does
your marketing automation and customer data platform enable this?

Implementation Guide: Success by Design: Guide 4, Drive app value 64

 ▪ New channels  Customer channels that the business uses to
deliver their product and services are evolving. For example,
customers expect remote delivery of services via digital channels
instead of an in-person visit to a store, which is further necessi-
tated by evolving circumstances such as the COVID-19 pandemic.
What does this change in channels mean for your order manage-
ment systems or other related business processes?
▪ Changes in value proposition  Businesses might differentiate
themselves by coming up with unique customer value proposi-
tions in highly competitive sectors, such as offering a lucrative
subscription model, to drive recurring revenue and shareholder
value. For example, customers aged 25–40 are more open to
subscription-based purchases versus outright ownership. The
automobile industry now offers cars on a subscription basis and
charges based on miles driven, as opposed to customers paying
full purchase price. How are your processes and business applica-
tions affected by this change in business model?
▪ Changes in customer relationships  An organization’s ap-
proach to engaging with customers evolves with the business.
For example, an organization could change from an aggressive
customer acquisition phase to retaining and upselling within their
existing customer base after acquiring a fair market share. How is
your customer relationship management (CRM) strategy chang-
ing, what does this mean for your CRM system?
▪ Changes in revenue streams  As businesses evolve, they create
new revenue streams. For example, an energy provider in a highly
competitive market might provide home services to expand their
revenue stream, which triggers changes to customer processes
and related business applications. Do you know how your revenue
stream changes impact your business applications?.
▪ Resources and sustainability  More organizations are commit-
ting to environmental sustainability. How do your manufacturing
and operations track your carbon footprint, impact on emissions,
and procurement from sustainable sources?
▪ Custom manufacturing  With the advances in 3D printing tech-
nology, a manufacturing company can offer custom, made-to-order
options to their customers. How does this impact your operations
processes and enterprise resource planning (ERP) application?

Implementation Guide: Success by Design: Guide 4, Drive app value 65

 ▪ Partnerships and acquisitions  Inorganic, acquisition-based
growth is common in several industries, and so is the strategic
partnership between organizations to grow market share where
they have complementary services. How do these changes affect
process alignment and your supporting business applications?
▪ Cost and efficiency  New technologies offer efficiency advan-
tages that can reduce operational costs. For example, delivery
companies use electronic signatures instead of paper-based
solutions to get real-time notifications and provide new value
to customers. Are there opportunities to drive process efficiency
using the latest technology?
▪ External factors  Changes in the socioeconomic situation, po-
litical climate, and regulations can affect the way organizations
conduct business. For example, the EU’s decision to bring in the
General Data Protection Regulation (GDPR) affected every busi-
ness with employees and customers in EU, which required changes
to their business applications to ensure compliance. How are new
and changing regulations impacting your business?
Changes to the
These triggers can all disrupt your existing business model. You can use
business model can the Business Model Canvas to highlight what aspects of your business
kickstart digital model are changing and align your digital transformation strategy to
transformation. these changes, which will eventually help your business adapt and grow.

Digital transformation goals

Changes to the business model (caused by a variety of triggers) can
kickstart digital transformation. These transformation goals must be
defined in business terms, with measurable metrics that could poten-
tially involve several milestones. Let’s look at an example in which an
organization’s goal is to create a 360-degree view of their customer.
They want a holistic view of their customer, their activity history, prod-
uct purchases, subscriptions, and more. But this goal doesn’t relate to a
business outcome or a measurable key performance indicator (KPI). A
measurable goal for your digital transformation could be the business’s
ability to increase upsell to existing customers by X percent or reduce
the time to address customer queries by Y percent. This requires a
360-degree view of the customer.

Implementation Guide: Success by Design: Guide 4, Drive app value 66

 Although most programs begin with a set of business goals, it’s equally
important to communicate and reiterate these goals to the IT and
technical teams implementing solutions. In long-running transfor-
mation programs involving several technologies, teams and partners
might lose sight of the actual business goals by letting technology
become a distraction.

Business process
The next stage of the process is discovery, which involves key business
stakeholders. With a holistic view of the evolving business model and
transformation goals in place, you can identify the related business
processes and applications that need to change. The discovery exer-
cise should be focused on creating clarity around how the business
processes and applications identified need to evolve to meet the
corresponding transformation goals.

The scope of change could be automating a manual activity to

improve productivity, capture data accurately to improve the effec-
tiveness of strategy, drive user efficiency, or enable the business to
generate insights about your customers by eliminating data siloes.

Prioritizing the changes to maximize and deliver measurable business

impacts continuously (without having to wait for a multi-year trans-
formation program to complete) keeps digital transformation in the
foreground, with engagement from the business, end users, and exec-
utive sponsors maintaining its relevance. Business agility should be the
key focus for long-running, comprehensive transformation, or you risk
You can use objectives and key results missed business opportunities and loss of market share.
(OKR) as a goal-setting framework for
digital transformation.
For more information, refer to Chapter 7, “Process-focused solution.”

Change streams
So far, this chapter has discussed the approach to digital transforma-
tion and how to develop a transformation plan for your application
and the process involved. This has traditionally been a top-down

Implementation Guide: Success by Design: Guide 4, Drive app value 67

 approach, starting at the business model and then narrowing down
to business processes and applications that deliver the most impact.
However, you should always keep in line with the broader theme of
constant change and embrace the mindset of getting comfortable with
frequent and inevitable change. In this section, we explore the change
streams that impact the roadmap and scope of planned activities.
Planning for how to deal with these changes throughout the program
is a key determiner of success.

Business
As we embark on the journey of transformation and start building the
business application based on the goals set by the business, it’s also
important to appreciate that during design and implementation, you
may need to accommodate for further adjustments and refinements.
This support for agility and change needs to be fundamentally built
into the program. This is where some of the iterative methodologies
could be beneficial, enabling us to respond to the change without it
turning into a disruption.

Embrace the Those leading transformational change often find that all aspects of
change are not well defined in the early stages , so iterating to adopt
mindset of getting changes quickly and early is key, but so is flexibility as clarity reveals the
comfortable with resulting business model. When transformation occurs in an industry,
frequent and what the industry will look like following the transformation is often

inevitable change. not known. Companies must be able to adapt quickly in the middle of
the project to incorporate these inevitable changes.

User
A key stakeholder in your transformation plan is the business user, who
interacts with the application often. If the process being implemented
in the application doesn’t meet the needs of the user, or the applica-
tion doesn’t consider working patterns and end user usage, it could
lead to poor adoption and lack of engagement. Incorporating user
feedback continuously throughout the process using well-defined and
frequent touchpoints is key to achieving your transformation goals.

Implementation Guide: Success by Design: Guide 4, Drive app value 68

Your approach to designing and developing the application needs to
be user-centric; it must clearly articulate how the change being imple-
mented results in value for the user.

You should expect that user experience expectations constantly evolve

as users are exposed to a variety of software products both in the enter-
prise and consumer space. For example, the predictive text fill feature
popularized by web search engines is now the expected search expe-
rience in business applications. This makes it even more important to
adopt newer features and improvements delivered by SaaS applications,
which takes us to the next change stream: product and technology.

Product and technology

In the world of cloud-based SaaS applications, you can expect a con-
stant flow of new capabilities, improvements to existing features, and
transitions to the latest technology paradigms, which can impact your
current application. Traditionally, business applications were built by IT,
with features requested by the business and end users. With the SaaS
cloud, the service providers themselves are invested in understanding
the latest trends in business and delivering innovative capabilities to
help their customers stay ahead of the competition. For example, a
business trying to transform their sales application can assemble an
application using the best-in-class sales capabilities offered by the
SaaS provider instead of spending years in research, design, and de-
velopment to build from scratch, which would make the application
obsolete by the time it launches.

Similarly, continued investment in microservices by Dynamics 365,

such as planning as a service, can transform a core business model
for companies allowing them to run material requirements planning
(MRP) as needed to drive a significant and transformative way of
managing demand and the supply chain.

The enhancement delivered via continuous updates has a much

shorter adoption timeframe when compared to traditional major
version upgrades.

69
 SaaS application providers that are competing to build and deliver
business capabilities for customers are helping advance business appli-
cations in way that was unimaginable just a few years ago. Applications
that were just forms over data (mostly passive data capture systems used
to track, view, and report on known data) have evolved into applications
that can automatically capture data, learn from the data, deliver insights,
and guide users to the next best action. This makes it extremely import-
ant for business to watch for new capabilities being made available and
adopt them to accelerate their transformation to be a differentiator in
the industry. The key value proposition of SaaS is also the ability to tap
into the enhancements and features that are based on broad market
research. Activating these features can accelerate your transformation
with minimal effort and without any development cost.

Your SaaS application provider is no longer just a software vendor

providing an off-the-shelf product, they’re a strategic partner to your
business with significant impact on your ability to implement your
strategy. Developing a strong relationship with the SaaS providers
to not just learn about the roadmap but also influence the roadmap
through strategic feedback ensures that you’re getting the most value
out of the product.

External
External economic, social, and political drivers can disrupt your
transformation plan. The COVID-19 pandemic is an example of how
supporting remote work and creating online collaboration channels
became a top priority for most customers in 2020. This required co-
ordinated changes from infrastructure, to network, to the device and
application layer across the IT landscape of organizations. Although it’s
difficult to foresee and plan for external events, the iterative approach
to delivering continuous value in smaller batches allows you to pause
and pivot as needed. For example, consider an organization on a long-
term transformation program that releases a major capability to users
once a year versus an organization that has adopted the DevOps cul-
ture of delivering value with regular bi-monthly or monthly updates.
The latter company can realize value from investments much sooner
and is better positioned to pivot when change demands it.

Implementation Guide: Success by Design: Guide 4, Drive app value 70

 Transformation map
So far, we have discussed a high-level approach to digital transforma-
tion and the various change streams that you need to consider. The
ever-growing product backlog of feature asks, change requests, and
product capabilities can overwhelm and potentially distract from the
drivers for transformation. To determine time to value, you need a
shared understanding between the business, IT, and users on how a
product feature or change aligns with the goals of transformation and
impacts the maturity of process implementation.

One way to visualize and communicate your digital transformation

goals is to plot the application’s capabilities in a transformation map
(Figure 4-1). You can adjust the definition of each quadrant to your
own goals and ambitions for the process, but you can easily get start-
ed with the template we offer here. If all the feature asks, changes,
Fig.
4-1 and SaaS product capabilities are plotted on this quadrant, you can
determine the current state of
your business process implemen-
Enhanced Differentiator tation and agree on the future
state aligned to your transforma-
Predictive
forecasting tion goals.
Forecasting
Sequences

Premium autocapture For example, a sales system

that only has the basic features
Assistance
cards and studio of contact management and
opportunity management may
Essential Innovative
be considered essential, whereas
Conversion
Contact intelligence a system that can use AI-driven
management
insights with advanced forecast-
Standard cards ing might be a differentiator that
Opportunity
Advanced

pipeline gives you competitive advantage

management
Autocapture
Quotes
in your industry.

Creating a dashboard for

Innovative
stakeholders that organizes the
requirements and features in this

Implementation Guide: Success by Design: Guide 4, Drive app value 71

 fashion can bring great clarity, not just during initial implementation,
but for future increments. It’s also important to prepare for time-based
decay, in which a feature that is considered a differentiator today
might be considered essential in a few years. Similarly, something that
is innovative or even considered ahead of its time could become a dif-
ferentiator in the near future. You should plan to revisit and refresh this
transformation map for each phase, product release, pivot, or strategic
inflection in your business.

Phases and increments

Planning and scoping distinct phases of your digital transforma-
tion program directly impacts the time to value realization. What is
delivered and when it’s delivered also has an impact on the level of
engagement from the stakeholder and the end user’s adoption. These
phases and increments are based on the goals agreed upon with the
Fig.
4-2 business and allow you to select the appropriate capabilities in the
transformation map.

Enhanced Differentiator It seems reasonable to start with

Predictive
the essentials and basics in Phase
Phase 2 Phase 4 forecasting
1 of a solution and then deliver
Forecasting Sequences enhancements with most the in-
Premium autocapture novative, differentiating features
Assistance cards in later phases of the program.
and studio
But if the most valuable parts of
the transformation are planned
Essential Innovative
for much later in the lifecycle
Conversion
Phase 1 Phase 3 intelligence (Figure 4-2), you risk stakehold-
Contact
management ers losing interest, or poor user
Standard cards perception of the system (which
Opportunity
Advanced

pipeline is hard to correct). Additionally,

management
Autocapture
Quotes the features that were considered
differentiators could now be
essential due to time decay. A
Innovative
transformation program that fails
to deliver beyond the essentials or

Implementation Guide: Success by Design: Guide 4, Drive app value 72

 the initial MVP drives little transformation.

You should plan the phases so you can pick up elements from various
quadrants and deliver a holistic solution that has the key process
elements, integrations, and reporting that delivers value (Figure 4-3).
Look for quick wins: the latest SaaS features that you can easily adopt
with little effort to bring innovative and differentiator elements into
the earlier phases.

Digital transformation is as much about people as the process and

technology. The phases and incremental design should also help
product owners maintain the right level of engagement from users and
stakeholders. This drives excitement and curiosity, but most important-
ly feedback. You could draw an analogy from a TV series that shows
teasers at the end of each episode or the end of each series to attract
viewers back to the program. Do the same for your digital transforma-
tion story by delivering value beyond the essentials in each phase so
customers come back for more.

Fig.
4-3

Enhanced Differentiator

Predictive
Forecasting Phase 2 Phase 4 forecasting

Premium autocapture
Assistance cards
and studio

Sequences
Phase 1

Essential Innovative Phase 3

Conversion
intelligence
Contact
management
Standard cards
Opportunity
Advanced

pipeline
management
Autocapture

Quotes

Innovative

Implementation Guide: Success by Design: Guide 4, Drive app value 73

 In the next section, we look at how the right approach to MVP strategy
when getting started with your digital transformation journey can help
get early feedback and drive value sooner rather than later.

Minimal viable
product strategy
The concept of MVP, which enables you to quickly build, measure,
learn from feedback, and pivot if necessary, has been well established
in the startup world.

Bringing the invaluable lean startup mindset to digital transformation

can help organizations deal with uncertainty from different change
streams and find the value that drives maximum business impact in
the least amount of time. Even in the context of digital transformation,
MVP is often used to test the value of a solution to a business and
gather feedback.

When translating the lean startup methodology to a business appli-

cation, MVP should not be a solution with just the basic and essential
features; you should create a solution that will help transform your
process and deliver on the most important transformation goals with
the least amount of effort. Most importantly, an MVP is not the end
state; it should drive quick feedback, iterations, and pivot to program
to avoid being in an MPV limbo for years.

Going back to the example of a sales application, let’s consider if

users are currently using a homegrown sales system with a form to
enter and view opportunity data. The system has a backend to store
this information, which is also used for management reporting. The
business has invested in a Dynamics 365 SaaS-based sales applica-
tion after assessing its rich features, out-of-the-box AI insights, and
integration capabilities with Power BI for reporting and analytics,
which allows them to improve their opportunity closure rate and get
accurate growth forecasts. A common approach is to first implement
the out-of-the-box opportunity management process and reports,

Implementation Guide: Success by Design: Guide 4, Drive app value 74

 and plan future phases to implement AI-based insights with Power BI.
Although you can easily follow this route, it doesn’t achieve the actual
goal of sales implementation. The system delivers an improved user
experience but fundamentally does the same thing as the home-grown
sales system, with little direct value to the business. Alternatively, using
embedded intelligence with opportunity scoring and relationship
health can help users target the right opportunities and improve their
close rate, which directly impacts business from day one.

“MVP is not the A good MVP has the following characteristics (Figure 4-4):
▪ Delivers on the most important of the digital transformation goals.
smallest product ▪ Is production-ready and uses supported customization techniques
imaginable, though (this is different from a proof of concept or prototype designed to
it is the fastest way prove feasibility or get specific usability feedback).

to get through the ▪ Goes beyond the essentials and delivers on top of the existing
capability of the incumbent system.
build measure learn ▪ Focuses effort on delivering maximum business value in the short-
feedback loop with est amount of time, (you could compromise on some experiences
minimum amount that aren’t directly aligned to business outcomes).

of effort.” ▪ Is built in a way that allows for quick changes based on feedback,
making it possible to pivot. This translates to a greater reliance on
– Eric Ries, author of The Lean Startup
low-code configuration versus highly customized, professionally
developed, code-heavy experiences.
▪ Allows you to test and learn latest SaaS product features that could
be activated (this requires some level of change management but
Fig. almost no development effort).
4-4

Delivers on Production ready with Beyond the essentials Maximum value to business
goals for supported customization and delivers on in shortest time, minimal
transformation techniques (not proof of existing system compromises
concept)

Characteristics of a good MVP

Creates the foundation, Test hypothesis Test, learn, Built to pivot and
supports building blocks and deliver on activate change after feedback
for success expectations

Implementation Guide: Success by Design: Guide 4, Drive app value 75

 ▪ Allows you to test the hypothesis of a specific feature or solution
delivering the expected business outcome and adjust based on
what you learn.
▪ Creates the foundation for long-term success—it should support a
building block mindset in which you keep adding without disrupt-
ing previous steps. (For more information about the architectural
skills to realize this approach, see Chapter 6, “Solution architecture
Fig.
4-5 design pillars.”)

Build
Although the MVP approach naturally works with new greenfield im-
Ideas Emotional
Code
product plementations, you may have existing apps that are being replaced or
Reliable migrated to another platform. Your scope should consider the existing
functionality to ensure parity, but shouldn’t try to mimic or replicate
Accessible

Functional

Learn Measure the experiences of the outgoing application. Focus your MVP strategy
on delivering the most value to the business sooner without limiting
Data
the scope to essentials (Figure 4-5). An MVP with a very broad scope
that takes years to deliver may cease to be an MVP.

Drive expansion
All the approaches and techniques we have discussed so far in this
Fig.
chapter can help you create an effective digital transformation roadmap
4-6
for your business application. They’re relevant to
the Initiate phase, but you can apply these methods
Users
Feature New iteratively or revisit them for further expansion. This
adoption workloads
section focuses on expanding the usage and scope of
business applications in different areas (Figure 4-6)
Incremental Satellite
changes applications to drive even greater business impact, while consider-
ing additional impact on the following:
Expansion ▪ Security and compliance
Application Integrations ▪ Application lifecycle management (ALM)
standardization
▪ Administration and governance
▪ User readiness and change management
Mobile Aggregation
▪ Environment strategy
Pivot
▪ Limits and capacity
▪ Data and integrations

Implementation Guide: Success by Design: Guide 4, Drive app value 76

 Users
The most common expansion scenario for a business solution is add-
ing more users. This could be the result of onboarding new regions
or countries onto the same application. User-only expansion sounds
straightforward, but you could have data migrations on a live system,
security and compliance challenges based on the region, or other
impacts (Figure 4-7).
Fig.
4-7 Users

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

High Low Low Medium-high Low Low Medium

Potentially high If the solution Based on the Full-fledged Creation of a Although each Usually, data
impact based remains the environment user change new production user usually migration can
on regional same, impact on strategy and management environment comes with be complex on
regulations ALM is minimal the creation and readiness impacts the their own a live system
of additional effort is required environment capacity and in case
production strategy integration API, of additional
environments, capacity, and environments,
you could see storage could integrations
medium impact be impacted might have to be
replicated across
environments

Mobile
Ensuring support on multiple devices, at times with offline capabilities,
could be critical to adoption. In some cases, this needs to be a part
of the initial release (such as in warehousing systems in which goods
must be scanned when picked up), whereas in others it could be a
quick follow-up. Most out-of-the-box capabilities work as is on mobile
apps, with additional optimizations to support the move experience.
Our example sales mobile application offers the core sales application
features with additional targeted mobile experiences for field sellers
(Figure 4-8).

Implementation Guide: Success by Design: Guide 4, Drive app value 77

 Mobile
Fig.
4-8

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

High Low Low Medium-high Low Low Medium

Potentially high If the solution Additional User readiness Should not have It should not Custom
impact based on remains the device policies effort is any significant have any embedded
the regulations same, impact on could have an required to impact significant integrations
and need for ALM is minimal impact ensure adoption impact on user might require
mobile device on mobile and capacity work to make
management account for any them responsive
or mobile variations for mobile
application consumption
management
features

Incremental changes
Incremental changes are primarily driven through the feedback and
change requests from users and the business. These changes help ensure
that the application meets the expectation of ongoing enhancement
and continues to build on top of the initial MVP to maintain relevance.
It’s still important to look at these improvement requests through the
business value and transformation lens (Figure 4-9).
Fig.
4-9 Incremental changes

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Low Low Low Medium Low Low Low

Changes within No major No expected User readiness Usually remains Additional Assuming no
the approved impacts if the impact as long effort is unchanged integration and integration
SaaS service solution doesn’t as the data required app workloads changes, impact
boundaries have new PaaS boundaries could have is minimal
and data loss components don’t change some impact
prevention
policies should
have limited
security and
compliance
implications

Implementation Guide: Success by Design: Guide 4, Drive app value 78

 Feature adoption
As we have seen when discussing change streams, SaaS products make
significant investments in researching and implementing business
capabilities that their customers can easily enable. This is a key value
proposition of the SaaS cloud, and can play a key role in deriving value
from your investments. With continuous updates in Dynamics 365,
new features are made available to customers. A business analysis of
the new features with technical impact analysis on the existing solution
and customizations (Figure 4-10) is a key business as usual (BAU)
activity that must be conducted in collaboration with key stakeholders.
Most of these feature improvements are made available to customers
with no additional cost and can help advance your application to the
innovative or differentiator quadrant.
Fig.
4-10 Feature adoption

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

High Low Low Medium Low Low Medium

Usually No major No expected User readiness No impact Usually no Features might

no impact impact expected impact as long effort is required impact, generate
except when as the data to use the new sometimes additional data
connecting to boundaries features features might but shouldn’t
external data don’t change require additional impact
sources storage and integrations
capacity

New workloads
Dynamics 365 provides several business applications that are built on
the same scalable power platform. As end users adopt your application,
you will find additional opportunities to onboard related workloads
that could help eliminate data siloes by enabling stronger collabora-
tion and data sharing. For example, to achieve seamless collaboration
when you’re onboarding a marketing workload in addition to sales, the
sales team can get visibility into the nurture programs their customers
are part of, extend invite to marketing events, and so on. Similarly,
the accounts payable team could expand and move into a centralized

Implementation Guide: Success by Design: Guide 4, Drive app value 79

 accounts payable processing system. However, new workloads can
have significant impact (Figure 4-11).
New workloads
Fig.
4-11

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Low-medium Low-medium Low-medium Medium-high Low-medium Low-medium Medium-high

This could be App- Some apps User readiness New apps Additional apps Data and
medium impact specific ALM could require effort is on existing can impact integration
based on the app; requirements additional required, but environments storage and needs for the
core Dynamics can impact your admin tasks existing users won’t have a tenant API new app can
365 apps might release and will benefit from major impact capacity have an impact
have been build pipelines a familiar UI
already approved

Satellite applications
When deployed, your business application covers core use cases and
scenarios, some of which may only be relevant to a specific team,
region, or role. In such scenarios, you can deliver such targeted capa-
bilities via a satellite application. Satellite apps are a good opportunity
for user-developed apps and automations while using the data model
of the core application. You could also use these applications for in-
cubation before incorporating it in the core app. Regarding areas of
impact (Figure 4-12), it’s important to have strong governance around
the data model and creation of additional data stores, which can lead
to data fragmentation.

Integrations
Integrations can play a key role in eliminating data duplication and im-
proving data quality. Most importantly, they reduce the need for users
to navigate multiple applications, which prevents context switching.
The core integrations should be part of the initial phases and not be
left to future expansion phases. Even the out-of-the-box integrations
with existing productivity applications like Microsoft 365 can positively
impact adoption and make the information seamlessly available across
apps while enabling stronger collaboration. However, an abundance of

Implementation Guide: Success by Design: Guide 4, Drive app value 80

 caution is needed; integration can also lead to performance issues and
increase the overall complexity and cost of maintenance (Figure 4-13),
so using the right patterns and practices is critical.

Fig.
4-12 Satellite applications

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Low-medium Medium Medium-high Medium Medium-high Low-medium Medium

If the data flows Managing As the satellite User readiness Depending Potential impact You should
in accordance the lifecycle apps grow effort is low on the ALM to capacity; avoid
with data loss of satellite organically, the for citizen- and citizen you may also fragmenting
prevention apps requires appropriate developed development need licenses the data with
policies, impact changes to governance community strategy, the to support the additional apps
should be low existing ALM policies need to apps, but might environment makers and using their own
processes be in place to require readiness strategy is consumers if data stores,
manage them effort if they’re impacted they’re not but it may be
adopted more already covered required in
broadly by existing some cases
licenses

Fig.
4-13 Integrations

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Medium-high Medium-high Medium-high Low-medium Low-medium Medium-high High

Depending on ALM process Additional Depending You may need Potential Expect impact
the data flows, could be monitoring on frontend to have stubs impact to API on data
additional impacted and integration versus backend or downstream consumption movement and
security depending on telemetry integration, integration make efforts
approvals may complexity is needed the impact and environments to ensure
be needed for to support readiness effort integrations
integration reconciliation of could vary are built to
pattern failures standard

Implementation Guide: Success by Design: Guide 4, Drive app value 81

 Aggregation
Aggregation can help consolidate multiple business applications with
significant overlap of data and processes into a single application.
Aggregation is different from integration, it completely replaces the
application and brings the core feature set into an existing application
instead of exchanging or embedding information. For example, a
commercial banking sales system could become an aggregator for all
non-personal banking sales systems.

The cutover approach for old applications is critical; make sure to avoid
parallel usage leading to fragmentation of data and other aftereffects
(Figure 4-14).
Fig.
4-14 Aggregation

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Medium-high Medium-high Medium-high Medium-high Low Medium-high Medium-high

Level of impact Existing ALM Admin User readiness is This will have Additional users High impact
depends processes are processes are required impact on and increased with respect to
on data changed consolidated environment data volumes data alignment
classification strategy, impact capacity and integration
requiring a consolidation
potential merge

Application standardization
This model of growth is driven by creating a generic application that
isn’t targeted at a specific process but a host of similar processes. The
application could even provide a framework that enables the business
to onboard themselves. This can achieve hyper-scale, in which you can
broadly t-shirt size hundreds of processes or applications that serve
more than one business process.

A general-purpose application allows scale, but can require addition-

al training and user readiness, so it must be thoughtfully done to
be successful.

Implementation Guide: Success by Design: Guide 4, Drive app value 82

 For example, consider a generic case management system that can
accommodate hundreds of different case processes from different
internal business functions that broadly involve similar steps. The busi-
ness can create a team of resolvers, configure service-level agreements
(SLAs), provide email notification templates, and create Microsoft
Word or Excel templates to capture non-standard data, but they can
still use the generic application for core case management capabilities.
All these actions have important considerations (Figure 4-15).

Application standardization
Fig.
4-15

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Medium-high Medium-high Medium-high Medium-high Low Medium-high Medium-high

Once approved Existing ALM There could be User readiness is Should have Additional users There could
for appropriate processes process specific required for each minimal impact and increased be an increase
data shouldn’t configurations new process data volumes in the data
classification, change for a required for impact capacity generated by
more processes specific template each new each process
shouldn’t have workload on the
an impact template

Pivot
Pivots aren’t necessarily expansion, but can trigger major change to an
application, which could fundamentally change and further broaden
its scope (Figure 4-16). For example, a targeted Power App used by a
small team could go viral within an organization, and you must pivot
to make it an organization-wide application.
Pivot
Fig.
4-16

User readiness
Security and Admin and Environment Limits and Data and
ALM and change
compliance governance strategy capacity integrations
management

Medium-high Medium-high Medium-high Medium-high Low Medium-high Medium-high

Application Existing ALM Admin and User readiness Potential Additional users Expect high
will need to processes will governance might be changes to and increased impact to data
undergo further potentially scope will required the overall data volumes and integration
scrutiny change need to be environment will impact based on the
reevaluated strategy capacity pivot

Implementation Guide: Success by Design: Guide 4, Drive app value 83

 Conclusion
As we have seen throughout this chapter, achieving true digital trans-
formation is not a sprint, but a marathon. It’s a continuously evolving
journey with iterations and changes that need to be effectively man-
aged. It starts with understanding the changes to your business model,
using triggers to accelerate transformation, and adopting the lean
startup mindset to build with an intention to test, learn, and change
while staying focused on driving maximum business value in the short-
est amount of time.

Going live with your MVP is the first step to driving value. You will iter-
atively refresh your transformation roadmap, revisit and refresh your
program goals, adopt new product capabilities, and drive meaningful
expansion while staying true to the core program value and transfor-
mation goals. Following this approach can establish your organization
as a leader in your industry.

The holistic value of SaaS is much broader; it goes beyond just the
technology and features of your application. You’re investing not only
in the technology but also in the research and innovation that drives
the product changes. Simplifying experiences and improving interop-
erability further reduces costs and drives business innovation.

As a final takeaway, we can’t emphasize enough the importance of

staying engaged and investing in learning. Stay on top of key product
announcements, industry trends, and research studies on technology
trends. Engage with product leaders via events, meetups, and advisory
boards. This will significantly boost your ability to achieve success.

Implementation Guide: Success by Design: Guide 4, Drive app value 84

 Checklist

Approach to digital Transformation roadmap

transformation Create an effective MVP scope that goes beyond the
bare minimum to accelerate value realization.
Have a clear understanding of the core business model
and how the application being developed impacts the Define your transformation roadmap and structure the
key elements of the business model. phases to maximize time to value and drive maximum
business impact.
Clearly articulate and communicate business transfor-
mation goals with specific measurable KPIs. Identify the expansion models and incorporate that into
your program phases.
Ensure the implementation team understands the
business processes in scope and the business teams are Adopt the latest capabilities to deliver continuous
closely engaged to optimize the process where needed. innovation to business.

Change stream
Account for the various change streams that will impact
the program and ensure budget and timelines are
adjusted to accommodate.

Implementation Guide: Success by Design: Guide 4, Drive app value 85

 Case study
Dynamics 365 helps
adapt business
processes to change
One of our clients, an accounting network, is facing unique and unex-
pected challenges. The COVID-19 pandemic has inhibited face-to-face
work, threatened global supply chains, and shifted regulatory and
political landscapes.

To maintain business continuity and stay connected with customers,

the company is adapting their digital selling techniques with the help
of Dynamics 365. With 360-degree views into each profile, sales staff
can work remotely without interrupting the customer experience.
The solution also works with the firm’s risk assessment tool to identify
vulnerable areas caused by the pandemic, allowing staff to create
proactive plans for their clients. The company is keeping their business
personal, up-to-date, and resilient by optimizing digital solutions in a
time of uncertainty.

This successful pivot clearly highlights the power of Dynamics 365 as

an adaptable SaaS cloud platform that can not only deliver strategic
solutions but also adapt to changing business needs quickly.

Implementation Guide: Success by Design: Guide 4, Drive app value 86

5 Guide
Implementation
strategy
Implementation Guide: Success by Design: Guide 5, Implementation strategy 87
 Introduction
Organizations embarking on a digital
transformation journey consider implementing
business applications as key critical milestones
for their business success.
The journey from a business vision to a successful implementation of it
requires the business and IT owners to define an overarching strategy that
acts as a framework to plan for predictable outcomes. An implementation
strategy defines these strategic choices and decisions made in planning,
implementing, and deploying Microsoft Dynamics 365 applications.

Along with new technology comes substantial change to a company’s

business processes. It is critical to choose a process-focused, user-cen-
Key concepts that will help in defining
a successful implementation strategy.
tric implementation strategy and methodology that not only supports
1 Key aspects digital transformation but also helps drives change management and
2 Business drivers user adoption. All of these factors are critical to making an implemen-
3 Success metrics tation successful.
4 Methodology

5 Deployment
In this chapter, we explore common industry methodologies and look
6 Change management

7 Conclusion
at common deployment strategies and high-level views on change
management strategy.

Implementation Guide: Success by Design: Guide 5, Implementation strategy 88

 Key aspects of
implementation strategy
The foundation of a successful implementation begins with planning
for key factors that define the overall implementation strategy. Let us
explore these key considerations.

Understand vision and business drivers

Project teams should be aware of the vision and strategic intent of the
customer stakeholders. Having a clear grasp of the key reasons behind
the implementation of a project and a collective understanding helps
the project team stay aligned with the project’s vision of addressing
significant business challenges.

It is important that senior management and project sponsors are clear

about the business requirements and help to establish the scope and
expected outcomes of the project. These goals help drive any decisions
that may be needed during the implementation cycle, for example,
prioritizing delivery of certain functional features over others.

This helps deliver a solution aligned with business needs and allows for
change management and a higher user adoption.

Identify key success metrics

After establishing a common understanding of the business challenges
and priorities, it is equally important to have an agreement on what
the key performance indicators (KPIs) that measure success look like.
The KPIs should be tangible metrics, for example opportunity conver-
sion rate, or average call handling time.

Understanding and identifying the KPIs is critical to ensuring the scope

being delivered is aligned with the expectations from the business
stakeholders and generates a tangible return on investment on the
Dynamics 365 investments.

Implementation Guide: Success by Design: Guide 5, Implementation strategy 89

 Some of the examples of success metrics/KPIs include:
▪ Improve the opportunity conversion rate from x % to y %
over z period.
▪ Reduce sales cycle from x days to y days over z period.
▪ Increase net new customer adds from x to y over z period.
▪ Reduce average call handling time from x mins to y mins ove
z period.
▪ Improve service request turnaround time from x days to y days
over z period.

The KPIs mentioned above are not exhaustive, customers have specific
KPIs that cater to their business scenarios. It is important to acknowl-
edge that each KPI the customer identifies and prioritizes has a direct
impact on the functionality that would be deployed on the Dynamics
365 Business Applications solution. Consequently, having a definitive
list of metrics to refer back to enables prioritization of the right use
cases and allows customer stakeholders to gauge the project success in
a tangible way.

Identify and allocate right resources

Setting up a productive team structure is critical to the success of the
project, which includes identifying the stakeholders, business process
owners, the team that delivers the projects, and the end-user represen-
tatives who can be engaged from the beginning of the project lifecycle.

The key roles required for all projects can often be grouped into two
types of team structures.
▪ Steering committee  The group of senior stakeholders with the
authority to ensure that the project team stays in alignment with
KPIs. As they monitor the progress they can help make decisions
that have an impact on overall strategy, including budget, costs
and expected business functionality. Steering committees usually
consist of senior management members and group heads whose
business groups are part of the identified rollouts.
▪ Core implementation team  This is the team doing the actual
execution of the project. For any Dynamics 365 implementation
project, core implementation teams should include project

Implementation Guide: Success by Design: Guide 5, Implementation strategy 90

 manager(s), business subject matter experts, solution architects,
functional and technical consultants, developers, testers, change
management advisors, and user champions.

Process owners engagement is critical to ensure complete understand-

ing of business processes. Their understanding is required in case of
any process changes that may result from alignment with out of the
box features. User representation should also be planned properly to
adequately cover all the personas. Lack of availability of any of these
may result in delays and/or incomplete understanding of processes
and also impact user adoption. The involvement of process owners and
users can be staggered depending on the implementation phase. This
should be planned early on to allow business teams to make the re-
quired changes to their schedules and be able to contribute efficiently
during the project lifecycle.

The project may require additional roles depending on the scope and
methodology. Some roles may need to join temporarily during the
project lifecycle to meet specific needs for that period.

For more details on team setup considerations, refer to Chapter 2,

“Success by Design overview,” and Chapter 8, “Project governance.”

Allocate budget and funds

Planning a budget for Dynamics 365 cloud applications typically
includes costs such as subscription, storage, implementation lifecycle
activities of designing, testing and deploying the solution, migration
from legacy applications, training, change management, and continu-
ous improvements.

Project lifecycles require constant assessments with careful considerations

and actions for the variance between budgets versus actual costs.

A common decision that most teams face is finding the right balance
between delivering a high value solution faster by using the off the
shelf capabilities versus extending the product capabilities to imple-
ment the business requirements and needs. Extending the Dynamics

91
 365 applications not only requires initial development costs but also
maintainability and supportability costs in the long run. This is an area
where implementation teams should carefully revisit the cost impact.
In the cloud world with rapid availability of new features and capabil-
ities, there is a need to rethink the investments required for extending
the application. Refer to Chapter 15, “Extend your solution,” for more
details on assessing this impact.

Project management teams should give appropriate weight to areas

that help them stay aligned with the overall project goals and objec-
Forrester’s Total Economic Impact tives. It has often been seen that as project scope and timelines change,
(TEI) of deploying various Dynamics the activities that fall towards the end of the project lifecycle are usually
365 apps is a good recommendation
to assess the costs and benefits descoped to meet budget constraints. These activities can include
associated with Dynamics 365
performance testing, data validation testing, training, and change
Implementation.
management, which, when compromised, can significantly reduce the
Some of the exclusive TEI done by
Forrester are:
quality and success of the implementation. It is important to revisit the
The Total Economic Impact of project goals and success factors that are defined during initial stages
Microsoft Dynamics 365 for Finance
and Operations to carefully revise the scope and timelines without compromising on
overall quality and user centric approach.
The Total Economic Impact of
Microsoft Dynamics 365 Field
Service
An overall approach is recommended since it helps define the total
The Total Economic Impact of impact by identifying the cost, benefit, flexibility, and risk factors that
Microsoft Dynamics 365 Customer
Service affect the investment decision, helping organizations improve their
overall business goals of winning, serving, and retaining customers.

Choose a methodology
Before we discuss choosing a methodology for your Microsoft
Dynamics 365 project, we need to understand why a methodology is
important for Dynamics 365 projects.

Why methodology is important

A methodology is a prescriptive definition of activities to be under-
taken for a given project or engagement. It describes the use of a
collection of methods to achieve predictable outcomes.

Implementation Guide: Success by Design: Guide 5, Implementation strategy 92

 Methodology is the core of project execution that allows the team to
take the right path towards:
▪ Improved consistency and predictable outcomes.
▪ Setting clear definitions of project phases, milestones, deliverables,
and entry and exit criteria to each phase.
▪ Setting clear definitions of roles and responsibilities required for
the project.
▪ Reducing risks of missing critical activities required for a successful
outcome.

Following a methodology is akin to taking a planned journey; the

roadmap to success (as illustrated in Figure 5-1) involves the following:
▪ Plan ahead  Ensure you have clearly defined the destination
(project scope), and have all the right tools (resources, governance,
and methodology) and checklists (project monitoring) before
starting this journey.
▪ Know the road  Define the route (solution blueprint) to set clear
expectations on start and end dates at milestones so you can
meaningfully monitor progress and the team understands what is
expected from them.
▪ Avoid pitfalls  Utilize proven best practices, as advocated in the
methodology, that minimize risks, costs, and scope creep.
Fig. 5-1

Plan ahead Know the road

Make sure you have all the right Set clear expectations around start and
tools and checklists for the trip. end dates and project milestones.

Start

Go
live

Avoid pitfalls Get there faster No surprises

Minimize risks, costs, and scope Take advantage of strategies Delight executives and customers by
creep with proven best practices that save time and money. delivering the right results on-time
from thousands of employees. and on-budget.
93
 ▪ Get there faster  Take advantage of strategies that help you save
time and money. Plan, action, and regularly monitor and adjust to
ensure you are taking the most efficient route to your destination.
▪ No surprise  Delight users by delivering a user centric solution
and business stakeholders by delivering results on-time and within
budget.

Choose an implementation methodology

Let us take a look at the specific methodologies that can be used for a
successful Microsoft Dynamics 365 implementation.

We often see three types of models used to implement Dynamics 365

Business Applications projects.

Waterfall model
A delivery execution model based on sequential, steadily flowing, like a
Waterfall, series of phases from conception, initiation, analysis, design,
development, testing, deployment, operations, to maintenance.

Agile model
Agile development is not a methodology in itself, but an umbrella
term that describes several Agile methodologies. Some of the industry
known methodologies are Scrum, XP, DSDM, Sure Step for Agile (for
Business Applications).
▪ Scrum methodology  A delivery execution methodology based
on Agile principles where requirements are in flux or unknown
prior to beginning a sprint. Sprints involve requirements analysis,
design, build, test, and user review. Sprints are a fixed duration
(usually 30 days), as illustrated in Figure 5-2.

Fig. 5-2

24h
30 days

Product backlog Sprint backlog Sprint Working increment

of the software

Implementation Guide: Success by Design: Guide 5, Implementation strategy 94

 Hybrid model
A delivery execution model based on Agile principles where require-
ments within the context of business processes are defined, analyzed,
Iteration and sprint is a timeframe
and prioritized and an overview design solution blueprint is generated
in which teams deliver incremental
value in the form of working tested prior to beginning the sprints. Sprints involve refining designs, build,
software. Iterative approaches use
the term iterations while Scrum
test, and user reviews. Sprints can be of varying durations.
approaches use the term sprints.

Now let’s take a deeper look at each of these methodologies.

Agile
Success by Design Framework is
An Agile implementation is an iterative approach that uses a number
methodology agnostic and is aligned of iterations or sprints.
to ensure proactive guidance and
predictable outcomes irrespective
of chosen methodology. For more
In Microsoft Dynamics 365 Business Applications projects, the majority
information, refer to Chapter 2,
“Success by Design overview.” of requirements are delivered by the packaged software. There are
specific challenges and tasks that are out of the box and need to be
managed and mitigated as part of the project methodology with the
user stories contained in the form of a backlog. Using those stories, you
carve out plans and within them a number of iterations or sprints of
development and testing are executed. Within each sprint, you have a
number of user stories outlining the requirements.

Agile promotes a collaborative process between the resources that

own and specify the requirements for the solution and the resources
responsible for the development and rollout of the solution.

The idea is to release software faster, take early and regular feedback,
adapt, and release again. This cycle continues until all the requirements
are met or the backlog is cleared.

An example of Agile practices using the methodology prescribed by

“Sure Step for Agile” is shown in Figure 5-3.

Implementation Guide: Success by Design: Guide 5, Implementation strategy 95

 Fig. 5-3
1.1.1 4.1.1 5.1.1

management
Project management Project Project
Analysis

Program
management management
1.2.1 1.2.2
Conduct Gather user
solutions training
overview requirements
1.3.1 1.4.1 Design
Conduct detailed
Agile preparation

Gather
4.2.2
business process

Training
business
Conduct end
analysis requirements
user training
1.7.1
Set up DEV Development
and other
non-production 1.4.2
environments

process analysis
Conduct fit gap
analysis

Business
1.8.1
Establish Deployment
integration
strategy

1.9.1
1.4.3
Gather data
Define

Requirements and
migration Operation

configuration
solution backlog
requirements
5.4.2
4.4.2
Transition to
Go live
support

30-day sprint cycle

2.1.1
Define sprint

Custom
coding
backlog

2.1.4 2.1.2
Conduct sprint Conduct sprint
post mortem planning meeting

and testing
4.6.2

Quality
Daily sprint cycle Conduct UAT

3.1.1 3.1.2
Agile execution

Conduct daily Analysis and

sprint cycle design for
meeting sprint cycle

Infrastructure
3.1.3
3.1.6 4.7.1
Create scripts
Conduct Build production
for testing
sprint testing environment

3.1.4
3.1.5
Sprint
Generate
configuration
daily build
and development
and interfaces
Integration

2.1.3
Conduct sprint
technical preview

4.9.3
3.6.1 3.7.1
migration

Final data
Finalize
Data

Conduct
migration to
solution production
production
testing specification

Implementation Guide: Success by Design: Guide 5, Implementation strategy 96

 Waterfall
The Waterfall approach to solution delivery is a sequential process that
depicts a linear flow of activities from one phase to another, culmi-
nating with the solution being promoted to production and then into
operation.

This is a traditional methodology of implementing on premises busi-

ness applications. It is a linear, noniterative approach of implementing
and delivering projects.

The phases, milestones, and deliverables are clearly defined, and you
only move to the next phase when a prior phase is completed.

In the modern cloud world, early stakeholder buy in is critical which

requires delivering quicker results, which makes waterfall technique
not so suitable.

An example of the Waterfall model as prescribed in Sure Step

Methodology is depicted in Figure 5-4.
Fig.5-4

Analysis Design Develop Stabilization Deploy

Organization

Training plan

Deployment plan User training

Baseline
project plan

Functional design
Process and
functional UAT scripts UAT
Transition to
Solution

requirements
Technical design support
Solution testing
Fit gap analysis
Solution design

Data migration
Technology

requirements
Integration and Production Production
interface design specifics environment
Environment
specification

Implementation Guide: Success by Design: Guide 5, Implementation strategy 97

 Hybrid
In the current landscape of rapidly changing business needs and
changing technology trends, projects commonly adopt a hybrid
approach. This is a combination of Waterfall and Agile practices that
implementation teams use to meet their project needs. It allows teams
to tailor an approach that enables a successful implementation of
Dynamics 365 applications. It is also our recommended approach for
Dynamics 365 engagements.

Figure 5-5 shows an example of a hybrid approach, based on

Dynamics Sure Step 365.
Fig. 5-5

Noniterative Noniterative Iterative Noniterative Noniterative

Solution Solution Support

Mobilization Initiation Build Deployment Operation
modeling testing transition

Analysis Process
Design End-to-end
Development Performance
Iteration testing UAT

Once per
Once per project Once per release iteration Multiple cycles per release Once per release
in a release

With this approach, we can manage initial activities, like initiation and
solution modeling, and final activities like system integration testing,
user acceptance testing, and release to production, in a noniterative
way. Then the build activities, such as requirement detailing, design,
development, testing, are completed with an iterative approach.

This helps provide early visibility into the solution and allows the team
to take corrective actions early on in the overall cycle. This approach is
considered to use the best of both Waterfall and Agile approaches and
is a win-win for any size of implementation.

In the hybrid approach, the focus is on a regular cadence of iterative

and incremental releases of software features.

Implementation Guide: Success by Design: Guide 5, Implementation strategy 98

 We stress for customer stakeholders to participate in the process, es-
pecially in the business validation reviews that happen after each sprint
cycle. These reviews are placed at regular intervals during the project
schedule, and allow for faster valuation and better stakeholder buy in.

Through these reviews, customer stakeholders get an opportunity to

see working software, have discussions with the team, and provide early
feedback. The readiness of business processes is validated for produc-
tion use by an invited business audience by showing the incrementally
improving readiness across all the components in the system as working
software, like design, data configuration, data migration etc.

The idea is to get early feedback and validation from the business team
on the business requirements scope and understanding. This approach
allows the project to showcase working solutions faster than other
methodologies and achieves a higher rate of implementation success.

Define deployment
strategy
A deployment and rollout strategy describes how the company is
going to deploy the application to their end users. The strategy chosen
is based on business objectives, risk propensity, budget, resources, and
time availability, and its complexity varies depending on each imple-
mentation’s unique scenarios. There are several key factors to consider
prior to choosing any of the approaches detailed below.

▪ What is the MVP (minimum viable product) needed for faster val-
ue to customers and then later plan for continued enhancements?
Refer to Chapter 4, “Drive app value,” to define an MVP strategy.
▪ •Is this a multi-country rollout or single-country rollout?
▪ Is there a need to consider pilot rollouts prior to wider team rollouts?
▪ What are the localization requirements, especially in scenarios of
multi-org or multi-region rollouts?
▪ Do we need to phase out a legacy application by a certain key
timeline or can it be run in parallel?

Implementation Guide: Success by Design: Guide 5, Implementation strategy 99

 ▪ What is the extent of impact on end users’ day-to-day productivity?
▪ What training needs do end users have?
▪ What is the complexity of integrations during the rollout?

Taking these factors into consideration, you should define a rollout

strategy that helps the organization deploy the application in the most
effective manner.

The following are some typical deployment strategy models.

Big-bang
In the big-bang approach, as the name suggests, the final finished
software solution is deployed to production and the old system is re-
tired. All users start using the new system on the go-live date. The term
“big-bang” generally describes the approach of taking a large scope of
work live for the entire user base of your organization simultaneously.

The selling point of the big-bang approach is that it concentrates the

focus for a shorter rollout period compared to a phased rollout and all
users are on-boarded at the same time.

An example of big-bang rollout is when an organization that was using

a legacy business application for all of its corporate office finance team,
communicates to all end users to stop using the old system and start
using the new Dynamics 365 Finance on the go-live date.

The whole implementation can be very large and can encompass

multiple countries, departments, or business units. During transition,
the new solution is deployed through a number of activities while the
old system is turned off so the new system can go live.

The risks of this approach are that the entire project could be rushed,
important details could get ignored, and business processes trans-
formed may not be in the wider interest of the organization. In an
aggressive big-bang rollout, the risks are magnified due to the dual
combo of a large scope and shortened rollout period.

100
 Change management can also suffer as people are less inclined to use
the system that may not be solving the business problems it was meant
to solve.

With such large changes happening, there is a potential for issues to arise
post go-live and falling back on the older system is very hard, if even
possible. It is critical to ensure proper contingency planning for business
continuity, and support mechanisms to deal with post go-live issues.

Large project scopes are at much greater risk by using the big-bang
approach since the delivery of finished software to the production takes
up a longer timeline, there is more to organize during transition. If the
final rollout runs into challenges, the wider user community is impacted
Fig. 5-6
due to the resulting delays. If Waterfall methodology is used as the
implementation, then the end
Pros Cons users don’t have a feel for the real
▪ Shorter, condensed ▪ May not accommodate for rapidly
system when it is finally rolled out.
deployment times, minimizing changing market scenarios and
organization disruption product capabilities, leaving a
lack of solution alignment at However, if hybrid methodology
▪ Lower costs from a quick rollout
without loss of overall work, but deployment causing a negative is used, it is possible to involve
with an added need to account impact for end users end users from the beginning,
for resource costs to manage the ▪ Daily task delays from users keeping them updated on what
complexity learning to adapt to the new is landing on go-live and keeping
application without any prior
▪ Same go-live date for all users any surprises in check.
experience
▪ Reduced investment in
temporary interfaces ▪ Transitioning back to legacy
For large and complex imple-
systems can be costly and
difficult, if even possible
mentations, big-bang projects
can require many months or even
▪ Any failures during the rollout have
a high impact on maximum users years to complete. Given the pace

▪ Heightened probability of risks

of change in most industries it
emerging from introducing large may not be feasible to wait that
changes for large groups of users long for implementation, espe-
▪ Heightened probability of issues cially since the business that was
coming up in go-live analyzed for requirements could
▪ No opportunity to learn from have different needs and prior-
early launches to change things ities a year later, as illustrated in
▪ Costs of ramping up a large rollout Figure 5-6.
team for a short period of time

Implementation Guide: Success by Design: Guide 5, Implementation strategy 101

 Phased rollout
In a phased rollout approach, we are releasing the software in a num-
ber of phases or releases. The phases can be planned in several ways
and can be based on modules, business priority, business units, or
Fig. 5-7 geographies. One approach for a phased rollout can be to release an
initial set of capabilities in the first
Pros Cons phase and build on that by releas-
▪ Higher business value features can ▪ Longer implementation and deploy- ing the rest of the requirements in
be prioritized. The implementation ment timeline due to multiple go live subsequent phases. At the release
team may also choose to prioritize events.
of the first phase for a business
lower complexity features to adapt ▪ Complexity of data migration from
unit or module, it is expected that
to the new functionality. legacy application to the target
the specific business unit’s end
▪ Unlike big-bang, phased rollout application increases as it needs to be
releases can bring new use cases planned in a staggered approach. users stop using the legacy system
gradually, allowing for more buy in and move to the new one. There
▪ Organization needs to plan for contin-
from end users as they get used to uous disruption over longer periods of could still be temporary scaffolding
the new system. time, like parallel efforts of supporting integrations setup to integrate the
▪ Project teams can optimize later deployments while working on future new system with the legacy systems
phases by incorporating the phase implementation.
until all phases are rolled out.
learnings from earlier ones. ▪ Employee morale may suffer as they
▪ Risk of disrupting business is less as face change fatigue. Phased projects
For example, a team can start the
errors found in the initial phase are need lots of focus and coordination
limited to specific business areas as the team might start losing mo-
implementation with one business
and users and rest of the business is mentum after first few phases. These unit. Once implementation is
not affected. projects don’t have the benefit of complete, they travel to the next
▪ When phases are based on func- focused intensity as the big-bang business unit. As the team gains
approach does. It is important that
tionality, rather than module, experience and learns from past
geography, or business unit, it the change management team keeps
mistakes, the subsequent rollouts
typically results in a faster time to a tab on employee morale and plans
initiatives to keep their interest alive. become smoother. This leads to
value.
▪ Need to implement temporary
less risk and more employee adop-
▪ For large implementations, it
scaffolding solutions to manage tion, as illustrated in Figure 5-7.
reduces the level of investments
and resources needed for ramp up integrations between new and legacy

for each deployment relative to a systems until the new solution is fully

big-bang approach. rolled out. Parallel rollout

▪ Scope creep can occur since project
In a variant of phased rollout ap-
timelines are drawn out. There can be
proach, parallel rollout, the legacy
a tendency to incorporate changes
and rework, which can affect the system is not discarded but kept
work for subsequent phases and their alive along with the new system.
timelines. More than a rollout approach, it is

Implementation Guide: Success by Design: Guide 5, Implementation strategy 102

 a validation technique allowing users an opportunity to learn, test, and
get comfortable with the new system. Typically, for few months both
systems are actively used and users are required to key in information
in both systems.

▪ There is much more effort required from the users as they double
key information.
▪ This rollout may be preferred by projects where the business risk is
very high and cannot be easily mitigated by testing. At the same
time, we need to be mindful of making sure team is not cutting
corners on testing and training efforts.
▪ This rollout is less and less popular due to the extra efforts and
costs involved in keeping two systems.

Rollout strategy goes hand in hand with change management strategy

as several activities of change management play a key input to suc-
cessful rollout, as illustrated in Figure 5-8.
Fig. 5-8

Pros Cons
▪ Less risk ▪ High efforts and high costs

▪ Users take their time to gradually plan to migrate to the ▪ Needs for data duplication can get complicated. We
new system recommend having a clear exit plan that is time or
workload based, for example, all existing cases closed
in old system

▪ May mask poor training and testing efforts

Define a change
management strategy
Adoption and change management planning is a critical aspect of
implementation strategy. Change is never easy for any organization
and requires thoughtful planning and a structured approach to
achieve high adoption and usage. A change management strategy
with a clear action plan ensures smooth rollouts and alignment with
stakeholder and end user expectations. In the absence of a planned

Implementation Guide: Success by Design: Guide 5, Implementation strategy 103

 change management approach, many business applications tend to
become purely a reporting tool rather than a value driven user centric
application.

Bridge the gap between business

requirements and results
All too often, projects meet requirements without a true focus on
adoption and business outcomes. The gap that exists between technol-
ogy solutions and business outcomes directly relates to the users who
experience a change in their day to day work. Change management
focuses on bridging this gap by effectively supporting and equipping
the user community to successfully adopt the transition. A change
management mindset is required across different functions, including
the project development teams, to keep the end user experience as a
primary design consideration.

Increase likelihood of project success

PROSCI’s correlation data from over 2,000 data points and 10 years shows
that initiatives with planned change management are six times more likely
to meet objectives than those without a formal change management
strategy. The data is abundantly clear. The better we apply change
management, the more likely we are to deliver on project objectives.

Take the change out of change

Change is difficult. To the degree that we can, we want to remove the
Change is never chance or variability associated with change. Project management has

easy for any accomplished this by providing direction on sequencing milestones,

deliverables, activities, and resources over the lifecycle of an effort.
organization and Unless we proactively support and guide people through the changes
requires thoughtful our projects bring, we leave the likelihood of them embracing change
planning and a to chance. Change management addresses this by providing employees

structured approach with the preparation, support, and skills they need to succeed in change.

to achieve high PROSCI’s framework that is followed at Microsoft describes the three
adoption and usage. sequential steps that are followed in change planning.

104
 Preparing for change
The first phase in PROSCI’s methodology helps change management
and project teams prepare for designing their change management
plans. It answers these questions:
▪ How much change management does this project need?
▪ Who is impacted by this initiative and in what ways?
▪ Who are the sponsors we need to be involved with to make this
initiative successful?

Managing change
The second phase focuses on creating plans that integrate with the
project plan. These change management plans articulate the steps that
you can take to support the users being impacted by the project.
▪ Communication plan  Communications are a critical part of the
change process. This plan articulates key messages that need to go
to various impacted audience. It also accounts for who sends the
messages and when, ensuring employees are hearing messages
about the change from the people who have credibility with them at
the right time. Communication is an understated yet very significant
aspect that keeps the various teams connected throughout the entire
journey. Some of the key communication aspects to consider are:
▫ Providing a sneak peek into what is in it for end users
▫ Launch and availability communications
▫ Project sponsor vision and direction communication
▫ Communication frequency and approach
▫ End user feedback incorporation
▫ Status reporting, including steering committee reports
▪ Sponsor roadmap  The sponsor roadmap outlines the actions
needed from the project’s primary sponsor and the coalition of
sponsors across the business. In order to help executives be active
and visible sponsors of the change, it identifies specific areas that
require active and visible engagement from the various leadership
teams, what communications they should send, and which peers
across the coalition they need to align with to support the change.
▪ Training plan  Training is a required part of most changes and is
critical to help people build the knowledge and ability they need
to work in a new way. The training plan identifies the scope, the
intended audience, and the timeframe for when training should

Implementation Guide: Success by Design: Guide 5, Implementation strategy 105

 be planned for. It is important that the training plan be sequenced
in a way that allows for awareness and desire building before they
are sent to training. A common aspect that tends to be overlooked
is cloud focused training for IT administrators. As organizations
implement Dynamics 365 cloud applications, it also may mean a
significant change for IT teams, so it is important to plan training
that keeps their persona in perspective. They may require training
not only on Dynamics 365 administrative activities, but also for a
general understanding of cloud concepts.
▪ Coaching plan  The coaching plan outlines how you engage with
and equip managers and people leaders to lead the change with their
own teams. Managers can play a significant role in aiding the change
management efforts, but they need to be engaged as employees
themselves first and allowed to work through their own change
process. Once that’s done, you can give them the information and
tools to lead the same change process with their own teams.
▪ Resistance management plan  Resistance to change is ex-
pected, so proactively defining the activities to mitigate the areas
of concerns should be initiated early on in the project lifecycle.
Engaging user champions from the end user community early on
to build a user centric solution contributes significantly towards
addressing this risk.

Reinforcing change
Equally critical but often overlooked, this third phase helps you create
specific action plans for ensuring that the change is sustained. In this
phase, project and change management teams develop measures and
mechanisms to measure how well the change is taking hold, gauge
user adoption, identify alignment with KPIs, and correct gaps.

A successful change management strategy requires continuous exec-

utive sponsorship, integration with the project management teams,
employee engagement, and frequent and open communication.

Conclusion
When a team defines an implementation strategy, they set clear

106
 expectations and guidelines for the team and wider business on how
the project goals are going to be attained. This helps define a clear
translation of business goals into a structured implementation plan.

To summarize, a well-rounded implementation strategy takes the

following into consideration.
▪ Build a common understanding of vision.
▪ Identify the methodology based on the specific factors that are
driving the implementation.
▪ Follow the tried and tested approach that is defined by the meth-
odology, do not leave activities out just for convenience.
▪ Help to take feedback early in the implementation journey by
doing a “show and tell” after each sprint/phase depending on the
chosen methodology.
▪ Keep organizational culture under consideration while choosing a
methodology. Organizations that have followed a traditional ap-
proach such as Waterfall may need additional change management
planning to embark on a modern approach such as Agile or Hybrid.
▪ Consider which industry practices may also have a bearing on
methodology. Some industries may have their own tailored ap-
proach and that needs to be kept into consideration.
▪ Formulate a focused approach towards user adoption by creating
a change management plan.
▪ Plan an approach of communicating to various stakeholders,
including leadership and end users.
▪ Continue to improve throughout the process; solicit end-user
sentiment and act!

Implementation Guide: Success by Design: Guide 5, Implementation strategy 107

Initiate
6 Solution architecture design pillars
7 Process-focused solution
8 Project governance
9 Environment strategy

108
6 Guide
Solution
architecture
design pillars
You must be sure about the
solution before you start building,
because failure is not an option.

Introduction
You can’t have a solution without first having a vision.
When you know the answer you’re looking for, only
then you can find a solution to get you there.
But it’s not always easy to know and articulate what you want, let
alone identify the elements that are essential for creating a blueprint of
your solution.

This chapter outlines the pillars of a successful solution architecture de-

sign that support your vision and provides you with the building blocks
you can use to create a coherent and satisfying outcome. By utilizing
the Success by Design framework, we present a structured approach
for designing a solution that’s enabled by Dynamics 365 apps.

Solution architecture
design pillars
Most of us know what “architecture” means in the building industry,
it includes the job of the architect, the scope of the work, and what’s
ultimately delivered. The architect starts with the project requirements

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 110
 and the budget, and then works on a blueprint of the design. Once the
physical labor begins, the foundation is laid, the house is built, and the
interior is completed. Many years of maintaining and improving the
building may follow. But why and when do you need an architect to
design a solution for your organization?

Let’s take the simple case of building a new house for a dog. In most
cases, this is a single-person job with a general vision of the final prod-
uct. Building a doghouse is cheap, and the risk of failure is low, so a
trial-and-error approach is one option.

But what if you were asked to build something as complex as the

Sydney Opera House in Australia? The approach would be completely
different: the job can’t be completed by a single person, the design
needs to be carefully planned, and architects, construction work-
ers, plumbers, electricians, and decorators have to coordinate their
workflow. With so many moving parts, there must be well-defined
processes, an explicit design, and powerful tools. When all the right
pillars are in place, the likelihood of failure is dramatically reduced.

Solution architecture follows a systematic approach by identifying the

desired solution and the necessary building blocks, where each builds
on the previous block and provides a foundation for the next block.
Solution architecture breaks up the work into logical sections, and
coordinates those sections, managing and mitigating the risks along
the way.

Building blocks of
solution architecture design
The journey starts with a Vision. The vision details are articulated in
a Business strategy. To build the solution Processes, People and Data
needs to be joined in a Solution architecture (Figure 6-1). A solution
strategy outlines how those elements can be addressed by using the
technology of Dynamics 365 apps. Project and change methodologies,
supported by governance and control methods, define the workflows
and the delivery.

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 111
 Solution architecture design incorporates the business vision and its
Fig.
implementation into a blueprint. The first version is created as part of
6-1
the pre-sales process and forms the initial high-level
understanding of what you plan to build.

Vision Solution architecture enables implementation of a

vision, and typically begins after the business strate-
gy has been crafted, but before the solution is built
Business strategy and implemented (Figure 6-2). Architecture creates
a picture of the finished product and includes the
Processes
specifications needed to build it.

People A solution architect provides the necessary quality con-

trol by ensuring that key requirements are met, with all
Data relevant interdependencies considered and managed.
A solution architect also serves as a translator between
the business and IT teams to articulate the impact of

Solution strategy technology on the business, and vice versa.

Technology
of Dynamics 365 apps
Vision
A vision is the desire to achieve something, to
change the present and improve the future.
Methodology
When an organization decides to go through a
transformation, it’s usually because one or more
Fig.
6-2

Business Solution Solution to be built

vision architecture and implemented
design

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 112
 people had the ability to predict what was coming and expressed a
desire for change, or a change was forced by industry disruption or
regulations. Such a change could be in the form of a mission statement
or a business case listing its objectives, which might include:
▪ Modernizing legacy enterprise business applications.
▪ Finding unified and faster ways of working.
▪ Earning higher profits.
▪ Achieving better service or product quality.
▪ Improving the user experience.
▪ Empowering users to drive greater value by building apps.

As the vision comes together, the plan for achieving it can start
taking shape.

Business strategy
Every vision serves a purpose, as does every organization, and any
solution needs to be aligned with this purpose. A business strategy
supports your vision by answering fundamental questions, such as:
▪ Why are you making this change, and what are the anticipated
benefits? What is the business value sought by the solution?
▪ Where do you imagine the organization will be in five, 10, or 20 years?
▪ What business capabilities can your organization offer with the
new solution? What business processes can you run? What infor-
mation and data would you like to record and report on, in line
with your organization’s services or products?
▪ Which clients, customers, or people inside the organization will be
served by the new solution, and who will be affected by it?
▪ Would you like to improve your current line of business or are you
open to a new industry?
▪ When do you plan to have the vision materialized? What is the
high-level timeline? And do you intend to deliver the solution at
once or grow in stages?
▪ Where are the regions, geographically and in terms of business
functions, to which the solution will apply? Will you apply it to all
or just some of them?
▪ Who is going to plan, design, and deliver the solution? Do you

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 113
 have a preferred partner, or do you need to select a vendor? Who
Why should you
will support and maintain the solution post go-live?
define your processes? ▪ How will you incorporate technology into your solution? (This is
It creates consistency by the first step of solution design and the link to your solution strate-
allowing for consolidation and gy, as well as the business case for IT transformation.)
unification of processes across
locations, teams, and systems.

It reduces complexity by
standardizing and rationalizing
Processes
previously complicated and
cumbersome procedures. Process architecture is a commonly understood, shared view of all busi-
ness processes that an organization uses to deliver a product or service.
It eliminates guesswork and It represents how an organization operates, in a structured order.
potential ambiguity by stream-
Complete, accurate, and well-organized process architecture is the first
lining operations and improving
communications. and most important pillar for a sound solution design. It confirms end-
to-end business understanding, provides a structure for understanding
It promotes productivity by
the scope, and serves as the basis for testing and training. It also forms
eliminating inefficiencies and
establishing one workflow for a map for requirements gathering. The processes are constructed at
all users. different levels to reflect specific areas, functions, locations, and teams.
Process architecture is often referred to as a target operating model, a
It guarantees quality by
maximizing attention to detail
process library, or a catalog.
and ensuring work is done in
a pre-defined, optimized way Any transformation starts with defining your processes, which is also a
each time.
good time to revisit and improve them.
It boosts morale by helping
team members take pride in
mastering the process, refining Dependent activities
their skills, and avoiding mis-
takes and missed deadlines. While we’re not going to take a deep dive into the business process
management life cycle and capability maturity, following the Success
It encourages continuous
improvement by giving team
by Design framework can help you identify important elements of your
members who follow the pro- solution design.
cesses a chance to give input on
how to improve them.
Process architecture includes multiple dependent activities:
It increases user acceptance ▪ Scope management  Defining the scope of the solution is the
by providing a higher-quality first step of the design. Ideally, you have a baseline process taxon-
experience at each launch.
omy with at least three levels of the process architecture. Start by
mapping requirements against it and marking which processes are
in scope. You can add to and take away processes from your initial
process library.

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 114
 Key deliverables related to processes

Process architecture map  Linear and cross-functional process flows

These show the input, output, and activities at process and
sub-process levels to help you figure out how a process must
change to meet requirements; you can also add details such as
assumptions, metrics, timelines, and costs.

Linear flow

This is a visual representation Product Product Product Product

of your business processes. ideation design development testing

Product
prototyping

Process catalog/ Product ideation phase

inventory and taxonomy Collect customer

suggestions

Test against
Conduct Put ideas into ideation Select ideas for
strategic
competitive analysis pipeline conceptual design
imperatives

Conduct ongoing
research

This is a sequenced list of the

processes, usually uploaded to
Process flows Route Load Outbound
Microsoft Dynamics Lifecycle planning building logistics

Services (LCS) and synchronized

with Azure DevOps.

Inbound Wave
Receiving Put away Count Pick Pack
logistics processing

Planning

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 115
 ▪ Business analysis  Initially, it’s likely that business users don’t
know or understand the application, and the tech partner doesn’t
understand the business, so a business analysis helps to join them
together. A business analyst gathers requirements and links them
to processes; conversely, a good processes structure drives the
requirements gathering by asking all necessary questions.
▪ Requirements management  Every functional and nonfunc-
tional requirement needs to be linked to at least one process. If an
applicable process doesn’t exist, you must slot it into the relevant
section of the process architecture. Azure DevOps is a good tool to
use for this success measure, as shown in Figure 6-3.

Fig. ▪ Solution design  Once you

6-3
Requirements have a better end-to-end
traceability matrix (RTM) Fit gap assessment business understanding,
it’s time to translate it into
your system processes. With
the Dynamics 365 solution
design, a Microsoft partner
works with the process leads
This links requirements throughout This is where you mark which
to develop the solution, and
the validation process, and requirement and respective
it can be helpful to create
is a key deliverable for process can be addressed by
requirements management. the system, and which ones process flows that show how
require customization. the processes can be run in
the system.

Fig.
▫ As part of the process, data goes in as input exchanged be-
6-4
tween people and applications, and data goes out as output in
People the form of documents, analysis, and reports (Figure 6-4).
▪ Test management  Once the solution is designed and devel-
oped, the process architecture establishes a baseline for testing.
Solution
design When you test every process, you ensure that every requirement is
s

ta addressed and the solution is appropriate. For more information,

Da

se
es

c
P ro refer to Chapter 14, “Testing strategy.”
▪ Training  Process architecture also defines and drives your training
content. You can use your process flows and guides as a first draft
for your training materials. Learn more in Chapter 7, “Process-
focused solution,” and Chapter 2, “Success by Design overview.”

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 116
 Key deliverables
related to people People
Organizational structure Even though not every process or activity is performed by a person,

Cross-functional there’s always a step or output that involves people, whether internal
process flows and maps (employees and contractors) or external (customers and vendors). This
is the key reason why and how the solution design is influenced by the
List of business users
or personas people pillar. People shape the solution design in many ways, including:
▪ Geographical location
Dynamics 365
security roles list ▪ Time zones
▪ Languages
Mapping between personas
and security roles
▪ Customs
▪ Internal and external hierarchies
▪ Organizational structure

This pillar of solution design is usually represented by the organizational

architecture, which visually intersects with process architecture in
process maps, and includes:
▪ Geographical structure
▪ Organizational groupings
▪ Line of business
▪ Reporting lines
▪ Segregation of duties

Key deliverables
related to data Data
Data governance strategy The third pillar of solution design is data.

Data architecture
A system with inaccurate, misleading, or partial data will lead to a failed
Data migration and implementation, so you must understand your data and how it fits into
integration strategy the overall solution. With the right data, you can identify actionable and
analytical information that improves business intelligence. Think about:
Data quality strategy
▪ Master data  This is static information, such as details about
customers and products, that you use and exchange in your busi-
ness activities.
▪ Transactional records  This refers to recorded business infor-
mation, such as customer payments and expense reports, that you

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 117
 create, store, and report as part of your business process.
▪ Documents  This refers to documentation, such as sales orders
and customer invoices, that you create as part of your business
process.
▪ Reports  This is organized or filtered information, such as a trial
balance or aged debt, serving as input or output for your business
activities.

At a minimum, you must start with an understanding of how your data

is currently utilized and how it supports your future vision. Planning
your data strategy upfront also allows for better scalability.

The world is evolving from silos of data to connected enterprise data

Refer to Chapter 10, “Data management,”
to gain a broader understanding of the that enables the digital feedback loop. That is why a data strategy that
various data functions you’ll need before
applies AI and analytics to make the data actionable is so critical to the
starting a project.
overall success of your design.

Solution strategy
Your solution strategy is a consolidated view and approach that defines
your overall solution. A solution blueprint is a living document with
several review points (Figure 6-5) during a project’s lifespan to help
you identify and take necessary actions, mitigate risks, and resolve
issues as they arise. In the Success by Design framework, the blueprint
is considered essential for a project’s success, and provides a view of
Fig. the overall solution architecture and dependent technologies.
6-5

Kickoff Go live
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

Initiate

Implement

Prepare

Operate

Solution SBR 2 SBR 3 SBR 4

Blueprint
Review Gap solution workshop Performance workshop Go live review
(SBR) 1 Environment workshop Test workshop Cutover plan workshop
Scheduling workshop ALM workshop
Customer kickoff BI workshop
Partner kickoff Integration workshop
Data migration workshop

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 118
 The Solution Blueprint Review workshop is a mandatory part of the
solution design experience, and helps you keep track of your solution
design’s progress to ensure that your vision and objectives are still
viable. It also allows solution architects and the implementation team
to review and gain an understanding of your:
▪ Program strategy
▪ Test strategy
▪ Business process strategy
▪ Application strategy
▪ Data strategy
▪ Integration strategy
▪ Intelligence strategy
▪ Security strategy
▪ Application lifecycle management (ALM) strategy
▪ Environment and capacity strategy

For more information and a list of activities Capturing these details helps you understand the project, and validates
in each Success by Design phase, refer to
Chapter 2, “Success by Design overview.” your solution design via the complete and well-organized processes,
data, and people enabled by Dynamics 365.

Technology
While technology does not drive all the processes, it provides the
backbone of products and services required to fulfill your businesses
strategy. Your processes dictate which technology to use and when,
and that technology will bring your digital transformation to life. For
example, the Dynamics 365 Sales app provides value-added details
about your organization’s lead-to-opportunity pipeline. Other exam-
ples of technology that’s under the hood to support users include:
▪ Infrastructure as a service (IaaS)
▪ Software as a service (SaaS)
▪ Integration tools
▪ Business intelligence tools
▪ Azure AI
▪ Azure Machine Learning
▪ Data connectors
▪ Customer portals
▪ Mobility solutions

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 119
 The technology and system selection are usually performed during the
sales process as part of the first high-level fit gap assessment.

Finding the right balance

Fig. 6-6

Identify potential It is easy to say that technology can be used to solve any business pro-
solutions using Fit-Gap cess scenario, but technology is simply what turns the gears for your
Analysis processes, people, and data. It’s important to include all stakeholders
in the design decisions, because processes don’t always align with the
technology. This is what we call the “gap” in a requirement, and it’s
where you might decide to purchase an Add-on solution or develop an
Conduct the Gap Solution
extension to meet the needs of the process (Figure 6-6). Think through
Design Workshop to
review proposed solutions the different options to find the right balance for your solution, but be
sure to challenge the business requirements before customizing the
system. You might be surprised to find out that what seemed like a
“must-have” customization is more of a “nice-to-have” feature.

After the fit-gap analysis is completed, the gap solution design workshop
For more on this topic, refer to Chapter 15, helps to discuss and further assess the identified potential solutions to
“Extend your solution.”
meet the requirements. This workshop provides structure to help analyze
For more information about the Success the gap solutions in terms of risk, ease of use, long term support
by Design framework, refer to Chapter 2,
“Success by Design overview.”
considerations, product roadmap alignment, etc. with the purpose of
selecting the solution that best fits the process and requirements.

Methodologies
A methodology is a collection of methods used to achieve predictable
outcomes. Good methodology also demonstrates why things need
to be done in a particular order and fashion. The Success by Design
framework is a higher-level abstraction through which a range of
concepts, models, techniques, and methodologies can be clarified.
It can bend around any methodology, including The Open Group
Architecture Framework (TOGAF), the Microsoft Solutions Framework
(MSF), and the Zachman Framework.

To achieve your solution, the workflows and delivery need to be

planned, organized, communicated by various methods, including:

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 120
 ▪ Project management
▪ Change management
▪ Governance and control
For more detailed information about
project management approaches, review
Chapter 5, “Implementation strategy.”
Project management
There are multiple project management methodologies and approaches,
and methodology for enterprise resource planning (ERP) is a topic that
can generate more heat than light. Partners often rely on their own brand-
ed methodology to reassure customers about their governance. Project
management approaches can be grouped into three core categories:
▪ Waterfall
▪ Agile
▪ Hybrid

As we move increasingly toward a low-code world, there are faster

deployments and fewer customizations, and your approach should
match the changing world and your customers’ expectations.

Change management
Project management (the “how”) and change management (the “who”)
are both tools that support your project’s benefits realization. Without
a change management plan in place, your organization’s objectives
are at risk. To drive adoption with your end users and accelerate value
realization, apply a vigorous change management approach, which is
most effective when launched at the beginning of a project and inte-
grated into your project activities.

The benefits of change management:

“When you have ▪ It focuses on the people side of organizational change.

▪ It seeks individual and organizational perspectives.
bad governance, ▪ It requires action and involvement by leaders and managers
resources are throughout the organization.
destroyed.”
– Wangarĩ Muta Maathai, Governance and control
Nobel Peace Prize winner The term “governance” often brings to mind just project management

121
 or possibly even stakeholder management, but such a limited view of
governance may cause an organization to ignore core elements re-
Fig.
quired for success (Figure 6-7).
6-7

Governance domains
Project Stakeholder Solution Risk and ​issues Organizational change
management engagement management management Change control and communication

▪ Weekly status ▪ Executive steering ▪ Oversight of ▪ Objective risk ▪ Identification and ▪ Intentional and
reports​ committee​ functional and assessment and prioritization of customized
nonfunctional mitigation planning​ requirements and communication to
▪ Project plan and ▪ Communication
attributes solution attributes​ stakeholders​
schedule​ and promotion​ ▪ Issue identification,
▪ Data, integration, ownership, and ▪ Tracking and ▪ Periodic
▪ Resource ▪ Escalation of
infrastructure, and resolution reporting via ADO effectiveness check​
management​ unresolved issues​
performance
▪ Prioritization and ▪ Organization and ▪ Solution adoption
▪ Financial ▪ Solution decision
▪ Customizations and severity assessment​ process changes​ and operational
management​ making and
process adaptation​ transformation
collaboration​ ▪ Risk response
▪ Risk management​ management​
▪ Security and management​
▪ Business alignment
▪ Change compliance​
and tradeoffs​
management

▪ Project status ▪ Project change ▪ Architecture board​ ▪ Project status ▪ Statement of work​ ▪ Communication plan​
meetings and reports​ management​ meetings and report​
▪ Governance ▪ Project artifacts​ ▪ Stakeholder
▪ ADO (project tools)​ ▪ Solution framework ▪ ADO (project tools) engagement​
▪ Project status report​
management​
▪ Governance ▪ Organizational
▪ Governance
framework ▪ Organizational change management
framework
change management

Every project component contains an element of uncertainty and is

based upon assumptions made before details are known or fully un-
derstood. So, if we know this, why would we expect a project manager
to be solely responsible for project governance? The project manager
is of course accountable for the project outcome, but 360-degree
governance requires everyone, especially the architects and leads, to
play a role.

A 360-degree governance method directly affects how well the project

team performs and the ultimate result, and is executed on multiple levels:
▪ A governance board or steering committee has final responsibility
for meeting the project goals. It typically is comprised of represen-
tatives or stakeholders from each of the major participants.
▪ A project manager leads the team and administers project ele-
ments that include alignment, risk identification and mitigation,
escalation, and cross-Microsoft collaboration.

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 122
 ▪ All project leaders, including the project manager, solution ar-
chitect, and technical architect, closely communicate and align
toward the common goal of solution deployment.
▪ The technical architect leads and orchestrates the technical
architecture across all application components to ensure optimal
performance.
▪ The solution architect guides users toward the new vision and
transformation.
Read more in Chapter 5, “Implementation
▪ The change manager ensures all internal and external communica-
strategy,” and Chapter 8, “Project governance.”
tions are in place.

Conclusion
As we’ve explained in this chapter, solution architecture follows a sys-
tematic approach that identifies your desired solution and the building
blocks needed to construct it. Solution architecture design takes
your business vision and breaks it into logical sections that become a
blueprint for building your solution. Here is a list of the key steps for
solution architecture design:
▪ Define your vision.
▪ Create a business strategy.
▪ Outline your business processes.
▪ Determine how people are connected in and around your
organization.
▪ Know your data.
▪ Develop a solution strategy using technology and tools.
▪ Apply project management, change management, and gover-
nance and control methodologies.

Implementation Guide: Success by Design: Guide 6, Solution architecture design pillars 123
7 Guide
Process-focused
solution
Implementation Guide: Success by Design: Guide 7, Process-focused solution 124
 Introduction
Fast changing technologies and markets demand a
rapid rhythm of delivering products and services.
Business applications improve inefficient processes by automating, opti-
mizing, and standardizing. The result: A more scalable business solution.

An approach based on business process mapping and management

in an implementation project is fundamental if organizations are to
achieve their goals. When implementations use this approach, there is
In this chapter we cover the more consistency in project phases, communication flows better, and
following topics related to a
processes run better.
process-focused solution:

Start your implementation project

with business processes This chapter introduces the importance of a process-focused approach
when implementing Microsoft Dynamics 365 Business Applications.
Opportunity for optimization
The goal is to emphasize the benefits of employing business processes
Defining the scope of
as the primary framework for a project implementation cycle.
the implementation

Defining your requirements

Fit to standard and

fit gap analysis Start with business
Implementation lifecycle
connected to processes processes
A solution that helps you to
operate your business
Business processes are the main drivers to start defining the solution
implemented in your project.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 125

 The business language
Every organization has an internal language they use to describe
daily operations. This language is immersed in business processes
and is often based on terminology common to the industry. When
implementing business applications like Dynamics 365, this business
language is the best, most familiar way for the business to think about
their needs, the tools and technologies that they use for work. This
business language is not simply terminology. It includes the processes
used to conduct day-to-day transactions and activities. Curating and
utilizing these processes in the language of the business is important
to keep sight of the basic truths of business requirements. It also reduc-
es the risk of the implementation team getting lost in the jargon and
When discussions about the project and
business processes begin with third parties,
nomenclature of the technical tasks involved in the implementation.
it’s helpful for those who may not be
familiar with the organization’s business to
begin learning this language. When the project starts, putting the business process view of the
organization at the core pays dividends.

Start with a business process future vision

Creating a vision of how a business is transformed through technology
makes it is easier to focus on the activities needed to achieve that goal.
This vision needs to include the understanding of how the business
currently works and the state that it wants to reach after.

Successful implementations start with a clear understanding of an

organization’s business model. That includes understanding how the
organization creates value through its products and services, as well
as its relationships with customers and suppliers. There are multiple
business processes strategically aligned to make that business model
work. Chapter 4, “Drive App Value,” describes how to define a business
model and the importance of connecting the business model to busi-
ness processes. Business processes provide the baseline to draw the
roadmap of your solution for the digital transformation.

The processes derived from the business model provide insight on

what it takes to reach the desired future. Business processes define the
functional scope of the solution being implemented as illustrated in

Implementation Guide: Success by Design: Guide 7, Process-focused solution 126

 Figure 7-1. Any set of business processes being implemented needs
Fig.
7-1
Requirements
definitions before you can start working on requirements and other
Business
aspects of the solution.
process

When defining your business model, assess the state of your processes.
Some organizations may want to use new technology to optimize exist-
ing processes through automation. Others are moving to e-commerce
Solution
or outsourcing parts of the operation, which requires new business
processes. Still others might be moving to lean manufacturing.

Many organizations also seek to consolidate operations and in-

troduce standard processes across different parts of the business.
Business process model Understanding the state of the processes allows you to understand
how to reach your goals. The effort required to implement an existing
process using new technology is very different than what is required to
define new processes or gain consensus from stakeholders.

New business processes won’t always be the same across a business.

One division, for example, might decide to ship inventory directly to
customers rather than to a warehouse, while others continue to use a
warehouse. On the other hand, it may be that the model for employee
purchasing is the same, so a shared product catalog and ordering
Start your implementation process is implemented. In any case, business model analysis, process
project with business processes
reengineering, and standardization strategies should be considered
Opportunity for part of the project definition. That firmly anchors the processes to the
optimization business strategy and also helps generate process-based project goals.

Defining the

Opportunity for
scope of the
implementation

optimization
Defining your
requirements

Fit to standard and

fit gap analysis It is easier to find good opportunities to optimize your process using
business process mapping.
Implementation
lifecycle connected
to processes

A solution that helps you to

Mapping processes
operate your business Processes are the heart of every business. They describe how the

Implementation Guide: Success by Design: Guide 7, Process-focused solution 127

business operates. The fact that your products or services reach your
customer safely and on time is due to a series of tasks executed by dif-
ferent roles and different business units. Business mapping represents
these steps.

Baseline business processes are called “as-is” processes. Depending on

the organization, these may be legacy processes or they may be new
processes engineered as part of a new solution or business model.

Regardless of where an as-is process set comes from, it is critical that

they are described at the appropriate level. Keeping as-is processes at a
more abstract level helps drive the scenarios in scope and enables the
team to focus on the desired business outcomes. As-is processes that
are very detailed about how current implementations are in legacy
systems introduce a risk of reimplementing the legacy system.

It’s worth reminding ourselves that the objective of as-is business

process mapping is not to define step-by-step how the solution is
designed in the new solution.

The goal of mapping the as-is business process is to:

▪ Provide structure/skeleton for the definition of the scope of the
project.
▪ Express the project aims in the natural business language of processes.
▪ Enable a natural way to define the main end-to-end business
scenarios (process variants).
▪ Provide the best language to identify and highlight areas for pro-
cess innovation and improvements, as well as risks and constraints.
▪ Become the basis for mapping to the equivalent process design
within the new system.
▪ Create a workable definition of the project that the business can
directly interact with without going to the level of individually
written requirements.

The value of having business processes mapped early is that it allows

everyone involved to understand how the processes operate by using
them as a common business language. This helps both third parties
and the organization agree on processes across multiple departments

128
 and organization layers. It is common for senior managers to be sur-
prised by how processes actually work on the ground.It is essential that
the right team is involved in mapping business processes to the new
technology. This team should be guided by the business stakeholder
who owns the process in the business operation. This individual is
charged with making the right decisions about structure and investment
in the process to achieve their goals. The stakeholder is supported by
one or more subject matter experts (SME) who are familiar with how the
business process operates in the real world. These experts can provide
depth around the various scenarios under which a process is executed.
Their knowledge, coupled with a mindset open to new ways of work-
ing, helps drive the right conversation on process mapping.

As an example, a company is looking to move to a B2B business model

via e-commerce and seeks to deploy their telephone salespeople
strictly to consumers. There may not be existing patterns within the
business to replicate, so the new processes should be based on ex-
pected outcomes rather copying legacy steps. This mindset should be
applied to all processes, as the key deliverables should be the expected
“business scenarios” and the desired business outcomes.

When reviewing new process diagrams, the team should consider the
following:
▪ Do the business SMEs recognize their business in the processes?
▪ Do they provide adequate coverage of the business activities in scope?
▪ Do they show the key interconnections between the processes?
▪ Do they show meaningful end-to-end business flows?
▪ Will they represent a good level at which the project team can
communicate with business stakeholders on risks, progress, decisions,
and actions?
▪ Do they illustrate processes sufficiently to identify opportunities
for improvement?
▪ Are they sufficiently well-defined to be able to be used in the next
stage where the processes can be applied to drive the design in
the Dynamics 365 system such that they can reflect the improve-
ments and the project goals and vision?

These process diagrams help define the baseline for the organization’s

Implementation Guide: Success by Design: Guide 7, Process-focused solution 129

 current business and are a good place to start when plotting optimiza-
tion objectives. They should help to identify the processes or parts of
processes that are not efficient or well understood.

The key to getting the processes mapped well and quickly is to ensure
the following:
▪ The right people are in the room to direct the mapping.
▪ Any mapping software or tools that help facilitate the rapid defini-
tion of the process and do not slow things down.
▪ The level to which the process is described is directly proportional
to its importance to the business. For example, once the end-to-
end scope of a widely used and very standard finance process is
defined to a certain level it should not be further broken down to
step-by-step processes.
▪ The process mapping effort is interactive, not a series of docu-
ments that go through long handoffs and approvals.
▪ The tools for the mapping can be anything from sticky notes, to
Microsoft Visio, to specialized process drawing software. The process
maps are visual and not hidden in wordy Microsoft Excel lists.

This is a vital step that needs to be executed early. Plan so that processes
are defined as fast as possible using the simplest tools in the workshop,
like sticky notes, pens, and whiteboards. The drawings and comments
Mapping business process is a not can then be transferred offline to more sophisticated applications.
one-time effort that occurs only during
the initial implementation. Rather,
it should be a continuous focus to
Remember that doing this mapping at the start of the project imple-
ensure the processes are aligned to
the evolving needs of a business and mentation is important. It could become a fundamental part of helping
are making the best use of the latest
capabilities in the product.
the analysis of the opportunities for process improvement and creating
the right baseline for the project.

Modeling your solution for the future

Digital transformation is fast paced. Processes that work today may not
remain competitive tomorrow. Technologies like the cloud, artificial
intelligence, machine learning, robotic process automation, virtual
reality, the Internet of Things, big data, virtual conference rooms, and
many others create possibilities that did not exist a few years ago.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 130

 When reviewing your business processes, it is important to keep in mind
how innovative technologies can improve process outcomes and collab-
oration. Similarly, there may be multiple areas in your processes which
For example, the Microsoft
HoloLens is being introduced to would benefit from modern options now available. Technology is a key
some manufacturing processes with
differentiator. Remember, chances are that your competitors may well
Dynamics 365 applications. Using the
HoloLens, the user can understand how be considering these options and looking to incorporate them.
to operate a machine. Or if a machine
is broken, and support is on the other
side of the world, virtual reality enables
collaboration between the user and
support to repair the machine.
Standardize your features with a business
Or if you have a workflow with
processes approach
approvals that require a signature, you Earlier, we discussed how a business model analysis can introduce
could use digital signature technology
opportunities for process reengineering and standardization.
to drive process efficiency and reduce
the time for approvals.

Assuming there is a good plan for process standardization grounded

in the business strategy, this standardization can be aided with a new
business applications system.

Of course, standardization across business units assumes that the

business models are sufficiently similar and there are no local market
constraints or competitive reasons for there to be differences.

Various benefits can come from standardized processes assuming the

business models are sufficiently similar between the business units so
that the other barriers to standardization are low.
▪ Enables reduction in cost by reducing the variations of business
processes that need to be implemented, trained, and maintained.
▪ Facilitates comparative analysis across multiple groups of users
that are executing the same process.
▪ Facilitates mobility of resources within the organization through
standard processes to deal with changes in demand.
▪ Promotes the adoption of best practices across the organizations.
Optimization for anyone improves outcomes for everyone.

Standardization also helps with rolling out the implementation to other

business units as it reduces the time and effort compared to building a
process from scratch. After the first implementation, there is a proven case
of it working within the business. The next business unit to be implemented
can discuss the pros and cons of the standard process with business peers.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 131

 Similarly, it is important to confirm that the constraints and implica-
tions are identified and well understood prior to starting design work.
▪ Standardization of a process can bring benefits at a group level,
but may be seen as additional work by a local business unit.
▪ Standardization can require significant change management.
▪ Is the standardization being defined at an industry level and then
implemented in the system? This can create a conflict if standard
processes are not a natural fit with Dynamics.
▪ Will the process be designed in Dynamics 365? And will these
processes become the new standard?
▪ Is the entire end-to-end process to be standardized or will there be
areas and functions for local variations? If so, does that align with the
standard capabilities in Dynamics, or could it need customizations?
▪ Does the business case for standardization remain sufficiently
strong after all the constraints and implications are balanced
against the benefits?

In multiple, phased rollouts, process standardization often takes the

shape of a “core template” with the expectation of rolling out this
“template” to subsequent business units.

Creating process maps is essential to identifying the opportunities and

to realizing the potential benefits outlined previously. The use of the
Start your implementation common business language of processes is the vital mechanism that
project with business processes
brings all of these aspects into focus.
Opportunity for optimization

Defining the
scope of the Defining the scope of the
implementation
Defining your requirements
implementation
A business application implementation is essentially the delivery of a
Fit to standard and
fit gap analysis new capability for end-to-end business transactions using the appli-
cation. Processes are the foundation for the definition of the solution
Implementation lifecycle scope. Processes embody many of the properties that make for a good
connected to processes
scope definition.
A solution that helps you to ▪ Business processes are well understood by the customer as they
operate your business are expressed in the languages they know best.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 132

 ▪ Processes are richer in context than a requirements list.
▪ Processes show the connections between different business units,
roles, functions, task-level steps, and systems.
▪ They are naturally hierarchical and can be expressed in different
levels of detail, depending on their intended purpose.
▪ Processes are the connectors that show the data and people
feedback loop.

Process mappings can be collected in a process catalog. This catalog

helps turn the visual process diagram into data. The business pro-
cesses in the catalog can then be numbered, labelled, and uniquely
referenced. Process hierarchy can be managed as part of the data
structures. Typically, these can then be managed in Azure DevOps (or
similar tools) where they can then be assigned to various tasks such as
configure, build, design, or test and to different project roles.

Business processes are usually interwoven with other business pro-

cesses. These interconnected processes can span systems, and the
connection can be a simple endpoint or an entire subprocess per-
formed in the external system. If the latter, it should be considered
for your solution scope. Where the process is in the external system is
directly relevant to the end-to-end process, include the full process in
the process mapping. This not only helps in understanding the process,
but also with making better decisions on customizations and selecting
ISVs, which are also stated in your solution scope.

Using process mapping as part of the definition of the scope also

helps when working with implementation partners. The process flows
helps express the ultimate outcomes for the project and having them
reflected the statement of work can help with a more comprehensive
definition of the project that results in fewer change orders.

Keep in mind that business processes are always subject to change. As

you move forward you are enriching the business process flows, and
possibly adjusting the solution scope. That is exactly the flexibility that
a business process flowchart offers.

133
 Fit to standard and fit
Start your implementation
project with business processes

Opportunity for optimization

gap analysis
Defining the scope of the
implementation In many projects, inertia drives the old ways of working. There are sev-
eral dangers in assuming that what works now also works just as well in
Fit to standard and a new Dynamics system. Recreating processes based on legacy systems
fit gap analysis
(often very old systems) can lead to the following:
Defining your requirements ▪ Unnecessary and excessive customizations and costs related to
design, coding, testing, training, documentation, etc.
Implementation ▪ Stagnation in the business as the natural opportunity for innova-
lifecycle connected
tion and improvements is lost
to processes
▪ Creating unique, siloed, inefficient processes when Dynamics 365
A solution that helps you to could provide more standard and recommended ways of working
operate your business
based on the learnings of thousands of customer organizations
▪ Damaging the natural usability engineered into Dynamics 365 by
forcing processes designed for legacy software into Dynamics
▪ Creating processes within Dynamics that were not designed to be
used in that way and therefore reducing the ability to use other
related functions because the processes are no longer a natural
flow or fit for the purpose
▪ Reducing the ability to immediately use the documentation, training,
sample code, and sample data available for the Dynamics system
▪ Reducing the ability to leverage market and industry knowledge,
insights, and experience in Dynamics 365 for the areas of interest
for the business
▪ Reducing the ability to directly apply tools and apps in the
marketplace
▪ Reduced capacity and resources available to apply customizations
that can provide significant business value as the available budget
and resources get used on customizations that do not add the
same value
▪ Creating barriers for end-user developers (citizen developers) to
create rapid, low-cost application using the Power Platform by
creating a more complex, customized data model/process model

Implementation Guide: Success by Design: Guide 7, Process-focused solution 134

 When it comes to understanding exactly what is going to be imple-
mented in your future solution, it is necessary to perform a deep
analysis about the current processes and what needs improvement. To
determine what functionalities of your business applications to keep,
and what others could be built or purchased from external vendors, we
suggest undergoing the following analysis. We suggest this because it
helps to have a standard solution, with minimal customizations, which
minimizes costs and maximizes value to the business.

Adapting to the standards of new system

It is recommended to start with a fit-to-standard approach for every
project. As part of the definition of the scope there should be a busi-
ness process catalog created in the early stages of the project. This
can now be used as the set of processes that can be configured and
enacted in Dynamics 365. Of course, initially, not all the configuration
is completed perfectly. However, starting with the core business pro-
cesses and iterating until there is a good, high-level understanding
of how the processes could be implemented within the system helps
create the solution blueprint. Putting the fit-to-standard approach at
the front of the analysis helps to set the right culture of “adopt wherev-
er possible, adapt only where justified.”

In most projects this analysis is not starting from a blank piece of

paper; the implementation partner has conducted some part of this
analysis as part of providing estimates on cost and time and a solution
overview. It is recommended that the customer project team get
deeply involved in this process to confirm that the key business pro-
cesses were correctly interpreted and to start gaining knowledge of the
standard processes embedded within Dynamics 365.

The continuation of the fit-to-standard leads to the definition of

requirements, fits, and gaps, but this is from the perspective of leverag-
Processes are the ing the processes enabled within the system first rather than trying to

foundation for the recreate the legacy system in Dynamics 365.

definition of the The advantages of starting with this approach using the process catalog
solution scope. are as follows.

135
 ▪ It promotes the reduction of customizations by supporting the de-
livery of the underlying business needs through configuration rather
than going directly into fit gap analysis with detailed requirements
that may be influenced by the existing or legacy system.
▪ It helps reduce the risk of missed requirements as the evaluation
of the fit with the new system is based on the richer and broader
context of business processes. As these business processes are the
natural language of business users, their evaluation is more com-
prehensive, meaningful, and effective compared to working with a
list of requirements.
▪ The process catalog can direct the fit-to-standard assessment by
working iteratively through the processes, starting with the high-
er-level core processes, and then working with the more detailed
sub processes and variants. This also helps the business users more
clearly see how well their underlying business requirements are
being met within the system.
▪ The project is more likely to adopt modern recommended stan-
dard processes embedded in the system.
▪ It creates higher-quality solutions as the processes are tested
by Microsoft and are more likely to be market-tested by others,
whereas custom developments and variants, especially complex
ones based on the legacy system, will need to be specifically
validated by the customer.
▪ The standard solution allows for more agility in adopting related
technologies and by keeping to standards where possible, make it
easier to add the real value-add custom extensions.
▪ It enables a faster delivery of the new solution; there is no need to
wait longer for a custom solution to be developed.
▪ Standard processes are more easily supported by internal and
external support teams, including Microsoft Dynamics Support.

The benefits of staying within the standard product wherever possible

are clear. The implementations following this approach, often called
vanilla implementations, adopt the Dynamics 365 system with its
standard configuration as shown in Figure 7-2.

Some businesses may have specialized business processes or an

innovative idea to improve their competitiveness in the market. In

Implementation Guide: Success by Design: Guide 7, Process-focused solution 136

 such cases, a process-centric description of the system allows these
improvements to be adopted with more transparency.
Fig.
7-2
Dynamics 365

Process Process Process Process

Start End
step 1 step 2 step 3 step 4

Gap analysis
As discussed in the previous section, adopting a process-centric
solution within Dynamics 365 has clear benefits. However, there may
be specialized functions that are not part of the standard solution as
shown in Figure 7-3. That is identified with the fit gap analysis. After
having the configurations set, you can look for gaps in the processes
and make a decision whether to customize.
Fig.
7-3 Dynamics 365

Process
Process Process Process End
Start step 3 with
step 1 step 2 step 4
Extension

Extending a solution can vary in terms of time, cost, and complexity.

There are many things to consider when extending. But it offers the
benefit of a solution tailored to business needs. See Chapter 15, “Extend
your solution,” for details on extending a solution. Here we look at the
implications of taking a process-centric approach.
▪ Some extensions can bring innovation into business processes. They
can add business value and provide good, long-term benefits.
▪ It is important to consider potential customizations in light of the
rapid pace of innovation in Dynamics 365 SaaS applications. New
features are released regularly and keeping an eye on the Dynamics
release plans avoids creating extensions that are quickly redundant.
▪ When evaluating a potential custom extension, using the process
maps can help determine the impact, not just on the process being
directly changed but also the impact on connected processes.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 137

 ▪ When looking at potential solutions for requirements that cannot
be met directly in the standard system, consider the different
methods available:
▫ For simpler designs, consider citizen developer-led applications
using the Power Platform, which can generate large value very
rapidly and at a low cost.
▫ For more complex or global solutions, some combination
of citizen and professional development through the Power
Platform or other technologies may be a more effective and
productive route. Using the process catalog can help pro-
vide an excellent backdrop to communicate and collaborate
between the project team, the citizen developer, and the
professional developer.

Third-party solutions
An alternative to extending is to buy an existing solution from a
third-party vendor, also known as an independent software vendor
(ISV). This option is more common when there is a significant gap
and developing a functionality in-house can be complex and costly.
Sometimes you don’t have enough resources, budget, or expertise to
develop and maintain such a significant solution.

ISV solutions tend to focus on more specialized process areas, normally

covering some end-to-end industry-specific process. In addition, ISVs
provide specific functionality to cover gaps, as illustrated in Figure 7-4.
Successful ISVs have expertise and experience in the vertical industry pro-
cesses they cover and can therefore add value to your business processes.

Fig.
7-4
Process 4 in
external app
or ISV
Dynamics 365

Process
Process Process Process End
Start step 3 with
step 1 step 2 step 5
Extension

Implementation Guide: Success by Design: Guide 7, Process-focused solution 138

 Dynamics 365 has been designed to meet standard business processes.
It enables projects to adopt the standard application more easily and
minimize the number of perceived gaps. It also has the flexibility to
customize and integrate external applications. Rather than choosing
one or the other, buying some third-party solution along with a stan-
dard implementation can build an optimal solution.

Representing the fits and gaps in a business process map in terms of

the processes drives a better understanding within the implementation
team and the organization. This is also needed to enlist the require-
ments for the design of the future solution.

Start your implementation

project with business processes Defining your
Opportunity for optimization
requirements
Defining the scope of the
implementation Some projects start with lists of requirements derived as a set of atomic
functional requirements, but without reference to mapped and docu-
Fit to standard and
mented business processes this can pose risks like the following:
fit gap analysis
▪ Late discovery of missing process steps
Defining your ▪ Misinterpretation of the requirement by parties not experts in the
requirements related process
▪ Lower business engagement in the project
Implementation
lifecycle connected ▪ Longer review cycles as the requirements and designs are less clear
to processes

A solution that helps you to Business process mapping helps draw the as-is processes to under-
operate your business stand how the business is running right now, and the to-be processes
to show how they work in the future. This also emphasizes the impor-
tance of business process mapping early in the project.

After determining what you keep or build, create the list of require-
ments, aligning them to the business processes for an accurate picture.
You can use tools equipped for the traceability of the requirements.

Typically, in a process-focused solution, once the processes have been

defined and agreed upon, the detailed requirements are derived and
relevant designs mapped. There are tools within Dynamics 365 to help

Implementation Guide: Success by Design: Guide 7, Process-focused solution 139

 with managing the requirements. For example, Azure DevOps can
work to create and track requirements, and if you are implementing
Dynamics 365 for Finance and Operations, you have the Business
Process Modeler (BPM) as part of your lifecycle services (LCS).

The list of requirements is easier to be crafted starting from the busi-

ness process mapping and it can be iteratively revised and refined.

Start your implementation

project with business processes Process-centric
Opportunity for optimization
implementation lifecycle
Defining the scope of the The value of taking a process-focused solution does not end with a
implementation
better definition of requirements. Adapting all the activities in the
subsequent phases to be more process-based helps deliver better
Fit to standard and
fit gap analysis outcomes, regardless of underlying methodology.

Defining your
requirements
At go-live, a business application such as Dynamics 365 sees system
users performing tasks and activities as part of a business process.
Process-centric The users are not executing on gaps, fits, or isolated requirements.
implementation It is important to keep this in mind when thinking of the full project
lifecycle lifecycle. The ultimate test of the project success is when the system
A solution that helps you to is in operation and the design build testing actions are intermediate
operate your business steps towards that goal. It is recommended to drive these phases with
the end-to-end business process as the framework upon which all the
related activities are planned, performed, and measured.

See the prospect to cash with Dual-Write integration to high-level

end-to-end process map in Figure 7-5.

Design
When creating the solution blueprint, the vision of the overall solution
architecture gives solidity and certainty to the project scope. This is the
transition from scope as business processes to scope as designs within
the system. As it is expressed in business language, it helps to ensure
that business users can be usefully engaged in the review and approval

Implementation Guide: Success by Design: Guide 7, Process-focused solution 140

 Fig.
7-5 Dynamics 365 Sales

Product and
Account Contact Quote Order Invoice
pricelist

Dual-write

Customer Contact person Product Quote Sales order Invoice

Finance and Operations

of the scope and proposed solution. There are, of course, other views
of the solution architecture in terms of data flow, systems landscape,
and integrations.

As part of the design phase in a Waterfall methodology, or during

design work in sprints in an Agile methodology, breaking down the
end-to-end business processes into meaningful subprocesses provides
manageable, bite sized units of work. It also allows better communica-
tion of the project tasks with business users.

The specific user stories, or configurations, and build work at a task

level can be related to the subprocesses so that they have the correct
business context. This makes it easier for business users to better un-
derstand the design and the business reviews and approvals of designs
are meaningful.

If the business processes are collected in a structured and hierarchical

process catalog, this catalog can be used to do the following.
▪ Help more easily and logically determine the sequence of work
starting with the more foundational process designs before tack-
ling the lower-level designs.
▪ Help highlight the interrelationships and dependencies between
the different business value streams, departments, roles, and

Implementation Guide: Success by Design: Guide 7, Process-focused solution 141

 functions. This helps with designing a more cohesive and integrated
overall solution.
▪ Help better manage the distribution of the actions and activities by
leveraging the sequence and interrelationship of the designs in scope.
▪ A process flow can directly be mapped to the system design from
configuration through data setup and functional run through.
▪ Help deliver working processes in Dynamics 365 software, by
implementing processes, including end-to-end process as early as
possible, and by taking a process-centric view.
▪ Help provide a better understanding of the solution by looking at
the current capability of the emerging system to execute end-to-
end processes.
▪ View the system in the context of a process, reducing unnecessary
customizations.
▪ Better engage the business subject matter experts (SMEs) and
business stakeholders by using the business language of processes.
▪ More rapid delivery of working software that better reflects and
makes better use of the Dynamics 365 SaaS cloud world.

Deliver processes in Dynamics 365 as early as possible, taking a pro-

cess-centric view. The benefits are practical and real, and often the only
real barrier to taking a process-centric view of design tends to be the
legacy of previous, more technical, and custom development-centric
approaches, better suited to the pre-cloud and SaaS world.

Having the high-level process, you can start breaking down the end-
to-end business processes into meaningful subprocesses, as shown in
Figure 7-6.

Development and configuration

A process-centric view provides an excellent basis for converting de-
signs into software. It can help directly address the question of how to
execute the business process within Dynamics 365. The configuration
and development tasks have the background of the business process
context, and do not need to be treated as an isolated delivery of a task
from a long list of such tasks.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 142

 Fig.
7-6

Dynamics 365 Sales

Product and
Account Contact Quote Order Invoice
pricelist

Dual-write

Customer Contact person Product Quote Sales order Invoice

Finance and Operations

Dynamics 365 Sales

Add products
Start Create order
to order
Price order

End Invoice order Pack order Pick order Confirm order

Finance and Operations

This isolation is still frequently seen in projects where the development

and configuration work becomes highly technical and is divorced from
the business process. This activity can become an end in itself, which
does not serve the project well. Instead, when a data- and process-cen-
tric view is taken, the configuration and development tasks can be
planned and managed in service of delivering a process embedded in
Dynamics 365 software.

A process-firstview allows the team to collaborate more productively.

The process designs, which now reflect the functions and configu-
ration and flow within the Dynamics system, help the various teams
communicate the context and better understand the delivery of any
development. Compared to working from a task in DevOps, or from a
paper design, the process background provides a better context and

Implementation Guide: Success by Design: Guide 7, Process-focused solution 143

 implications for a build task. For example, taking a process-centric view
can help with the mapping of the access necessary for performing
process tasks. That helps build a picture of the related security roles.

A process-focused view of the emerging solution also helps avoid

the dangers of a primarily “fit gap” approach which can create point
solutions. Instead, with a process-first background, there is a high
probability that the tasks and activities crystalize into a coherent and
well-functioning solution.

The project can also use the process-focused solution to conduct reviews
of the solution with the wider business to confirm that the project is
delivering on expectations. These reviews are much harder to do well
without the supporting process definition. That’s because without the
processes in focus, the project cannot speak the language of the busi-
ness. Many organizations are expecting some degree of business process
transformation as part of the implementation of Dynamics 365.

The reviews of the solution when the new processes are first explained
in business process terms helps ground the subsequent discussion
of the current state of the solution in clear, well understood business
language. This significantly improves the quality and usefulness of the
business review by minimizing the technical implementation jargon
and instead concentrating on the “business of business.”

Issues with resource constraints are well known. Showing incremental

progress of the build and development using business process language
improves the confidence of the wider business in the project. This can
be key for project success.

Testing
A project that has taken a process-centric view reaps the benefits during
testing. Other than the necessarily narrow focus of unit testing and some
non-functional tests, almost all other test types should be rooted in
some form of process definition. When there is an existing set of process
definitions and the associated designs, and a system with the designs
implemented, there are many advantages to the testing methods.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 144

 ▪ It helps ensure full testing coverage and enables earlier detection
of any missing or inadequately defined functions or processes.
▪ The testing can more easily follow the process-focused path al-
ready created earlier in the project.
▪ Testing has a natural focus on evaluating business process out-
comes which tends to be a closer match to the intention of the
design and eventual production use.
▪ It enables incremental testing of processes and subprocesses
which in turn helps engineer quality into the solution.
▪ Testing end-to-end processes, which is how the system is used in
production, is enabled.

Overall, testing in a business application like Dynamics 365 is a natural

match to a process-focused view of the solution. Having a process-cen-
tric approach throughout the lifecycle allows testing to become the
natural and obvious progression of the project delivery.

See Chapter 14, “Testing strategy,” for more details on the overall
approach to testing.

Training
Training to use business systems like Dynamics 365 applications is
fundamentally learning how to conduct day-to-day business processes
using the system. Most of the functional roles in the system interact
with multiple functions and processes. Rarely do roles exist in a vacuum
of their own process, but instead interact with other processes. Having
business process flows defined and available in the process catalog,
including flows across the seams of system roles, helps both in the
collation of process-based training materials and to guide the training.

During the design process, if the roles are mapped against the process-
es as part of security design, they can be directly used for testing and
for generating role-based training.

Even where some of the system roles may be very specifically restricted
to a specialized function in the system, there is often a need to under-
stand the upstream and downstream processes to perform a function

145
 well, and to understand any implications of any delays or changes in
the flow.

If whole project approach is process-focused, including the configure/

build activities, then the process of generating materials for training
(such as Training Guides for Finance and Operations and Guided Tasks for
Customer Engagement applications) goes smoothly, as there is a direct
correlation between the system processes and the training processes.

The process-based training not only helps prior to go-live, it also can
be used with new hires and when users change roles. It allows those
new to the business to absorb the business process simultaneously
while understanding how that process is performed within the system.

Roles can be easily defined in a business process and use the correct
security roles for testing and training as shown in Figure 7-7.
Fig.
7-7

Create order - E2E process

Dynamic 365 sales Finance and Operation apps

Add products
Sales manager Start Create order
to order
Price order

Sales clerk Confirm order

WHS worker Pick order Pack order

AR clerk End Invoice order

Support
The process catalog created in the project provides more than the
framework for implementation. It can also help with supporting the
solution. Functional issues raised to support usually need to be repro-
duced by the support team to enable the team to investigate the root
causes. A robust process-based definition of the flow allows support
to recreate the steps defined by the agreed standard process flow. The
visual depiction and description of the process allows the support team

Implementation Guide: Success by Design: Guide 7, Process-focused solution 146

 to speak a common language with the business user when discussing
the issues.

This ability to place yourself in the business process helps reduce the
number of back-and-forth cycles of communication and the overall
time taken to understand the issue in business terms. It increases the
confidence of the user that their reporting of the issue has been under-
stood. This reduces anxiety and improves user sentiment of the system.

A solution that helps you

Start your implementation
project with business processes

Opportunity for optimization

to operate your business
Defining the scope of the Designing and building a solution that has a great positive impact is
implementation
not an easy task. There are many aspects to consider, and everything
Fit to standard and is so connected that even missing a little piece can significantly affect
fit gap analysis the whole project. The business processes flows are the common
playground where we gather all the implementation artifacts together.
Defining your
requirements They keep all the solution components and phases aligned. They are
the line connecting the dots and the main instrument for communica-
Process-centric tion that helps to drive changes on how the business runs and helps to
implementation lifecycle
minimize missing those little pieces.
A solution that helps
you to operate your As you have read throughout the chapter, business processes are
business involved during all the journey of the implementation and they also
remain useful for future improvements in your business. They have
a perpetual impact in your business, as long as they are constantly
revised to align to the business goals and vision and changes are
reflected in the solution.

The creation and documentation of business process flows is to

prepare the fertile ground where the seed will be for all those imple-
mentation activities that allow you to harvest optimized and efficient
processes, risk depletion, cost minimization, role compliance, integral
solution, and, as an ultimate goal, a successful implementation.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 147

 Checklist

Business processes focus Ensure the business process definition is complete and
considers all activities and subprocesses.
Ensure the business process view of the organization is
Take advantage of the latest SaaS technology to drive
at the core of the definition of the project.
efficiency and effectiveness for the process optimization.
Clearly articulate the key business processes that are in
Ensure future readiness when mapping your business
scope and the respective personas so they are under-
process to the solution by incorporating configurability by
stood by all involved parties in the implementation.
design.
Ensure business model analysis, process engineering,
and standardization strategies are considered part of the
Fit gap analysis
project definition and deliver a strong process baseline
before implementation starts. Adopt a fit-to-standard approach and align to the
philosophy of adopting wherever possible and adapting
Collect the business processes in a structured and hierar-
only where justified.
chical process catalog during the requirements phase.

Process-centric solution
Use business processes for each phase of the project to
deliver better outcomes (all phase activities are better
planned, performed, and measured).

Opportunity for optimization

Explore opportunities to evolve, optimize, and consolidate
your processes as part of the implementation to meet the
digital transformation goals and drive user adoption.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 148

 Case study
The journey to a
process-focused solution
A large, multinational organization that provides contract-based
servicing with some manufacturing and distribution ser vices was plan-
ning to replace parts of its 15-year-old tier-one ERP and CRM systems
and other related legacy systems.

The wrong road

The project started with the implementation process that they had
traditionally applied to other IT projects. The traditional approach was
strictly Waterfall with a heavy documentation bias. The initial phase of
requirements gathering was conducted by using the old systems as the
reference point. A very long list of requirements was gathered, with the
assistance of an implementation partner, across the various functions.
The requirements gathering phase was extended a few times as the
requirements review took significantly longer than expected.

The reviews needed multiple parties to read, understand, and validate

the requirements and there was a concern among the business approv-
ers that if they missed any requirement, or even a nuance, their system
would not function well.

So, there were multiple document-based reviews where comments

from the approvers were formally submitted and the project team

Implementation Guide: Success by Design: Guide 7, Process-focused solution 149

 wrote responses to them, which in turn generated even more
comments from the approvers. The spiral of these reviews finally
concluded, but the business was not entirely convinced that they fully
understood what they had approved and hoped that the design phase
would provide more clarity.

The design phase was similarly based on writing and reviewing com-
plex design documents and was running late. The focus tended to be
on the “gaps” identified, and as these were not always placed in con-
text, the discussion with business users was not always productive and
was often at cross purposes. As further delays accrued and the business
users were becoming even less confident about the proposed solution,
the stakeholders decided to review the project direction and the rea-
sons for the continuous delays and general dissatisfaction.

The change of direction

As part of the review, which included third parties, several recommen-
dations were made. The primary recommendation was to adopt a
more process-focused approach and apply the processes as the frame-
work for a more Agile way of working. The first step in the process
was to define the end-to-end process at the highest level in order to
establish the boundaries of the project scope in business process terms.
The subsequent steps then generated two further levels of visualization
of the process detail for the business process streams. The project took
deliberate steps to try to define the processes in terms of the business
flow and desired business outcomes, rather than replicating the exist-
ing systems. This process mapping exercise was done at an accelerated
pace, working as combined business, project, and implementation
partner workstreams.

The right road

Once there was a reasonable map of the processes, the requirements
were mapped to the processes so that they could be understood in the
context of a process. As a result, many of the requirements were restat-
ed to be better anchored in the context of a business transaction and
thus less likely to be misinterpreted or removed for being redundant.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 150

 The processes at level two were put into a storyboard by using the
expertise and leadership of the partner solution architect and the
customer’s lead functional expert. The customer’s lead functional
expert used the process flows to create logical end-to-end process flows
across workstreams so the delivery of software would be meaningful.
The partner solution architect provided the Dynamics 365 application
view of the embedded standard processes, dependencies, and con-
straints to ensure that the sequence of process delivery would be
efficient within the Dynamics 365 applications. This storyboard would
drive the sequence of work from realizing foundational processes to
more peripheral ones. Furthermore, the level two end-to-end pro-
cesses (such as “Prospect to Cash”) were prioritized by considering
the core/most frequent/baseline path through the process as a higher
priority. The more specialized and less frequent variations were se-
quenced to be designed and delivered after the core processes.

This set of processes in the storyboard were then mapped to an overall

design within Dynamics 365 applications, generating a process-focused
solution blueprint. It was reviewed against the existing technical, system,
and data solution blueprint to create a rounded solution blueprint.

The delivery of the processes was distributed into sprints of four weeks
each, so each sprint delivered a meaningful part of the storyboard for each
of the workstreams, with an emphasis on delivery of end-to-end processes
as rapidly as possible. A high-level plan was constructed based on the
sequence, dependencies, and estimated effort related to process delivery.

The sprint level planning was performed on the processes in scope

for that sprint, defining a more detailed process flow when required.
Documentation was kept to a minimum and the processes were
designed in the Dynamics 365 system in collaborative workshop en-
vironments. All related activities such as data migration, integrations,
testing, training, change management, etc. were performed based on
the processes in scope. Each sprint culminated in a “conference room
pilot” (sometimes called “working demo” or “playback”), where the
business SMEs presented the new logical business process and how the
processes were designed and implemented in the system via a demo to
an invited business audience.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 151

 At each sprint end, the business attendees reviewed and approved the
incrementally complete and functioning processes in the Dynamics
365 system instead of reviewing and approving complex and technical
design documents. Individual gap designs that had previously been
circulating on paper for weeks and months were getting translated
into end-to-end working software. Business engagement significantly
increased as the project was talking their language and they were
working directly with the emerging Dynamics 365 system rather than
with abstract design documents and lists of requirements.

Arriving at the destination

The project further reinforced the process-centric project approach,
beyond design and build, by using the processes to script and drive
end-to-end testing, reporting status and progress as “ability to execute
a process within the system.” The senior business stakeholders on the
steering group also connected with the project more meaningfully as
they were finally able to understand the project readiness and business
operations implications.

The project successfully went live, and the customer continued to adopt
a process-centric view throughout the remainder of their go-lives in
other countries. The implementation partner decided to adopt this
process-centric approach as their new standard implementation approach
for their other projects because they could clearly see the benefits.

Implementation Guide: Success by Design: Guide 7, Process-focused solution 152

 8 Guide
Project
governance
Implementation Guide: Success by Design: Guide 8, Project governance 153
 Introduction
As the footprint of cloud-based solutions increases,
we need to work with and address different business
and technology challenges.
For example, Dynamics 365 applications are increasingly able to use
low-code/no-code technology. Also, customers have higher expecta-
tions on speed to value. The number of custom development-heavy,
multi-year, big-bang projects is diminishing; customers are looking for
early and incremental value, even on long-term, multi-country rollouts.
For a successful project Cloud-based Dynamics 365 software as a service (SaaS) applications
governance strategy, with regular, non-breaking system updates are changing the way we
consider these main areas: deliver business applications. Many of the limitations and sluggish
Project goals project governance processes of multi-year, on-premises implementa-
Project organization tions should not be carried into the cloud.

Project approach
We need to ensure that we’re creating a project governance model
Classic structures
that is fit for the world today and tomorrow. This includes considering
Key project areas new governance disciplines and reviewing the classic governance
Project plan model for effectiveness. We also need to consider that many partners,
system integrators, independent software vendors (ISVs), and custom-
We explore each of these areas in
more detail in this chapter. ers may have their own approach and governance standards that they
may have used previously.

Implementation Guide: Success by Design: Guide 8, Project governance 154

 Our objective in this chapter is to provide an overview of the impor-
tance of good project governance. We discuss recommended practices
on how to define the different areas of project governance and assess
the effectiveness of your governance. We also describe in more detail
some of the key areas of project governance and the specific consider-
ations that can help stakeholders, project leaders, and implementation
teams increase their chances of success.

The project governance topics discussed in this chapter are relevant to

any implementation methodology; we focus on the underlying prin-
ciples and provide guidance in a way that allows customers, partners,
and others to evaluate their own approach and adjust as necessary.

Objectives of project governance

Project governance should be designed to provide the framework and
techniques to do the following (Figure 8-1):
▪ Map business goals into actions and measures that deliver the goals
▪ Structure, plan, and reliably, safely, and efficiently drive the project
▪ Anticipate and avoid common issues
▪ Detect emerging risks and divergence from the planned path and
support corrective actions
▪ Provide the right amount of flexibility and agility to respond to
unexpected events and to adjust to the specific constraints of the
customer’s business or project
Fig. 8-1

When designing or reviewing

your project governance model,
you should examine how well it
Map business Structure, plan, Anticipate Detect risks and Provide
goals into actions and drive and avoid divergence from flexibility and addresses each criterion above.
and measures the project reliably common issues the planned path agility to respond
unexpected events Furthermore, it’s worth reexam-
ining these criteria in the specific
Project governance context of the goals and con-
straints of your current project.

In this chapter, we explore what project approaches, governance mod-

els, disciplines, and techniques project teams should consider when
preparing for and implementing a Dynamics 365 project. Similarly, we

Implementation Guide: Success by Design: Guide 8, Project governance 155

 use Microsoft’s Success by Design framework to highlight critical ques-
tions to consider throughout the project lifecycle to help drive success.

Why is project
governance important?
All system implementation projects, including Dynamics 365 applica-
tions, need good project governance to succeed. However, business
application projects have specific needs and challenges, and aren’t the
Fig.
8-2
easiest of projects to implement. Business applications directly impact
and are directly impacted by the processes and the people in the busi-
Typical ness. For a business application implementation to be successful, it’s not
sufficient to have good software and good technical skills; the project
governance also needs to have good governance processes that include the business.

issues A Dynamics 365 project needs to understand the business requirements

Unclear or ever-changing deeply and at a domain and industry level. It also needs significant and
project scope sustained participation from business users representing many different
roles, disciplines, and skills. Many, if not most business participants,
Late discovery of may not have previous experience implementing business application
project slippage
or business transformation projects, let alone Dynamics 365. This puts
an additional burden on ensuring that the methodology, approach,
Disputed areas of
accountability and governance models are sufficiently robust to support and drive the
Dynamics 365 business application project. It also requires business users
Low or sporadic user and to gain sufficient knowledge of the standard capabilities of the Dynamics
business engagement
365 application to better map their underlying business requirements to
the out-of-the-box solution.
Delays to go live at
a late stage
It’s perhaps worth reminding ourselves of the most common problems
Technical issues hiding related to implementing business applications (Figure 8-2):
underlying governance issues
▪ Unclear or ever-changing project scope
▪ Late discovery of project slippage
Mismatched expectations
between customer and partner ▪ Disputed areas of accountability or project responsibility
▪ Low or sporadic user and business engagement
Stakeholders blaming ▪ Delays to go live (often at a late stage)
each other
▪ Technical issues hiding underlying governance issues

Implementation Guide: Success by Design: Guide 8, Project governance 156

 ▪ Mismatched expectations between customer and partner
▪ Stakeholders blaming each other

The list is long, and you may have seen other issues, but most of these
issues result in project delays, quality issues, and impact on budgets.
However, the root cause of the issues tends to lie in gaps in the defi-
nition of the governance model, or in the effectiveness of operating
the project governance processes. Even after the project goes live and
meets most of the business requirements, if the project delivery isn’t
smooth, it can create stakeholder dissatisfaction and a lack of confi-
dence in the project.

In the context of the implementation lifecycle, you should define the

project governance model as early as possible, and certainly as part of
the Initiate phase of the project, because it directly influences the proj-
ect approach, oversight, resources, and project plan. During the Initiate
phase, if the customer or the partner has an underlying governance
model, you should review it and adapt it for this specific project, with
consideration for any specific business requirements and constraints.

Next, we explore the various areas that you should think about as part
of establishing your project governance model.

Project
governance
Project governance areas
Project governance is a wide topic and can encompass multiple dif-
Project goals ferent practices and disciplines. Various methodologies exist in the
market, including those that are customized by partners and cus-
Project organization
tomers. Irrespective of the methodology, you should consider some
Project approach fundamental principles when designing your governance, or when
reviewing and revising it.
Classic structures

The vast majority of projects have some form of governance; we’re

Key project areas
not describing setting up the common governance disciplines here.
Project plan Instead, based on our experience across thousands of projects, we look
at the areas where we often see the need for reinforcement and expan-
sion to drive a more reliable and successful project delivery experience:

Implementation Guide: Success by Design: Guide 8, Project governance 157

 ▪ Project goals  We look at whether the goals deliver solid foun-
dations for the implementation, are well defined, and provide the
necessary direction to the project
▪ Project organization  We explore how the right project orga-
nization and related roles and responsibilities help or hinder the
efficiency of the project
▪ Project approach  We examine the impact of the right method-
ology and the wider approach on project success
▪ Classic structures  We review how we can assess the effective-
ness of the most common governance processes to ensure they
don’t hide the real issues and provide false reassurance
▪ Key project areas  We look at how we can improve effectiveness
in areas of the project that get a lot of technical attention, but fre-
quently suffer from insufficient direction, planning, and oversight
▪ Project plan  We examine the critical importance of good proj-
ect planning and extract some of the planning-related lessons
from the previous sections

Project goals
Project goals
Project organization
Well-defined project goals are essential for steering a project and for
Project approach defining some of the key conditions of satisfaction for the stakehold-
ers. Often, the project goals are described in the project charter or as
Classic structures part of the project kickoff. In any case, it’s worth shining a spotlight on
them as part of the Initiate phase of the project and regularly through-
Key project areas
out the project. We recommend taking deliberate actions to reexamine
Project plan the goals in the context of your latest understanding of the project.

When reviewing or crafting project goals, consider the following:

Are the goals clear and realistic?

If your goals are overly ambitious or vague, the project may expend
unreasonable effort and time in the attempt to meet them, only to fall
short of the expectations set. Of course, no project intends to create
unclear or unrealistic project goals, yet there is a long history of such
examples, so we recommend conducting an honest review to confirm
these against the latest project scope.

Implementation Guide: Success by Design: Guide 8, Project governance 158

 Where possible, create objective goals, which are easier to mea-
sure, work with, and track. Some qualitative goals such as “the new
system should work at least as well as the old system” are open to
interpretation and can drive the wrong project behavior. They can be
misinterpreted and force the team to design customizations in the
new system to try and recreate the old system. Instead, consider the
goals in terms of the overall business outcomes you desire; this framing
allows the project to deliver them in a way that is most efficient in the
new system architecture.

Are the goals aligned with the business priorities?

If the goals aren’t well aligned with the priorities of the business stake-
holders and the business units in scope, the project won’t receive the
necessary business ownership, participation, or attention. Consider
whether the IT priorities and the business priorities expressed (or
implied) by the project goals are complementary. We also recommend
explicitly verifying that the priorities given by the organization are well
understood and aligned with those of the operating business units.
Sometimes group-led policies and requirements implied by the project
goals, which are expected to deliver additional controls and value at the
group level, may generate additional work at the operating unit level.

It’s essential to have all the stakeholders pull the project in the same di-
rection; conflicts at the level of misaligned goals are extremely hard for
the project team to solve. For example, if Finance leadership is looking
to improve compliance by adding more checks and approvals in a
purchase process, and Procurement leadership is looking for a faster,
It’s essential to have less bureaucratic and more streamlined purchasing process, unless the
all the stakeholders goals are balanced, the project delivery will falter. The end result will

pull the project in probably disappoint both stakeholders. Another common example is
when IT leadership has a goal of a single platform or instance for mul-
the same direction— tiple business units, but the business unit leadership has no goals to
conflicts at the level create common business processes. This mismatch can remain hidden
of misaligned goals and undermine the feasibility and efficiency of a single platform.

are extremely hard

Again, we recommend reviewing the goals with the stakeholders to
for the project team confirm that they will promote the right project delivery and business
to solve. outcomes. Successful projects have not only ownership of the project

159
 goals from the business, but also ownership of the successful delivery
of the project goals.

Are the project goals well understood by all the project members?
Some projects rely on a single kick-off meeting to communicate the
goals. However, many goals would benefit from more in-depth dis-
cussion (especially with project members from outside the business) to
better explain the underlying business reasons. Consider how you can
reinforce this communication not just during the initial induction of a
new project member, but also throughout the project lifecycle.

Have the project goals been correctly translated into project

deliverables?

Once a project starts Once a project starts the Implementation phase, the necessary at-
tention needed for the day-to-day management and delivery can
the Implementation sometimes overshadow the importance of the strategic project goals.
phase, the necessary This is one of the key findings from project post go-live reviews, the
attention needed original aims of the project faded into the background as the project

for the day-to- battled with the day-to-day challenges.

day management In many cases, projects are driven by the fit gap list, which takes a very
and delivery narrow view of the business expectations. Consider specifically reviewing
can sometimes the initial scope of the project and initial solution blueprint (and project

overshadow the plan) with the business stakeholders to assess how well the goals are
mapped to the project deliverables and take any corrective actions.
importance of the
strategic project goals. Is the governance process structured to formally monitor
progress against the goals?
This is an often-neglected area; sometimes projects only seriously
review the goals or success criteria as part of the final go-live assess-
ment, which is too late. Try to avoid this by defining specific structured
processes to ensure the assessment is ongoing. For example, look at
how the solution design blueprint helps meet the business process
objectives and how the data migration and reporting strategy meets
the information access objectives. Monitor to confirm that the project
priorities are aligned with the project goals and aren’t being diverted
by other considerations or day-to-day challenges.

Implementation Guide: Success by Design: Guide 8, Project governance 160

 Once a project starts the Implementation phase, the necessary at-
tention needed for the day-to-day management and delivery can
sometimes overshadow the importance of the strategic project goals.

Project goals
Project organization
Project organization Projects implementing business applications tend to have common
Project approach structures and roles (such as project manager, solution architect,
subject matter expert, functional consultant, and technical consultant)
Classic structures that are recognizable across different projects. The difference in the
effectiveness of project teams comes from the way in which the project
Key project areas
organization functions in practice compared to the theoretical con-
Project plan structs (Figure 8-3).

Fig.
8-3

Project organization

Project team aligned to business

Active and appropriate senior stakeholders

Structure promotes cross-team collaboration

Clear project leadership accountability and authority

Clear team roles and responsibilities

Roles aligned to solution and design requirements

Resources are proportional to effort

and complexity

When assessing your project organization, consider the following:

How well is the project team aligned to the business?

Teams that have good alignment between the business streams and
project functional workstreams tend to have more effective and

Implementation Guide: Success by Design: Guide 8, Project governance 161

 higher-velocity projects. A common model is for each key business
stream to be matched with a corresponding project workstream. An
experienced leader from the business stream is usually seconded to the
project and takes on the role of the lead subject matter expert (SME). For
larger projects, multiple SMEs may be appointed to a single workstream.

The direct involvement of a business stream with the project is usually

the most successful model because it engenders trust both ways, from
the business to the project and vice versa. If you have multiple orga-
nizational steps between the business and the project team SMEs, the
level of business ownership of the solution design diminishes, as well as
the quality of the input from the SMEs.

Does the project organization include active and appropriate

senior business stakeholders?
Projects are far more successful when senior business stakeholders have
clear roles and are active and deeply engaged in driving the project. A
good sign for the project is when the senior stakeholders are:
▪ Keenly interested in the status of critical processes
▪ Making a direct impact on the prioritization of tasks
▪ Making the corresponding resources available.

Projects in which the senior stakeholders are more passive and just
occasionally asking “How is it going?” or “Let me know if you need
something” tend to have poor outcomes.

Does the structure promote cross-team collaboration?

Examine if the project organization structure, reporting lines, team
leadership and composition, communication channels, team objec-
tives, and delivery approach naturally generate sufficient cross-team
collaboration, or if they naturally encourage silos.

Danger signs are project organization structures with the following:

▪ Team members performing the delivery work on the ground have
to navigate multiple layers of management to overcome issues
▪ Members of workstreams rarely work with other workstreams
▪ Responses to ad hoc questions on a team member’s task status
and issues are routed via the project manager; this also introduces

Implementation Guide: Success by Design: Guide 8, Project governance 162

 mistrust within the team
▪ Workstream communication channels are all hierarchical, with few
organic, cross-workstream objectives

A good way to determine the level of cross-stream collaboration is,

for example, to scrutinize how closely, and how often, the data mi-
gration workstream works with the various functional and technical
workstreams’ design. How closely entwined are the ongoing delivery
objectives of the data migration team and those of the functional
team? Do they primarily communicate hierarchically or across to their
teammates in other areas?

Is accountability and authority well defined at the project

leadership level?
Projects where the wrong role is accountable for delivery or where the
accountability is diffuse tend to have low velocity, with project deci-
sions stagnating. Projects where the accountability is given to someone
too junior in the organization to have the corresponding authority may
also languish without sufficient direction and resources.

How well are the team roles and responsibilities defined?

Every role on the project needs to be well defined, including roles that
may be part-time on the project. The critical role played by business
users that aren’t seconded to the project is often ill-defined. They
may typically be necessary for defining a detailed process or data
requirements, for approving business processes, or for preparing and
performing testing. These are critical activities, and the corresponding
roles should be well defined so the resources can be planned and their
tasks and status are visible to the project leadership.

How well are the team roles aligned to the solution complexity
and system design requirements?
You should conduct an honest analysis of the experience and ability
of the resources in key roles when compared to the complexity of the
design and its constraints. Be wary of job titles that don’t match the
experience and authority that normally accompany such titles.

163
 This is particularly important for the key roles of lead solution architect,
lead technical architect, and project manager.

Are the levels of resources proportional to the level of effort and

complexity?
By the end of the Initiate phase of the project, you should have a rea-
sonably credible high-level solution blueprint, a reasonable estimate
of the backlog and effort, and a high-level project plan. This should
provide a good grounding for reviewing the level of resources planned
for the key roles.

This honest review is especially recommended for the roles that are
likely to be the most constrained, so that mitigation plans can be
initiated in a timely manner.

During the implementation, you should regularly assess how the day-
to-day working of the control, communication, and feedback functions
are being helped or hindered by the project team organization (Figure
8-4). How well does the project organization structure facilitate or
constrict the undiluted and timely flow of direction, guidance, and
decisions from the business to the project workstreams? In the other
direction, does the structure enable or hinder the flow of accurate,
actionable, and timely feedback from the workstreams to the project
Fig. leadership and business stakeholders?
8-4

Accurate, honest, A typical setback on projects is

timely status: the leadership team not receiving
Progress, risks, timely and accurate feedback on
opportunities, blockers
project issues, therefore adding
unnecessary cost and delays. This
Workstreams is often due to structural issues
created by a project organization
Project
that doesn’t help the flow of mean-
leadership
ingful information. Projects with a
Based on a close lot of hierarchical communication
understanding of
and reporting-line hoops to clear
the project:
(especially on large projects) tend
Goals, direction,
decisions, resources to have difficulties in surfacing

Implementation Guide: Success by Design: Guide 8, Project governance 164

 accurate data promptly. This can result in project challenges stagnating
for too long, which leads to delays and additional cost.

Many projects start with the ambition of some model of partnership

and shared leadership with the various parties involved. The projects
that better achieve this ambition are the ones that meaningfully map
the commercial and partnership arrangements (or expectations if
informally agreed) and their respective expertise to the project orga-
nization. Projects tend to have better success when the customer has
a clear leadership (and ownership) role in the organization structure
driving the overall project. Similarly, unless the customer has experi-
ence and capability in implementing business applications, having the
implementation partner in a leadership (and accountability) role for
providing strategic guidance on the implementation methodologies,
system design, and technical procedures allows for the best mapping
of leadership roles to those with the best experience. Of course, de-
pending on the specific expertise between the customer and partners,
the actual balance of responsibility lies somewhere in the spectrum of
these arrangements, and the organization structure should be mapped
to that agreement.

In business application projects, some critical project roles have a

disproportionate impact on the success of a business application
project. For most projects, the role of the solution architects and the
project manager is particularly critical to success. The project manager
and solution architect roles are important both from the implementa-
tion partner and from the customer; the customer solution architect
is usually described as the lead business SME or the business lead.
Depending on which areas of your project are particularly significant or
risky, you may want to add other roles to the list of the critical project
roles. You should then pay extra attention to ensure that the individ-
uals chosen for these critical roles have the right level of experience
and capability. For example, an ineffective solution architect may not
be able to direct the solution design blueprint such that the different
elements of the solution design come together into a coherent and
efficient solution that matches the business needs.

You should also confirm that the project organization allows all roles to

Implementation Guide: Success by Design: Guide 8, Project governance 165

 be successful by enabling them to properly assert their expertise and
strongly influence the design and implementation, thereby delivering
the full positive impact of their roles.

Project goals
Project approach
Project organization
When talking about project approach, one of the dangers is that it
Project approach can be assumed to be synonymous with the project implementation
methodology. This can then leave a vacuum in the processes that need
Classic structures
to be defined outside of the implementation methodology. This can
be especially true if the implementation partner is providing a limited,
Key project areas
technical implementation methodology. If you’re the customer, you
Project plan should consider the wider set of processes required to define your
project scope, manage your resources, and manage the changes,
tasks, and processes. Prioritize areas that aren’t directly covered by the
partner but are necessary for every business to perform in support of
a business application implementation. You should explicitly identify
areas that would be the responsibility of the customer to manage.
Then confirm that you have adequate definition of the approach and
the right level of governance planned for each of these areas.

The areas excluded from the implementation partner’s responsibility

will vary, but the following areas are typically involved:
▪ Defining the scope of the project (the to-be business process and
system requirements)
▪ Managing internal project resources and liaising with wider
business teams
▪ Managing the process for the proper participation of all the rele-
vant business units and roles
▪ Managing the source data quality for data migration
▪ Training the internal project team and the end users
▪ Data, system, and process validation by the non-project business
users, such as user acceptance testing (UAT)
▪ Security and access definition and validation
▪ Implications on the wider IT enterprise architecture and wider
business processes

Implementation Guide: Success by Design: Guide 8, Project governance 166

 ▪ Interpreting, applying, and validating the requirements related to
relevant regulatory bodies
▪ Cutover management
▪ Managing the non-Dynamics 365 systems involved in system
integrations
▪ Managing communication within the business and with customers
and suppliers
▪ Budget management
▪ Business and organizational change management
▪ Non-functional requirements such as performance management,
system security, and business continuity planning
▪ Transitioning and operating the support model and other post-
operational duties

This is not intended as an exhaustive list, more of an indicator that

the overall project approach needs to consider a much wider set of
functions than what may be covered by a typical implementation
methodology or a commercial contract. Some projects define the
overall approach in a project charter or similar document. You should
examine if the project approach adequately covers the missing areas
from the implementation methodology.

One of the discussions that regularly occurs about project approach is

the debate on waterfall versus agile. Methodologies based on either
of these principles have their advantages and disadvantages. For
more detailed discussion on implementation methodologies, refer to
“Chapter 5, “Implementation strategy.” From a project approach per-
spective, whichever methodology (or hybrid) is adopted, it’s important
to identify and record the specific risks that logically emerge from that
methodology and have a mitigation plan for each one.

For example, if your chosen methodology doesn’t include early anal-

ysis and defining a solution blueprint that reflects the whole design
scope, the project may stumble from sprint to sprint, designing and
re-designing the solution. You may find that the designs in later sprints
can’t be built upon the previous sprints due to the constraints of not
having analyzed and considered the high-level design for the whole
end-to-end process. We often see this in pure agile projects where

167
there is no overview solution design blueprint.

Similarly, if the chosen methodology is strictly linear and mostly doc-

ument-based, you risk spending an enormous amount of time and
effort in analysis, design, and coding phases with little exposure to the
working solution and limited feedback on how the solution performs in
the system. This is a risk we see in many pure waterfall projects. You need
to identify such risks in your chosen methodology so you can discuss, ac-
cept, understand, and specifically address them in the project approach.

Sometimes when a methodology is being selected, there is an as-

sumption that a methodology that works well for a bespoke software
development project will apply with minimum changes to a highly
configurable Dynamics 365 cloud-based business application im-
plementation. Similarly, methodologies that suited the old world of
on-premises business application implementation with long implemen-
tation durations aren’t best adapted to today’s more rapidly evolving
cloud world. Consider the very different circumstances of a modern,
cloud-based Dynamics 365 business application implementation:
▪ The implementation is for a cloud-based, packaged SaaS software
application
▪ A significant (if not a majority) of the activities aren’t coding, but
business process design, setup, configuration, data migration,
validation, and more
▪ Some critical activities for a Dynamics 365 application implemen-
tation (such as extracting a fit gap report from a fit-to-standard
analysis) aren’t covered by more generic software development
implementation methodologies
▪ There is ever-increasing functionality in the Dynamics 365 busi-
ness application of powerful configuration options or low-code/
no-code options and the ability to use the Dynamics 365 Power
Platform in citizen-developer and professional developer modes
▪ The methodology must address the fact that a very significant part
of the implementation process revolves around the understanding
and configuration of business processes and working closely with
business users throughout the project lifecycle
▪ Although Dynamics 365 applications provide a platform for cus-
tom development, the chosen methodology must address the fact

168
 that the development process is building on, or extending, the
standard business process flows designed in the system
▪ Dynamics 365 applications are a cloud-based, SaaS business appli-
cation with a regular update rhythm that needs to be recognized
as part of the new way of working

All of these factors (and more) mean that the implementation meth-
odology needs to be directly relevant to the nature and needs of a
Dynamics 365 business application project.

When looking to determine what the overall project approach should

address, consider the specifics of your whole project solution build life-
cycle and make any additions, adjustments, or highlights to the standard
methodology. Confirm that the key processes and controls you expect
are adequately addressed as part of the governance of the project:
▪ Does the methodology and approach enable a good definition of
the scope and a solution blueprint?
▪ Is the testing strategy fit for purpose?
▪ Do controls such as regular and appropriate reviews by the project
team and the business exist?
▪ Is a good process for project status analysis and reporting in place?

You may want to define a set of critical processes, deliverables, and

controls to ensure you have the right coverage from the project ap-
proach ( just as you would for functional and technical requirements).

Once you are in project implementation mode, and the focus is on

day-to-day tasks, it’s easy to lose sight of the ideals agreed in the gov-
ernance model at the start of the project. You should establish regular
checkpoints to go through the formal discipline of honestly comparing
the actual ways of working on the project with the governance model
and taking corrective actions as necessary.

For example, analyze a random sample of how a given feature, require-

ment, or user story is actually progressed through the lifecycle, or take
a scope change item and trace how it was handled to resolution. Did
they progress through their lifecycle as you expected, and did your
governance model help deliver a fast and good quality result? Taking a

Implementation Guide: Success by Design: Guide 8, Project governance 169

 (preferably independent) regular pulse of the real, practical implemen-
tation of the processes and controls helps keep the project approach
relevant and reduces risk.

Project goals
Classic structures
Project organization
Most, if not all, business application projects have the common, classic
Project approach governance structures in place. This section doesn’t give a general
overview of historically well-known disciplines around these common
Classic structures governance areas; instead we look at how to assess the true effective-
ness of these processes and controls in practice. The mere presence of
Key project areas
these classic governance structures such as steering groups, program
Project plan boards, or risk registers can sometimes deceive projects into thinking
that they have adequate active governance. Let’s dig deeper into these
areas to explore how we can get better insight into their function and
evaluate their effectiveness.

Steering groups
Most projects have some form of steering group in which the proj-
ect sponsor, senior business stakeholders, and project leaders meet
regularly to discuss and review the project. A lesson learned from
participating in multiple such steering groups across multiple projects
is that the effectiveness of these meetings varies hugely. Common
factors that can impact the effectiveness of a steering group are when
Fig.
8-5 the purpose of steering group
meetings is diluted or unclear,
and when project status reporting
isn’t easily understood, accurate,
or actionable (Figure 8-5).

Steering group meetings

Diluted purpose Clear purpose Steering group meetings can neg-
Little understanding of project details Understand primary function atively influence project success if
the project manager presents the
No direct engagement Time invested
status and the steering group is
Lack deep knowledge Comprehensive knowledge
simply checking in with very little

Implementation Guide: Success by Design: Guide 8, Project governance 170

 understanding of the details of the project, no direct engagement,
and a lack of knowledge to challenge and explore issues more deeply.
Steering groups that rely entirely on project status updates with only a
perfunctory amount of independent analysis or review are often sur-
prised by a sudden crisis, resulting in delays and dissatisfaction.

An efficient and effective steering group has leadership team partici-

pants that understand and perform their primary function of steering
the project. This requires them to invest the time to understand the con-
tent of the project at a sufficient depth so that they can read between
the lines of the project status report, ask the right questions to under-
stand how and where they can help, and proactively direct the project.

Apply similar, regular assessments of the effectiveness of other such

boards, like the change control board or the design authority board.

Project status reporting

Accurate, meaningful, and actionable project status identification
and reporting is the lifeblood of a project. Business application im-
plementations often involve complex, multiple strands of work, all of
which need to come together seamlessly and on time. Multiple parties
are often involved in the project delivery, business SMEs, business IT,
business project managers, and their equivalent roles from the im-
plementation partner, as well as other third parties. In addition to the
main functional and technical delivery, there are multiple other disci-
plines, like data migration and security. All these roles have their own
specialist tasks to perform, and managing and monitoring all these
different types of tasks to provide a coherent project status isn’t easy.

Additionally, specialist teams may provide an overly technical view of

the status in terms that only that team understands. Often, project-lev-
el status reporting is driven by fit gap progress, which isn’t necessarily
the critical part of the project. Consider the following when looking for
effective project status reporting to the steering group:

Is the project status meaningful and actionable for the

intended audience?
For example, when presenting internally to your own specialist team,

Implementation Guide: Success by Design: Guide 8, Project governance 171

 a detailed and technical view is very appropriate; however, when
presenting the status to a steering group or to non-project team mem-
bers, consider how to express the status in terms that are understood
and actionable.

Is there evidence to believe that the project status is accurate?

Consider how the project status data has been gathered, analyzed, sum-
marized, and presented. Understand the level of uncertainty associated
with the data. Is there a history of the project status data and related
estimates versus actuals that can provide some idea of certainty?
Steering groups have to consider the degree of uncertainty associated
with the status when assessing risks and making critical decisions on
moving or adding resources, approving budgets, and setting timelines.
This means testing the accuracy of the status by looking at it from multi-
ple angles to ensure that it provides robust and actionable data.

Is there a comparison between the planned and actual status?

Project status reports only have meaning when they show an easily
understood comparison between the planned status and the actual
progress, the report must be able to answer questions like “Are we on
track?” “What is the size of the remaining work to completion?” “What
do we, as the steering group, need to do to get the project on track
and keep it on track?” The absence of a comparison of planned work
to actual and remaining work doesn’t allow the steering group to take
necessary actions to help recover.

Risk register
Most projects have some form of a risk register. When used well, this
register is a good way to communicate risks and help teams focus their
attention on removing barriers to success. The following are examples
of ineffective use of risk registers:
▪ Risks that are of little practical value to the project are being used
to shrug responsibility or provide cover against blame.
▪ Risks remain on the register for a long time with just new com-
ments and updates being added weekly at each risk review
meeting, but with no resolution. This implies that either the risk
isn’t getting the attention it deserves or it’s difficult to resolve and

Implementation Guide: Success by Design: Guide 8, Project governance 172

 consuming more and more resources without results. In any case,
you should either treat risks stuck in this loop urgently with focus
or accept them with a mitigation so they don’t drain the project
over time.
▪ Risk priorities don’t reflect the project priority at that stage of
the project. You should take care to ensure that the risk register
doesn’t create a parallel project effort with its own priority and
path. Risks and issue resolution should be incorporated as part of
the main project delivery.
▪ The risk register has a very large number of risks, many of which
are stagnant. Consider how many risks the project can realistically
work on at any given time and trim the register according to real
project priority.

Stage gates
Stage gates or milestone-driven planning and reviews are a common
feature of the majority of business application projects, including more
agile projects. These milestones are regarded as important checkpoints
spread throughout the project timeline, which the project can only
pass through if they have met certain criteria. The reality of many
projects is that the checkpoints don’t always perform their intended
A milestone function. There may be several reasons for this, and projects should
should have the examine if they are suffering from the following limitations:

primary purpose ▪ Criteria for the milestone are unclear  You should strive to
create very explicit criteria for entering and exiting a milestone. If
of confirming the criteria are difficult to measure, it’s difficult to take unequivocal
that the project decisions based on data.
has satisfactorily ▪ Exit and entry criteria aren’t respected  Projects are typically

completed the under great pressure to get past one milestone and move into the
next phase, either because of resource usability or commercial
precursor tasks payment triggers, or because the project optimistically believes
so that it’s safe, that it will somehow catch up with the overdue tasks that didn’t
efficient, and meet the milestone criteria. Sometimes a project is also pressured
to move to the next phase because it implies progress. This self-de-
meaningful to start lusion can create additional risks in the project; uncontrolled
the tasks of the disturbances in project deliverables can delay the ultimate goal of
next stage. a successful go live. The debt incurred by skipping the criteria has

173
 to be paid back at some point; the later it’s resolved, the higher
the cost. When considering allowing a milestone to pass when it
hasn’t met all its criteria, project leadership should strongly consid-
er the nature and size of the incomplete tasks, precisely how these
tasks will be addressed, what impact the lagging tasks will have on
dependent tasks. This allows for a managed transition between
milestones with the risks fully understood and a clear and realistic
action plan for the stragglers.
▪ Milestones don’t reflect true dependency  From a project
lifecycle perspective, a milestone should have the primary purpose
of confirming that the project has satisfactorily completed the
precursor tasks so that it’s safe, efficient, and meaningful to start
the tasks of the next stage. There may be other perspectives, such
as commercial ones to trigger stage payments or just periodic
points for the project to review progress or cost burndown, but
we’re focusing on a project lifecycle point of view. Some common
examples of stage gate criteria are as follows:
▫ To-be business processes defined  An approval step from
the business to confirm the business process scope helps
ensure that prior to spending a lot of effort on detailed
requirements and design, the process scope has been well
established and agreed upon.
▫ Solution blueprint defined  This helps ensure that the
requirements have been analyzed and understood sufficiently
well to be able to define an overall, high-level design that is
confirmed as feasible. Additionally, the key design interrela-
tionships and dependencies are established before working on
the individual detailed designs.
▫ Formal user acceptance testing start agreed  Starting
formal UAT with business users tends to be the final, full formal
test phase before go live, with the expectation that go live is
imminent and achievable. Prior to starting this phase, it makes
sense to validate the readiness of all the elements to ensure
that this test phase can complete within the allocated time
period and meet the necessary quality bar.

Consider the various milestones that help with a true understanding

of readiness, especially those that help you better understand when to

Implementation Guide: Success by Design: Guide 8, Project governance 174

 start dependent tasks, such as feature complete, code complete, code
frozen, business data validated in Dynamics 365, or solution ready for
user acceptance testing, and determine what entry and exit criteria
would help.

Design and change boards

A project with multiple parties involved in the design and delivery tend
to have some form of design review board that ensures compliance
with the enterprise IT policies and standards. This board also ensures
that the proposed designs from the project (and especially design
changes) are fit for purpose.

Effective design review boards (also called architecture review boards

or design change boards) tend to have a good grasp of the overall
solution blueprint. In addition to reviewing any individual design, they
can ensure new designs and changes will operate within the boundar-
ies of the solution blueprint and won’t adversely affect other designs.

The design review board should also have representatives, not just
from the project, but also from the wider business and IT, to help
ensure the business case for the design is sound and the impact on
related systems (integrations) and overall enterprise architecture are
considered. Another characteristic of a successful design review board
is that they have minimum bureaucracy and their focus is on helping
the project to move as securely and as fast as possible.

The process by which a design review board actually functions is as

important as its constituents and roles and responsibilities. Design
review boards should communicate the solution blueprint boundaries
and design standards expected so projects can proactively make sure
that they’re working within the expectations.

For the best project velocity, you should set an expectation that the
reviews will be interactive and expected to be resolved within a single
iteration. This requires the design review board to be pragmatic and
well prepared. It also requires the project to have the discipline to
provide sufficiently detailed proposed design information. This process

Implementation Guide: Success by Design: Guide 8, Project governance 175

 needs to be organized and communicated as part of the project
governance in the Initiate phase so you’re not trying to establish
these practices during the intensive pressures of a running project
implementation.

Project goals
Key project areas
Project organization
Project governance is often thought to be confined to the classic areas
Project approach we discussed in the previous section, but in Dynamics 365 business ap-
plications, we should consider the key activities that also need to have
Classic structures
governance processes embedded. We see many projects where the
strategies for the more specialist project delivery areas have their own
Key project areas
processes and disciplines that don’t always work in concert with the
Project plan project approach or to the benefit of the overall project. These areas
are often managed and driven by specialists with their own processes
and internal goals.

It’s worth considering what governance is appropriate for these areas

to ensure the project (and hence the business) derives the best value.
Fig.
8-6 In this section, we discuss the
key areas of application lifecycle
management (ALM), test strate-
gies, data migration, integration,
Key project areas cutover, and training strategies
(Figure 8-6).
Application lifecycle management

Test strategy Application lifecycle

Data migration management
Does the defined application
Integration
lifecycle match the project meth-
Cutover odology? For example, how will
Training strategy the configuration data be ana-
lyzed, defined, approved, secured,
managed, and distributed to the
different environments? Is this

Implementation Guide: Success by Design: Guide 8, Project governance 176

 process also governed by the requirements of the project approach,
and will it be designed to meet the expectations of the project timeta-
ble? The risks here are that ALM processes are sometimes considered
as a technical topic and not rolled into the overall project governance,
which creates mismatched delivery and project planning.

The build process (configuration, data, code) in a business application

can be presented as a complex, technical, and jargon-filled process.
Business stakeholders may retreat and let specialists take complete
oversight. However, project leadership can get meaningful insight into
the (necessarily technical) build process by devising a layer of gover-
nance that provides the necessary understanding and helps keep the
project on track. For example, ask the following questions:
▪ Are the build requirements directly mapped onto the end-to-end
business processes, so that any discussions related to the require-
ment can be well understood in that context by non-technical
stakeholders?
▪ Are the right controls in place so that progress can be tracked and
reported in a way that is meaningful outside of the technical teams?
▪ Do controls exist to ensure the right quality, standards, and com-
pliance needs will be applied?
▪ Does the project have a practical and efficient method to raise busi-
ness process questions with the business area owners, and a way for
the business to understand the system delivery of a business process?

A more in-depth exploration of the ALM and build process is available

in Chapter 11, “Application lifecycle management.”

Test strategy
A well-defined test strategy is a key enabler for project success. A high
level of governance is required to ensure that the right test strategy is
created and implemented, because multiple cross-workstream parties,
business departments, IT departments, and disciplines are involved.
When evaluating the suitability of the test strategy, in addition to the
technical angles, consider how well some governance areas are covered:
▪ Does the right level of governance exist to ensure that the testing
strategy is mapped to, and proportionate with, the project scope

Implementation Guide: Success by Design: Guide 8, Project governance 177

 and the business risks?
▪ Is the business sufficiently involved in the test coverage, test case
development, and approval of the testing outcomes?
▪ Is the project approach in line with the test strategy and test planning?
▪ Does the right governance exist to make sure the business pro-
cesses are tested end to end?
▪ Is the data to be used in the various testing phases and test types
sufficiently representative to fulfill the test objectives?
▪ Is the test coverage across the functional and non-functional
scope adequate to help assure the business of safe and efficient
operation in production use?

More details on defining a successful test strategy are in Chapter 14,

“Testing strategy.”

Data migration
For data migration, examine the related governance coverage to
ensure that this process is well understood at a business level:
▪ Do you have the right level of business ownership and oversight
on the types and quality of data being selected for migration?
▪ Is a data stewardship process in place to make sure that the
cleansed and migrated data is kept clean during the project and in
operational use?
▪ Will the data from the existing source systems be faithfully
transformed so that it’s still meaningful and fit for purpose in
Dynamics 365?
▪ Will the data strategy and plan provide the right type, quan-
tity, and quality of data to the project at the right time for key
milestones such as testing, training, mock cutovers, and other
data-dependent tasks throughout the project lifecycle?
▪ Does the data migration strategy adequately reflect the go live cu-
tover strategy, so the mock cutovers or migration dry runs closely
reflect the process for data migration at go live?

Review the data migration strategy from a governance view to ensure

the workstream is delivering to the needs of the project and the busi-
ness and isn’t becoming a silo with only technical oversight.

Implementation Guide: Success by Design: Guide 8, Project governance 178

 Integration
A common area missing governance is the management and owner-
ship of non-Dynamics 365 systems involved in the integrations, partly
because they tend to be managed by those outside of the project.

Make sure that the business impact on either side of the integration is
understood and managed. Confirm that the data exchange contracts
are defined with the business needs in mind. Examine if the security
and technology implications for the non-Dynamics 365 systems are
properly accounted for.

You also need project-level governance on the definition and manage-

ment of non-functional requirements, such as expected volumes and
performance, and clear processes to resolve issues that may stem from
the other systems in the integration.

Cutover
The cutover from the previous system to the new Dynamics 365 system
is a time-critical and multi-faceted process. It requires coordination
from multiple teams for the related tasks to all come together for a go
live. You almost certainly need to include business teams that aren’t
directly part of the project. Therefore, cutover needs to be shepherd-
ed with a deep understanding of the impact on the wider business.
Preparation for the cutover needs to start early in the project, and the
governance layer ensures that the cutover is a business-driven process
and not a purely technical data migration process. For example, early
definition and agreement on the related business calendar sets the
right milestones for the data migration to work with.
The cutover from
the previous The legacy systems shutdown window for the final cutover is typically
short, perhaps over a weekend. For some cutover migrations, that
system to the new
window may be too short to complete all the cutover activities, includ-
Dynamics 365 ing data migration. In such cases, the project team may perform the
system is a time- data migration as a staggered, incremental migration, starting with
critical and multi- slow-moving primary data and frozen transactional data. This leaves a
smaller, more manageable remainder to address during the shutdown.
faceted process.
179
 This needs careful governance, and the strategy needs to be decided
early because the data migration engine needs to be able to deliver in-
cremental data loads. You should also carefully consider what business
activities you may need to throttle or perform differently to reduce the
number and complexity of changes between the first migration and
the final cutover. The changes need to be meticulously recorded and
registered (automatically or manually) so that they can be reflected in
the final cutover.

Training strategy
Most projects have plans to train at least the project team members
and business users. All too often though, training is seen as a lower pri-
ority. If any project delays put pressure on timelines or budget, training
can be one of the first areas to be compromised.

This trade-off between respecting the go live date and completing the
full training plan can be easier for the implementation team to rational-
ize because the team is aiming for a go live and the risk of poor training
can be seen (without sufficient evidence) as manageable. The worst
consequences of poor user training are felt in the operational phase.

You can mitigate this (mostly unintentional) unbalanced trade-off with

the following actions:
▪ Clearly defining and communicating the implications on project
success in operational use of poor training delivery.
▪ Ensure that the business approval of the solution includes an
assessment of people readiness. This check should be part of the
entry criteria into user acceptance testing and the results evaluat-
ed as part of the UAT exit criteria and be a meaningful part of the
go/no go criteria for production use.

The test for people readiness needs to be meaningful; it should be an

evaluation of the effectiveness of the training, and not just that it was
made available. The test should be equivalent to assessing whether
enough people in key roles can conduct their critical day-to-day busi-
ness process safely and efficiently using the Dynamics 365 application
and related systems.

Implementation Guide: Success by Design: Guide 8, Project governance 180

 More details on how to put together a good training strategy are in
Chapter 19, “Training strategy.”

Project goals
Project plan
Project organization
Project plan analysis is where the outcomes of a lot of governance topics
Project approach become visible. The effects of good governance are felt on a project
by noting that the project plan is resilient to the smaller changes and
Classic structures unknowns that are a reality for all business application projects. Poor
governance, on the other hand, manifests itself as continuously missed
Key project areas
targets, unreliable delivery, and repeated re-baselining (pushing back) of
Project plan the project plan milestones. The key is for the project planning process
to have its own governance mechanisms to avoid poor practices, detect
risks early, and provide the agility and insights to fix and adjust quickly.

When determining how the project plan should be constructed, the

project plan should be able to demonstrate the following:

Is the project plan designed to cover all the critical areas?

Does the plan cover (at least at a milestone level) all the key areas of
activity? If not, key dependencies and effort will be missed.

Is the project plan actionable?

A project plan should be able to fulfil one if its primary purposes of di-
recting a project. If that isn’t the reality on the project, you should look
at how to restructure it so that it’s the main driver of project activity at
least at the milestone level.

If a plan is thousands of lines long, it’s unlikely to be directly actionable,

and will probably be in a state of continuous update and adjustment.

Similarly, if the project workstreams aren’t paying any attention to

the plan, it’s a clear sign that the project plan may not be actionable.
Actionable project plans help direct the project teams to deliver the
right activity at the right time and reduce uncontrolled activities.

Projects using agile practices may be using some form of prioritized

Implementation Guide: Success by Design: Guide 8, Project governance 181

 backlog management for much of the build and test activities, but there is
normally still a high-level milestone plan to allow for the communication
and management of related tasks that aren’t covered by the backlog.

Whatever methodology is in place, the plan should be regularly evalu-

ated to confirm that the teams are taking their direction from the plan.

Is the project plan up to date, accurate, and realistic?

Compared with the traditional on-premises business application
projects of a few years ago, today’s cloud-based Dynamics 365 imple-
mentations tend to move faster with shorter durations. A project plan
that isn’t up to date is ineffective.

Projects should institute thresholds for how out of date a project plan
is allowed to become. A project plan that is inaccurate in terms of the
estimated effort (and remaining effort), duration, and dependencies
will promote the wrong decisions and allow risks to remain hidden, and
ultimately give an inaccurate picture of the project status.

For a project plan to remain accurate, it needs a robust task

estimation process and a reliable and easy means of tracking tasks in
progress. Consider how to make this possible structurally and with
efficient processes.

Does the project plan give a clear indication of what tasks are critical?
Business application projects have many moving parts and many spe-
cialist areas. The project must be able to accurately derive the critical
path so that project leadership can focus their attention on those tasks.
For agile projects, agile crews prioritizing the backlog and managing
the critical path within short-duration sprints and iterations can pro-
vide the necessary insight.

Accountability for outcomes

Project planning should promote the clear identification of account-
ability for the outcomes (both at the workstream level and overall).
Often, projects create a RACI (responsible, accountable, consulted,
informed) matrix at the start of the project, but this may not provide

Implementation Guide: Success by Design: Guide 8, Project governance 182

 the right level of accountability for all areas. Additionally, this matrix
isn’t always referenced during the project implementation. Asking the
question via the project plan if the accountability for the outcome (not
responsibility) is clear, helps keep the accountability fresh in everyone’s
mind. It keeps the project more agile with faster decision-making.

Status reporting
A well-constructed project plan facilitates accurate project status
reporting. However, it needs to be deliberately designed into the
project plan with the right dependencies, estimated effort, and mile-
stones so it can be easily extracted from the plan. This means that the
project should be deliberate in the detail to which tasks are scheduled
so dependencies and milestones are created with this in mind. Because
status reporting often has milestone-level status as a key indicator, the
meaning of a milestone’s completion must be explicitly defined so the
implications are clear to the intended audience.

Project plans also need a regular and easy-to-use process to keep

task progress updated with an accurate status. You should also define
a uniform and reliable method for units of progress so you get the
same quality of updates from all teams. For example, tasks defined as
80 percent complete often under-estimate the effort to complete the
remaining 20 percent; instead, consider reporting the remaining effort
in hours for a given task.

Some projects, especially ones using more agile practices, may use
alternative or additional analysis and presentation methods such as a
backlog burndown, remaining cost to complete, or earned value, but
these principles apply nevertheless and they all rely on having accurate
underlying data on actual versus expected progress to date and the
remaining effort expected.

A useful way to report the status of a business application project is

to present a functional heatmap view (Figure 8-7), which shows the
readiness of the key processes for the next major milestone (such as
sprint review, SIT, UAT, or go live). The functional heatmap is a business
process perspective of the project plan status.

Implementation Guide: Success by Design: Guide 8, Project governance 183

 Fig.
8-7 Business process heat map example
Legend
Process not complete Process complete,
and not on track in repair, or testing

Process design blocked Process solution in Process complete and

or not started progress and on track ready for next test cycle

Finance

HR

Operations

Logistics

Legend
Process not complete Process complete,
and not on track in repair, or testing

Process design blocked Process solution in Process complete and

or not started progress and on track ready for next test cycle

Finance

HR

Operations

Logistics

Status reports can also include status from other project controls such
as risks and issues, sentiment surveys, or resource availability in prior
months. These aren’t directly derived from a project plan and can
provide a useful supplementary view. However, it’s important to ensure
that actionable conclusions are extracted from these alternative views;
these conclusions should generate tasks that can be measured, planned,

Implementation Guide: Success by Design: Guide 8, Project governance 184

allocated, and subsequently reflected at some level in the project plan.

Project feedback loop

Project status should provide a feedback loop to the project stakehold-
ers, project management, and the project team to adjust the project
based on the findings. This should be a defined process in the overall
project governance to create the discipline to regularly review the
status and generate explicit actions for the project that directly address
the status findings.

The project status should not be the only source for this feedback loop;
consider all the other controls and procedures in the project that can
generate actionable information to establish a systematic mechanism
defined to diligently listen and extract the actions. For example, up-
dates from daily stand-ups, design reviews, sprint playbacks, or even
informal risk discussions can help provide useful feedback.

In summary, create a culture from your governance processes that ac-

tively recognizes the feedback loop and translates feedback into actions.

Conclusion
The primary intention at the start of this chapter was to provide an
overview of the importance of good project governance and provide
guidance on how to assess the effectiveness of your existing or pro-
posed governance model.

We then moved on to describe how many of the critical issues on a

Dynamics 365 project that initially may manifest themselves as tech-
nical issues actually have their roots in failures of project governance.
We also saw that the governance requirements for a Dynamics 365
business application need to address the fact that the project is essen-
tially delivering the capability to perform business processes using the
application. So, the governance model must include strategies and
techniques that can help realize the project goals through the right
engagement with business users and business processes.

185
 We discussed how defining a project organization model that has an
engaged leadership and facilitates rapid and accurate communication
is more aware of the reality of the project. Similarly, project orga-
nizations that enable the right accountability, naturally encourage
cross-collaboration, and reflect the complexity and effort of the project
are much more likely to have good velocity and be successful.

We also discussed how the approach is much more than a technical

methodology, and that the approach must be practical and tailored to
the specific needs of a Dynamics 365 business application and to the
customer’s project and business needs and constraints.

We showed that the mere presence of the usual governance structures

doesn’t assure project success. We can evaluate the effectiveness of
the governance model in very practical ways and with a focus on the
specific needs of Dynamics 365 projects.

We also brought awareness to how a governance perspective on the

specific technical implementation strategies (such as test and build
strategies) can help a Dynamics 365 project achieve better, faster
results. We proposed that governance should be an organic part of
the strategy. These strategies shouldn’t be considered as being purely
subject to technical oversight and control.

Lastly, we discussed how a good project plan and planning process

will expose the outcomes of a lot of these governance topics. Project
planning shouldn’t just be an overall status reporting tool, but be
deliberately engineered to gain early and deep insights into the effec-
tiveness of the different elements of the project work and allow rapid
responses to the deficiencies.

You can use the key objectives of good project governance to judge
the overall suitability of a given governance model. Ideally, you should
test a proposed governance model against these objectives during the
Initiate phase of the project and throughout, but it’s never too late.

In summary, project governance for Dynamics 365 projects should be

a framework that permeates all aspects of our projects, not just the

Implementation Guide: Success by Design: Guide 8, Project governance 186

typical areas that were historically or conventionally regarded as the
domain of governance. Good project governance isn’t a set of gener-
alized bureaucratic or administrative procedures to follow, it’s about
creating a Dynamics 365-specific governance culture of driving the
project to reliably and efficiently deliver the project goals.

187
 Checklist

Project goals Identify relevant project roles and areas of ownership

and assign them to team members with the right exper-
Ensure goals are clear, realistic, mapped to actions and
tise and experience.
measures, correctly translated into project deliverables,
and regularly monitored throughout the project lifecycle.
Project approach
Align goals and ensure they are sponsored by relevant
stakeholders across the organization, and properly com- Analyze, review, and confirm that your chosen imple-
municated and understood by the implementation team. mentation methodology works well with the business
and project constraints and circumstances.

Project organization
structure Classic governance
Align business streams with functional workstreams Establish an effective steering group that understands
for better efficiency and structure, and make sure the enough of the project details to actively steer the
business stream is directly involved in the project. project based on meaningful, actionable, and evidence-
based information.
Secure strong executive sponsorship and active engage-
ment from senior business stakeholders within each Create a risk register with meaningful and actionable risks
business stream. that align with project priorities and are actively addressed.

Ensure cross-team collaboration in which members of Implement stage gate or milestone-driven planning and
each workstream are involved in other workstreams to reviews as checkpoints to better track and communicate
avoid working in silos. implications of the project status.

Plan and budget the project resources in proportion to Implement design review boards to ensure new designs
the effort and complexity of the project. and changes operate within the boundaries of the solu-
tion blueprint and don’t adversely affect other designs.
Define accountability and responsibility at the project
leadership level. Each stream identifies an owner with
autonomy who is empowered to make decisions.

Implementation Guide: Success by Design: Guide 8, Project governance 188

 Case study
Project governance as a
critical success factor
A large enterprise that was using legacy home-grown applications for
managing their sales and customer service business processes decided
to embark on a new cloud application, Dynamics 365. This was a mis-
sion-critical and complex implementation for this company; it would
affect their key business of providing customer service to their large
enterprise customers, and was the first cloud deployment within a
traditional IT managed landscape. The plan was to release a minimum
viable product (MVP) within a timeline of six months and follow a
phased approach for their users to adopt the new solution smoothly
and without too many challenges.

Although the implementation team ensured that the scope was

aligned with business priorities and clear project goals were agreed
upon, no clear processes were defined for the following key aspects of
the project:
▪ Realistic timelines for the activities planned within the engagement
▪ Understanding of technical and business complexity
▪ Well-defined project organization structure
▪ Assignment of adequate and qualified resources
▪ Effective project updates for the steering group
▪ Full understanding of the product’s out-of-the-box capabilities,
thereby avoiding unachievable product requirements

Implementation Guide: Success by Design: Guide 8, Project governance 189

 As the project began, complexities surfaced early in the cycle.
However, the team didn’t revisit and realign their activities. This result-
ed in the following impacts:
▪ Poor quality of deliverables
▪ Mismatch in customer versus partner understanding of requirements
▪ High number of defects identified during initial testing cycles
▪ Incomplete implementation of requirements due to lack of skilled
resources on the project
▪ A continuously shifting end date for final go live, resulting in a
delay of 6-8 months
▪ A complete lack of trust between the customer and partner

After several discussions between customer and partner stakeholders

and a high impact on cost and morale, the team agreed to clearly de-
fine a project organization structure and an approach to continuously
monitor and manage the overall project. The following activities were
incorporated as project governance processes were defined:
▪ Considering the new technology, training needs for both the
customer’s business and IT teams were identified so they could
understand both the SaaS world and Dynamics 365 product
capabilities.
▪ An architecture board was established that was responsible for
diligently reviewing architectural decisions coupled with the gap
analysis. For any gaps identified, this board provided input to
the change control board to ensure that a change management
approach was followed without any further impact on timelines.
▪ Both industry and product architects were assigned to the project
to avoid any further lack of domain understanding and perform a
clear fit gap analysis.
▪ The steering committee was empowered by the customer’s senior
management to take the relevant and necessary decisions through
the project.
▪ A key challenge that impacted the deliverable quality was that
several activities were planned almost parallel to each other and
were to be done by the same project team. As part of the project
reorganization, activities such as performance testing and vulner-
abilities testing were outsourced. This allowed the existing team to
stay focused on their core functional scope.

Implementation Guide: Success by Design: Guide 8, Project governance 190

 As the project team implemented these various controls and processes,
they were able to redefine their project timelines and have now gone
live within that timeline. This has led them to ensure that they can
apply and tailor these learnings to other projects to avoid similar high-
cost impacts.
As illustrated by this and other case
studies, project governance is the most In conclusion, project governance must not be thought of as an af-
critical factor that entails all the key
elements to make a project successful. terthought. It must be set up from the beginning to work as a flexible
backbone for the entire project.

Implementation Guide: Success by Design: Guide 8, Project governance 191

9 Guide
Environment
strategy
Implementation Guide: Success by Design: Guide 9, Environment strategy 192
It all starts with the foundation.

Introduction
Defining your environment strategy is one of the
most important steps in the implementation of your
business application.
Environment-related decisions affect every aspect of the application, from
application lifecycle management (ALM) to deployment and compliance.

At a fundamental level, a good environment strategy is about obtain-

ing the right balance with multiple layers of compliance and security,
productivity, collaboration, isolation, performance, and maintainability.
A strategy to manage environment provisioning and access, and con-
trolling resources within them, is key to:
▪ Securing data and resource access
▪ Organizing and structuring solution components
▪ Governing and managing capacity

In this chapter, we explore several deployment and transition scenarios.

By the end of this chapter, you should understand the factors to
consider when defining your environment strategy, and the available
patterns that will enable you to confidently deliver an environment
plan for your applications.

Implementation Guide: Success by Design: Guide 9, Environment strategy 193

 Environment strategy
Environments are containers that store, manage, and share your orga-
nization’s data. They also store the data model, application metadata,
process definitions, and the security constructs to control access to
data and apps.

A successful environment strategy defines the core principles and

policies for creating, granting access to, and decommissioning en-
vironments of different types, along with the necessary governance
process. These policies should provide a consistent approach for cen-
trally deployed, IT-driven implementations and for citizen-developed
applications for individuals or small business teams.

A data strategy is likely to inform or impose constraints on an environ-

To get started with creating ment strategy. For example, an organization could build a unified data
an environment strategy, platform that different apps and services consume, or there could be a
ask the following questions: fragmented legacy data estate that must be accounted for in the cloud
• Where are my users physically located? environment strategy. Chapter 10, “Data management,” delves into the
• What are my organization’s data
patterns and scenarios that are relevant to environment planning.
compliance and residency requirements?

• What types of environments are

required, and when? It should be apparent from the sidebar questions that environment
• What are the different roles (such as
strategy is not the responsibility of a single person or team. It can
developers, testers, admins, business
analysts and trainers), and which involve IT admins, the architects responsible for the design and imple-
environments do they need access to?
mentation of the solution, the information security and compliance
• What’s the access procedure for
an environment? teams, and the business stakeholders.
• Which apps will be installed in
the environment?
Environment-related decisions are hard and expensive to change. As
• Will the solution require more than one
production environment? we explain throughout this chapter, environment strategy can affect
• Which services and third-party how solutions are implemented. This exercise needs to be done at the
applications do I need to integrate with?
start of the project, even before beginning implementation, and is not
something that should be left to the end as a part of a go-live plan.

Tenant strategy
Understanding tenants is fundamental to defining an environment

Implementation Guide: Success by Design: Guide 9, Environment strategy 194

 Fig. strategy for your deployment. A tenant is a logical structure that rep-
9-1
resents your organization. Every cloud service to which an organization
Services subscribes will be associated with the customer’s tenant. A tenant is
Dynamics 365 not dedicated to a specific service; it has subscriptions for Microsoft
services, such as Exchange, Dynamics 365, Power BI,
Microsoft 365
Tenants and Azure. Every tenant is given a Microsoft domain,
Azure microsoft.onmicrosoft.com such as yourorganization.onmicrosoft.com, but
admins can also associate their own custom domains
yourorganization.onmicrosoft.com with the tenant. For example, as shown in Figure
9-1, contoso.onmicrosoft.com can be associated
contoso.onmicrosoft.com
with contoso.com and other domains owned by the
organization. You can also federate access to your tenant via your or-
ganization’s on-premises Azure Active Directory (Azure AD) to enable
single sign-on (SSO) across on-premises and cloud services.

A tenant provides the highest level of isolation for a cloud service, as

it represents an organization and differentiates one customer from
another. Tenants never share resources or licenses because they’re
meant to represent an organization and not a department or business
unit. (In some cases, subsidiaries could share a tenant with the parent
organization, depending on the organization structure, and the tenant
model might mirror that structure.)

Isolation approaches also may differ between services. For example, as

shown in Figure 9-2, you could have different subscriptions in Azure
within the same tenant, and each subscription could host different

Fig.
9-2

Test
Operations
environment
Dynamics 365
and Power Platform
Production
Tenant Sales app
environment

[email protected]

SharePoint
Microsoft 365 SharePoint
site

Azure Subscription Web roles

Implementation Guide: Success by Design: Guide 9, Environment strategy 195

 applications and have a different set of users, because Dynamics 365
and the Microsoft Power Platform have environments that isolate data
and manage user access.

Your environment strategy should provide a balanced level of isolation

to meet the security and compliance needs of your organization. It also
should take into consideration the administration, ALM, and collaboration
Fig.
9-3
needs of the project. To define the right tenant and environment strategy
for your organization, it’s necessary to understand the controls that are
Global single available for each service to isolate code, configuration, data, and users.
tenant vs. global It’s also important to note that the isolation provided by services doesn’t
multitenant setup necessarily reflect the underlying infrastructure or design of a cloud
service. A separate tenant doesn’t mean it’s a separate server farm, and
Global single tenant having a separate environment doesn’t give you a different front end.

Product Global single-tenant setup

Using a single Microsoft tenant to represent the organization in the
Account Contact Addresses Microsoft cloud (Figure 9-3) is a common setup. It provides unified
administration for user access and licenses, enables seamless integra-
tion between services, and lets the organization share tenant resources.
Global multitenant
All Dynamics 365 and Power Platform environments will be a part of
the same tenant. There could be different apps deployed in different
Product #1
environments, or they could have different users for each app and
environment, but all would belong to the same Azure AD that is asso-
Account Contact Addresses ciated at the tenant level. Using the Multi-Geo setup, you could create
environments in different countries or regions to meet your compliance
Product #2 and application needs, but they would remain a part of the same orga-
nization-wide tenant. Sovereign cloud deployment requires a separate
Azure AD and might have additional regulatory restrictions that limit
Account Contact Addresses
your ability to create environments in different countries or regions.

Product #3
Let’s examine some of the pros and cons for a global single-tenant setup.

Pros of a global single-tenant setup:

Account Contact Addresses
▪ It aligns to the concept of tenant and the way cloud products
are licensed.

Implementation Guide: Success by Design: Guide 9, Environment strategy 196

 ▪ It allows sharing of licenses and quotas across the tenant per
service. (Microsoft Dataverse storage is one such example.)
▪ It provides a central place for IT admins to manage the services,
users, and licenses.
▪ It is easier to deploy organization-wide IT policies such as data loss
prevention (DLP) and conditional access.
▪ It is simpler to manage service account configurations and con-
nections when moving from one environment to another in the
same tenant.
▪ It facilitates easier and safer integrations between tenant services,
without the need to elevate privileges to cross service boundaries.

Cons of a global single-tenant setup:

▪ A tenant can only be associated to a single org Active Directory
forest. If there are independent subsidiaries using a separate Active
Directory forest, they can’t be added to tenant, but can use alterna-
tives such as Azure business-to-business (B2B) for guest-user access.
▪ In a single-tenant setup, you can create tenant-wide policies. All
environments, including the production environment, are gov-
erned by the same policies. Creating exceptions for trying out new
services or testing might require the team to go through multiple
information security hoops. (For example, a proof of concept or
pilot requiring a different Active Directory conditional access
policy will attract scrutiny from security.)

Global multitenant setup

Usually, the tenant structure for the Microsoft cloud software as a
service (SaaS) mirrors the structure of the organization. But some orga-
nizations operate as separate business entities in a country or region,
even though they might be part of a single umbrella brand. In such
cases, an organization might adopt a multitenant setup (Figure 9-3), as
each business entity might have its own contractual needs, regulatory
constraints, and licensing agreements.

Alternatively, organizations can have separate tenants for develop-

ment, testing, and live systems. This primarily supports the information
security requirements to isolate data and keep users who aren’t

Implementation Guide: Success by Design: Guide 9, Environment strategy 197

involved in production from accessing production systems. Still, using
separate tenants is generally not recommended because there could
be data and access isolation at the service level to meet information
security requirements. (For example, you might use security groups on
environments to limit access.)

User accounts, identities, security groups, subscriptions, licenses, and

storage can’t be shared among different tenants. So, using separate
tenants might require replication of all other services or creation of
stubs for external systems, per tenant, for all the integrations. Also,
duplication of licenses and quotas across tenants may be needed
to support ALM and performance testing. Admins would have the
additional responsibility of ensuring that all the policies and settings
between tenants are in sync to avoid deployment failure between
tenants. Some organizations do go down this path because of their
historical setup or a lack of understanding of the controls available to
restrict access at the service level.

Let’s examine some of the pros and cons for a global multitenant setup.

Pros of a global multitenant setup:

▪ It works when different subsidiaries are fully independent, with
different systems to manage user identities in separate domains,
and have separate licensing needs.
▪ It provides the highest level of isolation that will simplify the
security approvals but is not generally required.
▪ It is easier to manage and monitor the cost per tenant without
the need for internal cross-charging and tracking usage of shared
resources on a single tenant. (For example, storage is shared on a
single tenant.)

Cons of a global multitenant setup:

▪ Licenses can’t be shared across different tenants, which means
that storage, API add-ons, and other tenant-level quotas must be
purchased for each tenant, and you may have to purchase sepa-
rate licenses for the same user to access separate tenants.
▪ Overhead to maintain tenant-level Active Directory and DLP
policies consistently across tenants can lead to surprises during
production deployment.

198
 ▪ Service-level admin actions, such as ability to copy an environ-
ment, may not be available across tenants, which can make testing
or troubleshooting difficult.
▪ The build automation and continuous integration and continuous
delivery (CI/CD) pipelines that span multiple tenants can be more
complicated and might require manual intervention, especially
when managing connections to the service.
▪ You may have to purchase a significant number of licenses for con-
ducting testing. With load testing, for example, you can’t reliably
simulate a load from 1,000 users using five test-user accounts.
▪ If you’re using capacity add-ons, you will have to make duplicate
purchases for each tenant to develop and test your solutions.
▪ Integrations with other services that can’t be done across tenants,
which means potentially purchasing licenses for other Microsoft
services for each tenant.

Overall, managing your ALM process across several tenants brings

unnecessary complexity to licensing and technology, and delivers little
value in terms of security, because you might already have the neces-
sary controls at the service level to limit access to business data.

Key factors affected by

an environment strategy
In this section, we explore the key factors directly affected by an environ-
ment strategy, including security and compliance, ALM, operations, and
functional and nonfunctional aspects of the solution. Considering these
key factors will help you define an environment strategy that meets the
needs of your organization, and set the direction for the future growth of
these applications to meet evolving business requirements.

Compliance
Security and compliance are critical considerations for an environ-
ment strategy. Each organization needs to ensure that data is stored
and processed in accordance with local or regional laws, such as

Implementation Guide: Success by Design: Guide 9, Environment strategy 199

data-residency requirements for a specific region. For example, the
European Union (EU) enforces the General Data Protection Regulation
(GDPR) for EU residents, as well as its citizens outside of the EU.

At the onset of the project, your organization must determine your envi-
ronment’s compliance requirements. These can vary widely depending
on the industry, regulations, business type, and user base, and will need
to be approved by your internal security and compliance teams.

The most common elements affecting compliance are:

▪ The physical location of the environment(s), which usually deter-
mines data residency
▪ The specific features and apps in the scope and their data-
processing requirements
▪ The encryption (at rest and in transit for confidential information)
▪ The disaster recovery procedures to restore the service and data
▪ The data-retention policies, and the ability to back up and restore data
▪ The personally identifiable information (PII) standards for data
protection, and ensuring the service is compliant with the regula-
tory requirements for the industry and region

Application design
The environment strategy can affect the application design. Conversely,
the needs of an application can drive the environment requirements,
and it’s not uncommon for IT systems within an organization to reflect
the actual structure of the organization. Depending on your organiza-
tion’s strategy for data isolation, collaboration, and security between
different departments, you could choose to have a single shared envi-
ronment or create isolated environments. For example, a bank might
allow data sharing and collaboration between commercial and business
banking divisions while isolating the personal banking division, which
reflects the bank’s application design and environment strategy.

Environment policies should not be created in isolation. Creating a

generic organization-wide policy, such as “Every application should
use a separate environment” or “All applications should share a single
environment,” without considering the business requirements or

200
 understanding the underlying data and integration dependencies,
could lead to unnecessary complexity and fragmentation.

The data store for an application and the supporting business pro-
cess plays a key role in the environment decision. If multiple apps for
different departments can benefit from leveraging each other’s data,
a single environment with integrations can improve consistency and
collaboration. The user experience can be tailored via dedicated apps
for different personas and secure data access using the security model.

For sophisticated systems that need to use multiple environments and

potentially sync data between them, it could be technically possible,
but such patterns require careful consideration of performance and
API capacity and are best avoided.

App integrations also play a key role, as multiplexing with different

environments may not work. For example, a user may not be able to
simultaneously sync emails to several environments.

Performance
Microsoft cloud services provide a high degree of scalability and per-
formance. Based on considerations such as network latency, firewalls,
network traffic monitoring, organizational proxies, and routing by in-
ternet service provider (ISP), globally distributed users can experience
higher latencies when accessing the cloud. This is why we recommend
creating a latency analysis matrix (Figure 9-4) for solutions that have a
globally distributed user base.

Fig.
9-4

Environment
User location Device Network Latency Bandwidth
location

North America Corporate

Canada Browser 80 ms 6.4 Mbps
(NAM) network

North America Corporate

Amsterdam Tablet 120 ms 8 Mbps
(NAM) Wi-Fi

Implementation Guide: Success by Design: Guide 9, Environment strategy 201

 This exercise gives a fair idea of how the network latency will affect the
user experience based on the location of your environment. This infor-
mation can be used to make a balanced choice so most users have an
acceptable level of latency, and application design can be optimized
for users in high-latency locations. For example, using lightweight
forms or reducing the number of records shown on a page can enable
faster loads.

Also, environment types should be chosen carefully, based on the

workload and the capacity needed. Using trial or low tier sandbox envi-
ronments, for example, will not deliver the same level of performance
as live production environments, so it may not be suitable to develop
or test the solution on a trial or low tier sandboxes.

Scalability
Scalability of the SaaS platform is a critical consideration for business
applications. Traditionally, scaling up in on-premises deployments was
about adding more servers or more CPU, memory, or storage capacity
to existing servers. In a cloud world with elastic scale and microservice
architecture, the server could be replaced by an environment and the
compute and data transfer units by the API capacity. (This is just used
as analogy, it’s not a one-to-one mapping where one environment
corresponds to a server in the SaaS infrastructure.)

The scalability parameters for a SaaS could be:

▪ How many parallel connections or concurrent users can you sup-
port with one application or in a single environment?
▪ How many API calls is a user entitled to make?
▪ What are the limits on workflow or code executions?

The following is a potential SaaS cloud interpretation of vertical and

horizontal scalability.

Vertical scalability
Organizations commonly operate single environments supporting
thousands of users, and each user has a defined API entitlement based
on the license type assigned. The environment’s storage grows as more

Implementation Guide: Success by Design: Guide 9, Environment strategy 202

 users and applications store their data. Dynamics 365 SaaS services
are built on a scalable cloud infrastructure that can store terabytes of
data to meet the requirements of large enterprises. When it comes to
workflows automation, each environment can have any number of
Microsoft Power Automate flows, each with thousands of steps per-
formed per day, which can be further scaled using appropriate license
types and add-ons. There are no hard limits on the number of apps you
can have per environment, but you can’t have multiple first-party apps
of the same type in the same environment.

Horizontal scalability
With horizontal scalability, organizations can have several separate en-
vironments, with any number of Power Automate flows on the tenant.
There are no native sync capabilities between environments, and you
still need to take license entitlements into consideration, especially
when it comes to tenant-wide storage and API entitlement.

Maintainability
Maintainability measures the ease and speed of maintaining a solution,
including service updates, bug fixes, and rolling out change requests
and new functionality.

If you have several applications sharing common components in

the same environment, you should consider maintainability when
upgrading or deploying a new release. It’s also important to invest in

The effort to automation testing so you can run regression tests and quickly identify
any dependencies that could cause issues.
maintain the
solution is directly The effort to maintain the solution is directly proportional to the number
proportional to of environments involved. For example, testing a release wave or analyz-

the number of ing the impact of deprecations is easier when there is just one production
environment with the Dynamics 365 Sales app and the Dynamics 365
environments Customer Service app, compared to a solution that uses different
involved. production environments for the Sales and Customer Service apps.

203
 ALM
ALM includes the tools and processes that manage the solution’s
lifecycle and can affect the long-term success of a solution. When
following the ALM of a solution, consider the entire lifespan of the
solution, along with maintainability and future-proofing. Changes to
your environment strategy will directly affect the ALM (and vice versa),
but it’s important to be clear that environments are not the repository
of your code and customizations.

Environments can be used to separate business functions or for differ-

ent purposes, such as development, testing, and production. Typically,
at least three environments are necessary to support a mature re-
lease-management process.

Types of environments
Production environments are meant to support the business. By de-
fault, production environments are more protected for operations that
can cause disruption, such as copy and restore operations. Sandbox

Environments can be used to develop, test, and delete as required.

Purposes of environments
▪ Development  One or more development environments are usu-
ally required, depending on the customization requirements and
time available. Development environments should be set up with
proper DevOps to allow for smooth CI/CD. This topic is covered in
more detail in Chapter 11, “Application lifecycle management.”
▪ Quality assurance (QA)  Allows for solution testing from both
a functionality and deployment perspective before the solutions
are given to the business teams in a user acceptance testing (UAT)
environment. Only managed solutions should be deployed here.
Depending on project requirements, there can be multiple testing
environments, including regression testing, performance testing,
and data-migration testing.
▪ UAT  Generally the first environment that business users will have
access to. It will allow them to perform UAT testing before signing
off deployment to production environment for go live.
▪ Training  Utilized to deliver training. It allows business users to

Implementation Guide: Success by Design: Guide 9, Environment strategy 204

 practice in a simulated environment without the risk of interfering
with live data.
▪ Production  The live system for business users.

Citizen development
One of the key value propositions of the Power Platform, the underlying
no-code/low-code platform that powers Dynamics 365 Customer
Engagement apps, is that it enables people who aren’t professional de-
velopers to build apps and create solutions to solve their own problems.

When planning Central IT still has governance responsibility and creates guidelines
to secure data with DLP policies. But business users should be able to
your environment build apps and automate their work while accessing data in a secure
strategy, consider way. This organic growth of applications by business users facilitates
any future phases digital transformation at a massive scale. Thousands of organizations

and rollouts of have adopted this “citizen developer” philosophy to quickly roll out
hundreds of apps across the organization, creating a community of
the solution, as engaged employees who are helping realize the vision of an agile busi-
well as changing ness that can evolve to meet customer needs in days or weeks instead
requirements. of months or years.

The environment strategy that enables this kind of organic growth will
be different from a traditional IT-developed solution. It’s crucial to de-
liver an agile and automated process where business users can request
environments to enable the maker experience and connect to data in
a secure way while conforming to the standards for data modeling and
application design set by the organization.

Future-proofing
Future-proofing is the process of developing methods to minimize
any potential negative effects in the future. It can also be referred to as
resilience of the solution to future events.

Your environment strategy needs to take into consideration any future

phases and rollouts of the solution, and allow you to deal with changing
requirements and build a solution that will not limit the business or take

Implementation Guide: Success by Design: Guide 9, Environment strategy 205

 away necessary control. As an example, consider country-specific needs,
as well as variations to and deviations from the standard process.

Environment app strategy

If you’re deploying multiple business apps on the Dynamics 365 plat-
form, you will need to decide on an environment app strategy. Should
all apps be deployed in the same environment? Or should each app
have an environment of its own?

There isn’t a standard answer or blanket approach that will work for
all apps within your organization. The best approach for you is the
one that facilitates collaboration and cross-pollination of information
between apps, while also reducing data silos and meeting your organi-
zation’s security, compliance, and data protection requirements.

Multiple-app environment
In a multiple-app environment scenario (Figure 9-5), a single pro-
duction environment is used for multiple apps. For example, the
production environment might have the Dynamics 365 Marketing and
Sales apps to enable collaboration between the marketing and sales
teams, and facilitate a quick transfer of qualified leads to sales repre-
sentatives. Similarly, having the Sales and Customer Service apps in the
Fig. same environment gives the sales team insights into customer support
9-5
experiences, which could affect

Multiple-app environment ongoing deals.

With multiple-app deployment,

Production environment the app data, security models,
and data models are in a common
Dynamics 365 apps repository, allowing the apps
to reuse any integrations. The
security model design will be used
Sales Customer Marketing Field Project to limit access to data for each
Service Service Operations
app, with the user experience
defined by the individual app. For

Implementation Guide: Success by Design: Guide 9, Environment strategy 206

 example, the Sales and Marketing apps might use the same lead table,
but security roles and the app definition will control access to records,
fields, and processes.

Let’s examine some of the pros and cons for the multiple-app deploy-
ment model.

Pros of a multiple-app deployment model:

▪ It enables stronger collaboration between business teams by
decreasing data silos.
▪ It reduces the number of environments to manage, as well as the
amount of effort needed for platform updates.
▪ It allows reuse of integrations, and its simpler design lowers
API consumption while avoiding the need for data sync across
environments.
▪ It simplifies reporting and the business intelligence architecture.

Cons of a multiple-app deployment model:

▪ Its security model could become complex, depending on access
restrictions.
▪ Its ALM and release process needs to be coordinated for each
app and will require full regression testing when changing shared
components, making automation testing even more critical.
▪ Its capacity utilization for individual apps is more difficult to track.
▪ Its security, data models, or integrations, if poorly designed on one
app, can affect other apps.
▪ It might have limitations on deploying several apps of same type in
a single environment. (For example, you can’t have several Power
Apps portals with the same template in a single environment.)
▪ It can’t be used if you need to segregate the environments for a
globally distributed user base due to performance, compliance
regulations, or process differences.

Per-app environment
In a per-app deployment model (Figure 9-6), every application gets
its own production environment, with a separate set of environments
to support the underlying ALM and release process. There is com-
plete isolation of the data, schema, and security model. A per-app

Implementation Guide: Success by Design: Guide 9, Environment strategy 207

 environment might seem simpler from a deployment perspective, but
it can create data silos such that one business process cannot easily
Fig.
benefit from sharing information with another, leading to additional
9-6
effort in building complex integrations.

Per-app environment Also, the security model is defined

per environment, so you can’t use
Production Production Production the platform security constructs
environment environment environment to define which data from one
environment will be accessible to
App App App a user in a different environment.

Consider an example where the

Sales and Customer Service apps
Sales Customer Field
Service Service are deployed in separate environ-
ments, requiring core customer
records to be synchronized across
the two. In this scenario, providing sales representatives with visibility into
support experiences will require some level of integration, either by copy-
ing over the data periodically or integrating the visual user interface (UI).

This approach might be more relevant for isolated functions within an

organization where there is little need for data sharing and collaboration.
For example, an internal HR ticketing system for employees has little to do
with the sales process. Network latency, compliance, and region-specific
requirements could mandate the use of per-app environments, but will
introduce data-replication challenges, integration redundancy, and
increased costs for storage, integration, and maintainability.

Let’s examine some of the pros and cons for the per-app
deployment model.

Pros of a per-app deployment model:

▪ Its security model is separately managed for each app, which simpli-
fies configuration when there is no need for data sharing between
environments or development of custom security components.
▪ Its ALM and release process for each app can be independent.
▪ Its capacity utilization for individual apps can be tracked easily for

Implementation Guide: Success by Design: Guide 9, Environment strategy 208

 cross-charging within the business.
▪ Its design issues in one app will not directly affect other apps.
▪ It is the preferred approach if there is a need to segregate the
environments for a globally distributed user base due to network
latency, compliance regulations, or process differences.

Cons of a per-app deployment model:

▪ It can create data silos between business units and departments,
potentially reducing collaboration and value to the business.
▪ It means a linear increase in the number of environments to sup-
port ALM for each app and the administration effort for platform
updates.
▪ It largely prevents reuse of integrations and is a potentially com-
plex solution requiring data sync and complex custom security
logic. In some cases, integrations are not possible across several
environments. (For example, you can’t sync users’ Exchange mail-
boxes to several environments at the same time.)
▪ It potentially brings higher API and storage costs due to data and
integration redundancy.
▪ Its reporting and intelligence architecture will be more complex
and require data unification in an external data store, such as a
data lake.
If you’re deploying
more than one
app, you will Global deployment
need to carefully scenarios
consider these
This section will focus on three common app-deployment models you
factors in could use, along with the tradeoffs, pros, and cons when choosing
coordination with one approach over the other, to help you find the best option for
the multiple- your deployment.

app and per-app

We’re discussing global deployment scenarios in the context of a
environment single-app deployment. If you’re deploying more than one app, you
strategies covered will need to carefully consider these factors in coordination with the
in the previous multiple-app and per-app environment strategies covered in the

section. previous section.

209
 Global single environment
A global single environment is just one environment deploying an
app that’s accessed by users across the globe. All the data, processes,
code, customizations reside in a single environment. Based on the
app’s security requirements and regional data-sharing policies, your
organization will have to manage access via security roles. This is the
most common approach, as it enables strong global collaboration and
a unified view of data and processes across different regions, business
units, and legal entities.

While a global single environment simplifies deployment and main-

tenance of the solution, centralized process governance is needed to
ensure that different regions conform to a unified process, with devia-
tions kept to a minimum. If processes are not unified globally, handling
region-specific requirements will require a complex solution, likely with
poor performance, which ultimately will affect usage and adoption.

Your organization might have to adjust the security model for each
region to meet local regulations and accommodate cultural differences
around sharing and collaboration.

Let’s examine some of the pros and cons of a global single environment.

Pros of a global single environment:

▪ Data unification across regions and countries enables greater
collaboration and helps standardize processes
▪ Settings, updates, configurations, and ALM are easier to manage
▪ Costs of storage and integrations are reduced by avoiding duplication
▪ Intelligence and reporting architecture, as well as master data
management, are simplified

Cons of a global single environment:

▪ Change management and governance could be harder when
trying to make a process work for everyone while avoiding
political challenges
▪ Poor adoption is possible in regions where users aren’t engaged or
if their requirements are deprioritized
▪ Security model configuration could become complex, as each

Implementation Guide: Success by Design: Guide 9, Environment strategy 210

 persona might have a different regional flavor
▪ Some regulatory and data-residency requirements would be hard
to meet in a global single environment, and could slow down
change and business agility if the effort isn’t well coordinated
▪ High network latency for users in different countries or regions can
affect performance and adoption
▪ Maintenance schedules and downtime are difficult to manage

Global multiple environment

Your organization Multiple-environment scenarios support different business units, de-
partments, teams, countries, or other structures inside an organization.
might have to
adjust the security Each environment has its own database, and data isn’t shared across
model for each environments. This builds a physical security aspect on top of logical,

region to meet out-of-the-box security based on security roles. In separating environ-

ments by region or country, solutions included in each environment
local regulations can support local business and compliance requirements.
and accommodate
cultural differences If users are distributed over large distances, particularly for global

around sharing and deployments, multiple environments may be more suitable because
of the implications (such as latency) associated with the connection
collaboration. infrastructure, which can significantly affect the user experience.
Distributing environments to provide users with more local access
can reduce or overcome network-related issues, as the access occurs
over shorter network connections. For example, with Dynamics 365
Commerce, customers can deploy region specific Commerce Cloud
Scale Units to provide better latency to store and ecommerce users
and elastic scale and easier ALM processes.

In terms of ALM, the solution-release process can be unique for each

environment, which will support specific business units, departments,
and countries. However, it will increase the number of environments
needed to support development, testing, and training activities, as
each production environment needs its own environment to support
the solution-release process.

Because the data is not unified in a global multiple-environment setup,

Implementation Guide: Success by Design: Guide 9, Environment strategy 211

 the reporting and analytics strategy is more complex, and typically
there needs to be a data-consolidation exercise for reporting on key
performance indicators (KPIs) and insights about the overall business.

This approach is common when companies have different require-

ments across subsidiaries, business units, or countries. By creating
multiple production environments, it’s possible to easily implement
specific business requirements and have different teams delivering the
solution within different time frames. Each subsidiary, business unit, or
country operation has control over its own environment.

This model makes it easier for local operations to run and maintain
their solutions, but it also increases the costs, and may lead to siloed
work and deployment.

Let’s examine some of the pros and cons of a global multiple environment.

Pros of a global multiple environment:

▪ Individual environments are only responsible for fulfilling their
own requirements, making them simpler and smaller
▪ Local administration minimizes language and time-zone barriers
▪ Local environments may perform better when closer to the end
user, reducing latency
▪ Compliance with local requirements, such as GDPR, is easier
By creating ▪ Data can be physically separated, and there’s flexibility in manag-
multiple production ing data residency
environments, ▪ Maintenance time windows are easier to adjust

it’s possible to Cons of a global multiple environment:

easily implement ▪ Data may need to be unified in a central repository for reporting

specific business and analytics

▪ Data may not be updated across environments or may become
requirements and specific to each environment
have different ▪ It’s more costly
teams delivering ▪ Users may need to access different environments for different

the solution within purposes, affecting the user experience

▪ Multiple instances must be managed
different time
frames.
212
 Sp t
ok n
Hub-and-spoke modele C environ me
A hub-and-spoke model (Figure 9-7)
Hu depicts the company’s
n core
b-an io
d
-spoke C sol ut
unified business solution as the hub. The core is then distributed and
extended by the subsidiaries, business units, or country operations,
Fig.
9-7 which are depicted as spokes.

Global multiple environment Global single environment

n Hu
tio b-
lu Spo a
n Sp
ent ke tio ok
so

nd
nm
A

-sp

lu
ke

eB
ok
o

so
po

vir

eB
en
d-s

sol
Spoke A
Spoke A en

solut
viro
Hub-an

ution
Hub environment
ion
nment

Hub solution
Hub solution

Sp t
ok n
eCe me Sp
o ke C on
n v iro n s o l u ti
Hu
b-an io n
d-spoke C solut

The hub-and-spoke model could be a variation of the multiple-en-

vironment model where each of the regions have an independent
Global single environment
environment, but share the core data model.

Alternatively, this could also be achieved with multiple apps in a sin-

n Sp gle environment, where each region is able to independently manage
tio ok
the variations.
lu

eB
so

sol
Spoke A

ution

Hub solution
Environment lifecycle
scenarios
Sp Environments transition through various states before they are decom-
o ke C on
s olu ti missioned. Some of these transitions are natural to support ALM, but
Implementation Guide: Success by Design: Guide 9, Environment strategy 213
 there could be business, compliance, and technical reasons that trigger
these transitions. As described earlier, environment-related changes are
complex, and they require significant planning and (usually) downtime.

There are multiple factors that could trigger environment transitions,

ranging from the overall ALM process to changes in business struc-
ture. Let’s explore the various types of transitions that you might see
throughout the lifecycle of an environment. These transitions could
require admins with appropriate roles to self-serve and, in some cases,
seek assistance from product support teams.

Creation
An organization might create a new environment for several reasons.
The purpose of the environment, and considerations such as the envi-
ronment type, country or region, apps required, access security groups,
URL and agreed-upon naming conventions, and tiers and service
levels, should be well defined and understood before its creation.

Environment transition
An organization might have multiple types of environments, each
targeting specific use cases, including trial environments, default
environments, sandboxes, and production environments. Be aware
of possible transitions between types, as there could be limitations
and implications to the service. For instance, changing a production
environment to a sandbox might change the data-retention policy and
applicable service-level agreements (SLAs).

Copy
Most often used in troubleshooting scenarios, environment copy lets
you create a copy of an existing environment with all its data or only
the customizations and configuration. Be aware of the effect of storage
capacity and compliance requirements, which could restrict copying of
customer data. Sometimes copy is also used to support ALM instead of
using the source control. This pattern should be avoided, and environ-
ments should never be used as a repository for code and customizations.

Implementation Guide: Success by Design: Guide 9, Environment strategy 214

Restore
Sometimes, you might need to restore an environment. Depending
on the type of environment and service, there could be several restore
points or a feature allowing you to restore an environment to a precise
time. Restore could lead to data loss, so it’s not really an option for
production environments unless performed in a limited window.

Geo-migration
Geo-migrations could be triggered by changes in regulations, or
simply because of a need to move an environment to a lower-laten-
cy region for a better user experience. This would normally involve
creating a copy of an environment and then physically moving it to
a different region. It almost certainly will involve downtime for users,
and requires approval from the compliance and security teams. It also
could lead to change in service URLs and IP ranges, affecting integra-
tions and network security.

Tenant-to-tenant move
Any change in tenant strategy might trigger a request to move an en-
vironment from one customer tenant to another. This is not a common
request and will require several steps before the actual move. Most
importantly, users must be correctly mapped between tenants and the
record ownerships must be restored. It might require a regional migra-
tion first and could involve several hours of downtime.

Merged environments
Merging multiple environments into a single environment takes
substantial planning and could be extraordinarily complex. Merging
involves the data model and the processes in the app, followed by the
actual data with necessary transformations and deduplication, then the
security model and, finally, the integrations.

215
 Split environments
Splitting an environment could be required if an organization
transitions to a multiple-environment strategy. This could involve pro-
visioning new environments and migrating all previous customizations,
along with the relevant data subset. Alternatively, an organization
could create the split by copying the environment and moving it to
another region.

Migration from on-premises to cloud

Many organizations have on-premises infrastructure, but the evolv-
ing needs of businesses and the potential and benefits of the cloud
lead them to migrate to a SaaS cloud solution. Traditionally, this has
involved creating new cloud environments, running a cloud-compat-
ibility assessment, moving the solution components and data, and
redesigning parts of the solution to take advantage of the latest cloud
features. To help speed up this process, Dynamics 365 services offer
a “lift and shift” solution to move data into the cloud. It’s fast and
cost-effective, but by moving a legacy solution into the cloud, organi-
zations may miss an opportunity to optimize their processes.

Administration mode
Administration mode or maintenance mode can be used for mainte-
nance when only selected users can access the environment. Before
you do this, assess the systems that will be affected by putting an
environment in administration or maintenance mode, such as a pub-
lic-facing portal that connects to the environment.

Deletion
Deleting an environment removes all the data and customizations, as
well as the option to restore and recover. The process of decommis-
sioning or deleting an environment needs gated approvals.

Implementation Guide: Success by Design: Guide 9, Environment strategy 216

 Governance and control
Establishing a center of excellence (CoE) for business applications that is
responsible for defining guidelines, creating best practices, and nurturing
organic growth plays a fundamental role in driving business innovation.

Good governance is critical during deployment and for long-term

system operation. But the governance and control processes for a
centrally deployed and managed IT solution might be vastly different
from processes used to deliver a secure ecosystem of business-user-de-
veloped applications.

A CoE helps enforce A CoE helps enforce security and policies, but also fosters creativity and
innovation across the organization. It empowers users to digitize their
security and policies, business processes, while maintaining the necessary level of oversight.
but also fosters
creativity and With the changing technology landscape and evolution of security

innovation across mechanisms, organizations should review and update policies on a

continuing basis. The better organizations understand the service ar-
the organization. chitecture, the greater the opportunity to fine-tune the controls, rather
It empowers users than creating overly restrictive policies.
to digitize their
business processes, To define a robust environment governance policy:
▪ Examine your organization’s usage, application types, user perso-
while maintaining nas and location, data confidentiality, and access control
the necessary level ▪ Evaluate your capacity needs in terms of storage, API calls, and
of oversight. other factors
▪ Understand the release and rollout strategy, and how it affects the
environment lifecycle
▪ Enforce audit requirements at the environment and application levels.
▪ Ensure DLP policies apply tenant-wide
▪ Define admin roles and processes to obtain admin privileges
▪ Monitor deployment and usage

Implementation Guide: Success by Design: Guide 9, Environment strategy 217

 Product-specific
guidance: Operations
Throughout this chapter, we have discussed concepts related to
environment strategy that are product-agnostic and apply to most
cloud deployments.

Now we’d like to offer some guidance specific to the finance and op-
erations apps like Dynamics 365 Finance, Dynamics 365 Supply Chain
Management, and Dynamics 365 Commerce. A team implementing a
project with these apps requires environments to develop, test, train,
and configure before production. These nonproduction environments
come in a range of tiers with pre-defined sizing, topologies, and costs.
Understanding these basics is key to charting out a well-designed
environment plan.

What is environment planning?

Environment planning is the process of describing the details of all the
environments required during and after implementation. It includes
procuring and preparing all environments that the team will need to
adequately develop and test the Dynamics 365 solution.

For Finance, Supply Chain Management, and Commerce projects, the

FastTrack team provides a sample environment plan in a matrix format
that lists the type, topology, deployment option, and due date for each
environment. This plan is a living document that should be revisited
after each milestone of the project.

The importance of environment planning

Implementing an enterprise resource planning (ERP) and a customer
relationship management (CRM) project typically requires collaboration
among multiple teams that will need to access the apps for different rea-
sons: to develop, build, test, and train. So, part of your plan should cover
To learn more, check out the Environment making environments available for different teams to ensure they can
Planning for Finance and Operations
Implementations Tech Talk. deliver their respective tasks on time. Also, if you plan for all needed envi-
ronments early, you will know what the costs are going to be throughout

Implementation Guide: Success by Design: Guide 9, Environment strategy 218

 the project and at which stage(s) you need the funds, and you will avoid
A note on LCS projects project delays from last-minute discoveries of missing environments.
In the case of Dynamics 365 Finance,
Supply Chain Management, and
Commerce, all environments are
contained in one project on the Environments and project complexity
Microsoft Dynamics Lifecycle Services
(LCS) portal. This LCS project is hosted When you buy the Finance, Supply Chain Management, and
on one Azure AD tenant, and includes
all the environments used in the Commerce apps from Microsoft, you get two environments. One is the
implementation, testing, and for going
production environment, which is where your company business will
live, such as the production environment.
be conducted. It’s a robust environment, with high availability (HA) and
Typically, an implementation project has
only one LCS project per Azure AD tenant.
disaster recovery (DR) capabilities.
In rare cases, an organization may decide
to have multiple LCS projects. Reasons
to have more than one LCS project or In addition to the production environment, you will receive one sand-
instance include times when a global box environment that’s a standard acceptance testing (tier 2) instance
implementation’s requirements for data
residency, latency, or data volume can’t for the life of the tenant. It’s a nonproduction, multibox instance that
be met by one instance. Or perhaps
can be used for testing, UAT, integration, and training needs.
different business units are implementing
the product separately as independent
applications.
(Because it’s a basic tier 2 instance, it may not be enough for every type
Each LCS project must meet the minimum of testing. For example, performance testing needs a bigger instance
licensing requirement. Currently, you
need to open a support ticket to request to handle larger data volumes and more users.)
additional project, but soon this will be a
self-service activity.
Depending on the project complexity, these environments may not be
sufficient. Figure 9-8 depicts a sample chart comparing the additional
Fig.
9-8
requirements of two projects.

Environment plan sample

Finance and
Environment purpose Operations Dataverse Standard subscription Less complex project More complex project

Testing or training 2 Sandbox

Production Sized Production

Development 1 Development

Build 1 n/a

Golden configuration 1 or 2 n/a

Data migration 2+ n/a

Performance testing 4 or 5 Sandbox

Training 2 or 3 Sandbox

System integration testing 2 or 3 Sandbox

Other/ad hoc n/a Trial or sandbox

Finance and Operations environment Dataverse environment

Implementation Guide: Success by Design: Guide 9, Environment strategy 219

 For less complex projects, you can use the out-of-the-box tier 2 instance
for testing. You still need to consider development environments for de-
velopment activities, with one development environment per developer.
These environments have Microsoft Visual Studio installed.

You should also think about the build environment. Source code devel-
oped in the development environment is synched to Azure DevOps. The
build process will use the build environment, which is a tier 1, to produce
code packages that can be applied to sandbox and production.

Alternatively, you can use Microsoft-hosted build agents to do the

builds instead of having a dedicated build environment. The limitation
of this approach is that you can’t run unit tests, and there is an overall
time limit per month. Assess the limitation and decide whether you
need a dedicated build environment.

Next is the “Golden configuration” environment, for configurations

that are fully tested and those you want to retain and transfer to other
sandboxes and production environments.

You can consider a data-migration environment to move data from

legacy systems, and a performance-testing environment, which is a
higher tier (4 or 5) environment, to conduct performance tests.

When it comes to more complex projects, you may need more envi-
ronments. For example, you may wish to conduct testing in parallel if
multiple teams are developing and testing parallel workstreams.

Production environment and go-live

assessment
When you’re preparing to go live, the first thing you need is access to
your production environment. Project-wise, a production environment
is deployed after all customizations are code-complete, UAT is finished,
the customer has signed off, and there are no blocking issues.

For a Finance, Supply Chain Management, and Commerce project, you

won’t have access to this environment on day one, as Microsoft needs

Implementation Guide: Success by Design: Guide 9, Environment strategy 220

 to complete few prerequisites, including a go-live assessment, before
we can release the production environment.

In a go-live assessment, a solution architect is assigned to review the

Please note that the production project and provide an assessment of potential risks, best practices,
environment is used exclusively for running
and recommendations for a successful go live. In some cases, the solu-
your business operations and shouldn’t
be used for testing. You will be able to tion architect might highlight risk factors and ask for a mitigation plan.
perform the cutover and, if planned,
“mock” the cutover in production.
When the assessment is completed, the solution architect will indicate
that you’re ready to request the production environment in LCS.

General recommendations
▪ Plan for environments early in the project and revisit the plan at
regular intervals.
▪ Define a consistent naming standard for your environments. For
example, a gold environment should have “gold” in the name.
▪ Have a regular schedule to deploy updates and import fresh data
(if needed).
▪ Keep all environments in the same region if your business is in
one region. For example, avoid having a test environment in one
geographical location and production in another.
▪ Deploy environments by using an unnamed account, such as
dynadmin@your_organization_name.com. Assign the environ-
ments to an owner who will be responsible for their status and
maintenance. We strongly recommend using the same dedicated
admin account on all environments.

Assign the Test environments

environments to Consider the number of testing environments you will need through-

an owner who will out the project. A lack of environments may prevent concurrent testing
activities and delay the project.
be responsible for
their status and
maintenance. Training environments
▪ Make sure all users have access with appropriate roles and permis-
sions, which should be the same roles and permissions they will

Implementation Guide: Success by Design: Guide 9, Environment strategy 221

 have in production.
▪ Plan for downtime and have a communication plan to alert users.
(Zero downtime is the eventual goal.)
▪ Ensure all integrations are set up and working, so users can experi-
ence the end-to-end cycle of a business process.

Gold environment
Make sure no transactions happen in a gold environment. There should
be a process to bring tested configurations into the gold environment.

Data-migration environments
▪ Assess whether you need a dedicated environment for data
migration, which is a disruptive task that can’t generally coexist
with other types of test activities. Multiple data-migration envi-
ronments may be needed to avoid conflicts if multiple parties are
migrating data concurrently.
▪ Account for data-migration performance in environment plan-
ning. Depending on the size of the data-migration task, it may be
necessary to use a tier 3 or higher environment to perform da-
ta-migration testing. (You can also use an elevated cloud-hosted
environment.)

Pre-production environments
▪ Assess whether there is a need for a separate environment to test
code or configuration changes before they’re applied to production.
▪ If there will be continuing development of the solution after you
go live, you may need a separate pre-production environment to
support concurrent hotfix and hotfix test activities. (This environ-
ment should have the same code base and data as production, so
a like-for-like comparison can be performed for any new changes
before they’re applied to production.)

Performance-testing environments
▪ Plan a specific environment for performance testing, or you won’t

Implementation Guide: Success by Design: Guide 9, Environment strategy 222

 be able to do performance testing activities in parallel with other
test activities
▪ Avoid doing performance testing in a cloud-hosted environment,
the production environment, and tier 1, tier 2, or tier 3 environments
▪ Use of tier 3 or smaller environments typically won’t provide the
resources required to perform a performance test
▪ Limitations on the number of threads and database transaction
units (DTUs) available in these environments won’t allow for mod-
eling of most concurrent workload profiles
▪ Turn off these environments when not in use to save costs

Developer environments
Ensure each developer has an independent development environment.

Production environment
Raise production environment requests through support, as you don’t
have direct access to this environment.

Product-specific
guidance: Customer
Engagement
Now we’re going to focus on product-specific resources that apply to
Power Platform and customer-engagement apps.

Environments and Dataverse in

Power Platform
For each environment created under an Azure AD tenant, its resources
can only be accessed by users within that tenant. An environment
is also bound to a geographic location, such as the United States.
When you create an app in an environment, that app is routed only to
datacenters in that geographic location. Any items that you create in

Implementation Guide: Success by Design: Guide 9, Environment strategy 223

 that environment, including connections, gateways, and flows using
Power Automate, are also bound to their environment’s location. Every
environment can have up to one Microsoft Dataverse database (Figure
9-9), which provides storage for your apps.

Fig.
9-9 Tenant: Business
Development Development Development

Testing Testing Testing

Production Production Production

Environment

Business US (default) Business Canada Business Australia

Apps Flows Dataverse Apps Flows Dataverse Apps Flows Dataverse

Dynamics 365 products and Power Apps

Microsoft business applications are a set of intelligent solutions that
supply a comprehensive view of an organization’s business. These
solutions include the Dynamics 365 products that are connected by
data and intelligence and supported by the Power Platform.

Microsoft unified CRM and ERP within Dynamics 365 and the Power
Platform brands to make it easier to create apps and share data across
all Dynamics 365 applications. The combination also creates a set of
purpose-built intelligent apps to connect front-office and back-office
functions through shared data. Rich analytical capabilities provide orga-
nizations with deep insights into each functional area of their business.

Power Apps is a suite of apps, services, connectors, and a data plat-

form that provides an app-development environment where you can
quickly build custom business apps that connect to your business data

Implementation Guide: Success by Design: Guide 9, Environment strategy 224

 stored in either the underlying data platform (Dataverse) or online and
on-premises data sources.

Customer-engagement apps such as Sales and Customer Service are

first-party Power Apps developed by Microsoft.

Administration and governance

The Power Platform admin center provides a unified portal for ad-
ministrators to create and manage environments, including sandbox
environments, backing up and restoring environments, and copying
an environment. Admins can also manage settings for Power Apps and
Power Automate.

The Microsoft Power Platform Center of Excellence (CoE) Starter Kit is

a collection of components and tools that are designed to help you
get started with developing a strategy for adopting, administrating,
and supporting the Power Platform, with a focus on Power Apps and
Power Automate.

On-premises to online migration services

The Dynamics 365 migration program enables on-premises customers to
simplify and accelerate their move to the cloud. The migration program
also features access to a dedicated team of migration advisors, no-charge
migration assessments, pricing offers, tools (including a CRM platform
assessment tool), and migration support for qualified customers.

Conclusion
As we have seen throughout this chapter, defining an environment strat-
egy for your organization is a critical step when planning for deployment.
Decisions made about environment strategies are hard to change and
could be very costly later in the program lifecycle. The wrong environment
strategy can create unnecessary data fragmentation and increase the
complexity of your solutions. Early environment planning aligned to
your organization’s digital roadmap is fundamental to success.

Implementation Guide: Success by Design: Guide 9, Environment strategy 225

 Checklist

Organization environment Assess the impact on the overall data estate. Avoid
excessive fragmentation and promote reuse of existing
and tenant strategy integrations and data flows.
Define environment and tenant strategies and obtain
agreement for them at the program level with all
Global deployment
key stakeholders, including business, IT, security, and
compliance. Account for global deployment scenarios; additional
coordination and agreement may be required to meet
Create a strategy that considers the future growth of the
regional requirements and compliance needs.
solution.
Assess the latency to choose the optimal location—
Create an environment strategy that can support the
global deployments are prone to network latency
ALM processes and necessary automations.
related performance issues.
Create an environment strategy that considers the short
and long-term impact on licensing, compliance, appli-
Governance and control
cation design, performance, scalability, maintainability,
and ALM of the solution. Establish governance processes for provisioning, mon-
itoring, managing the lifecycle, and decommissioning
Create a strategy that considers the potential need for
the environments early on.
citizen development scenarios or collaborative develop-
ment with both IT and business users. Ensure the different security personas involved in the
environment management are understood and appro-
Create an environment planning matrix with key consid-
priately assigned.
erations of the pros and cons to help plan and visualize
the impact. Use the CoE Starter Kit to make necessary adjustments
to citizen development use cases as needed because

Environment app strategy they require further governance consideration.

Create an organization environment strategy that

defines the guidelines for onboarding additional apps.

Implementation Guide: Success by Design: Guide 9, Environment strategy 226

 Case study
Timely environment planning
is a critical factor to success
A multinational retailer with stores in Southeast Asia, Australia, and
New Zealand implemented the finance and operations apps like
Dynamics 365 Finance, Supply Chain Management, and Commerce in
a series of phased rollouts spread over eight months.

The management team was not clear that the environments, services,
and default capacities included in the SaaS subscription license, mainly
the production and sandbox environments and hosted build auto-
mation services, were the minimum needed to run the solution on an
ongoing basis, but were insufficient on their own to complete all the
implementation activities required for the moderately complex rollout.

When the FastTrack team was engaged early in the implementation

process, this issue was addressed as part of the Solution Blueprint
Review (SBR). In discussions with the solution architect, the retailer
and partner members of the implementation team gained a better
understanding of the deployment plans and risks posed by their cur-
rent environment strategy (or lack thereof). After the review, the team
agreed to conduct a deep-dive workshop on environment strategy
with FastTrack, which was one of the recommendations based on the
findings from the SBR.

Implementation Guide: Success by Design: Guide 9, Environment strategy 227

 In the workshop, the solution architect explained that environment
planning is not a one-time, set-it-and-forget-it event. Rather, it’s a
continuing activity that requires close collaboration between key
stakeholders. Careful planning aligned with the project schedule
ensures that the team can be productive, conduct important activities
concurrently without environment contention, and achieve milestones
on time, within budget, and with the desired level of quality.

Next, the project team charted out an environment plan to document

the different environments, their purpose, the exact project phase for
which they were needed, and how long they would be needed. For
example, they acquired a high-grade testing environment for three
months just to conduct their performance testing. They also identified
key activities that might affect the environment plan and incorporated
recurring reviews of the environment plan into the project plan. This
simple but important exercise helped the team feel confident that
they understood and accounted for all important implementation
requirements, and that they were prepared to procure the needed
environments on time to reach all milestones.

The development team also worked on configuring DevOps for multi-

ple release branches, with a separate pipeline of sandboxes to facilitate
go-live phases and ongoing support for each one.

Ultimately, this retailer successfully performed the implementation

and rolled out the solution across multiple countries and realized that
timely environment planning was a critical factor in that success.

Implementation Guide: Success by Design: Guide 9, Environment strategy 228

Implement
10 Data management
11 Application lifecycle management
12 Security
13 Business intelligence, reporting, and analytics
14 Testing strategy
15 Extend your solution
16 Integrate with other solutions
17 A performing solution, beyond infrastructure

229
10
Implementation Guide: Success by Design: Guide 10, Data management
Guide
Data
management
230
 Deploy faster and
more efficiently.

Introduction
Data surrounds us every day, like a blanket. Whether
you are posting on your social media, scheduling
a doctor’s appointment, or shopping online, the
information collected is one of the most valuable
assets of any business.
With the right data, organizations can make informed decisions, im-
prove customer engagement, and gather real-time information about
products in the field.

This chapter aims to break down the various functions within data man-
In this chapter, we discuss the many
agement that collectively ensure information is accessible, accurate, and
ways data plays a part in defining a
solution. Data plays a vital role in the relevant for the application’s end users. We focus on the most common
success of every deployment.
discussion points surrounding the lifecycle of data within a project.
You learn about:
• Data governance
• Data architecture Regardless of your role, take the time to consider what is important to
• Data modelling each person interacting with the data. For example, users of a system
• Data migration
• Data integration focus on their data quality, ability to search, data relevance and per-
• Data storage
formance, while architects and administrators are focused on security,
• Data quality
licensing, storage costs, archival, and scalability.

Implementation Guide: Success by Design: Guide 10, Data management 231

 Make sure to always ask these questions:
▪ How is the data used?
▪ What use cases require the data?
▪ Where is the data coming from?
▪ Where is the data stored?
▪ Is there a cost (internal/external) to access the data?
▪ Do users need their data on a mobile phone?
▪ Do customers need access to their data?

We use the Success by Design framework to highlight critical questions

like these that must be asked throughout the project lifecycle to ensure
the desired results for everyone.

Governance Data governance

Architecture Let us start the discussion with data governance before we start un-
packing different principles of data management.
Modeling

While data management is usually seen as dealing with operational

Storage
issues of data, data governance is about taking a high-level strategic
Migration view of policies and procedures that define enterprise data availability,
usability, quality, and security.
Integration
With artificial intelligence (AI) and machine learning (ML) taking center
Quality
stage in most digital transformation projects, and the fact that the
success of these initiatives is highly dependent on data quality, it is
prudent that executives start considering data governance seriously.

You may have heard of instances where different departments in the

same company do not agree on the definitions of business terms and
the values reported. For example, a sales report generated last month
may experience two departments reporting different values. This could
be happening because the two departments are using two different
systems, which is caused by there not being a common definition of
business terms and values that everyone in the company agrees upon.
With advanced data governance in place, the walls that create data
silos effectively crumble as data is shared in a trustworthy way with a
consistent understanding of business concepts and definitions.

Implementation Guide: Success by Design: Guide 10, Data management 232

 Sharing data within the company is crucial for two economic reasons:
growth and distribution. Data is a non rival resource. It is not a material
resource that if one person uses it, others cannot use it. If data is shared
with everyone in your company, people can start becoming self-suffi-
cient and build on top of it increasing the overall value as it becomes
available to all.

Though data governance is a broad topic, for our discussion we want to

focus on key aspects that can help drive a successful implementation.

Consider the following key pillars when analyzing data governance

within your implementation.

Data stewardship
A data steward is a role within an organization responsible for the
management and oversight of an organization’s data assets, with the
goal of providing that data to end users in a usable, safe, and trusted
way. By using established data governance processes, a data steward
ensures the fitness of data elements, both the content and metadata,
and they have a specialist role that incorporates processes, policies,
guidelines, and responsibilities for administering an organization’s
entire data in compliance with policy and/or regulatory obligations.

Data quality
The quality of the data cannot be thought of as a second or third pro-
cess. Data quality is at the front of data governance policies. It should
be thought of as high quality and fit for the purpose of whatever the
intended use is. Driving data’s accuracy and completeness across the
organization is typically managed by dedicated teams who may use a
variety of tools to scrub the content for its accuracy. Although these tools
aid the process more and more, this is still typically a human responsibility.

The users that are most familiar with their data should typically be the
gatekeepers for cleansing, including standardization and adherence to
the policies outlined by the data steward.

Implementation Guide: Success by Design: Guide 10, Data management 233

 Master data management
Most enterprises keep a primary set of data used across the organization
to supplement systems and reporting. The data elements are typically
stored within a master data management (MDM) solution. This includes
customers, accounts, products, and other areas according to business
needs. Typically there are rules of engagement that internal teams
requesting access to the master data must follow, which is documented
within the data governance plan and managed by the data steward.

Proper use cases

A data governance plan cannot be set up without understanding the
proper use cases. Every line of business (LOB) has a level of ownership
of their data along with the use cases that drive the solution design. The
same holds true for data governance. The use cases aid in naming the
primary data elements the organization is taking proactive control of.

For example, one retailer wants to better target its customers through
email campaigns, which in the past failed to deliver expected results
due to incorrect contact information being captured. While defining
this use case, the company also set up the right data governance
principles that ensure what data quality is required for email cam-
paigns. They were challenged to define “good” data to satisfy the use
case. This simple use case uncovered other issues. The company found
that online customers, who buy as guests, could enter any value in the
email field and there was no validation. This led to data stewards and
LOB process owners setting up new validation processes.

Without data
Without data governance in place, organizations struggle to control
governance in corporate assets needed to ensure data quality. During the requirements
place, organizations gathering stage of your implementation, start paying particular atten-
struggle to control tion to the availability of data required for your solution. Early discussion
and identification goes a long way in defining your use cases.
corporate assets
needed to ensure For an example of a proper use case, see “A Show-Don’t-Tell Approach
data quality. to Data Governance.”

234
 Governance
Data architecture
Architecture
After data governance is understood within the organization, the
Modeling next step is to look at data architecture and the different types of
enterprise data.
Storage

According to Data Management Book of Knowledge 2 (DMBoK 2):

Migration
“Data architecture defines the blueprint for managing data assets by
Integration aligning with organizational strategy to establish strategic data re-
quirements and designs to meet these requirements.”
Quality

Data architecture is a high-level concept, traditionally part of enter-

prise architecture. In simpler words, data architecture describes how
data can be supplied to meet a business’s strategic objectives.

The right data architecture is central to the success of your business’s

data strategy. It focuses on where and how you use your data across
various sources and what technologies are available to harness. It is
important to know where the data is coming from, if it is accurate, and
if it provides the insights needed by the business within the solution.

If your organization has an enterprise-level data architecture in place,

make sure to research policies, procedures, and restrictions that are in
place for accessing the information you need.

For example, an organization may have a number of disparate ap-

plications to support different business units, and one of the legacy
applications (not Dynamics 365) is traditionally used as a central
repository for customer master data. The enterprise data strategy may
be to support a master version of the customer data in order to supply
better analytics and visibility of customer interactions across all lines
of business. However, interactions with a customer, like sales oppor-
tunities, sales transactions, support cases, or general activities, may be
managed within your Dynamics 365 solution.

When there is no enterprise data architecture, if you simply draw out

all the existing business applications, their interfaces, and connections,

Implementation Guide: Success by Design: Guide 10, Data management 235

 you come up with a spaghetti diagram. That is a good start to getting
an overview of your integration and data architecture. In fact, the
resulting data flow diagram is an important input into defining your
integration strategy.

Read more about integrations in Chapter 16, “Integrate with

other solutions.”

The key point to data architecture is to create a holistic view of the data
repositories, their relationships with each other, and ownership. Failure to
implement data architecture best practices often leads to misalignment
issues, such as a lack of cohesion between business and technical teams.

During your design and analysis phases, it is important to capture

the data owners, systems, and conceptual flow between systems. This
information helps to meet the integration requirements, as well as
providing a reference for the broader project team.

When you prepare your data architecture, use the questions below to
guide you.
▪ How do you use your data?
▪ Where do you store your data?
▪ How you manage your data and integrate it across your lines of
business?
▪ What are the privacy and security aspects of the data?

Before you can start working with data, it is important to know what a
typical enterprise data is made up of, as illustrated in Figure 10-1.

Implementation Guide: Success by Design: Guide 10, Data management 236

 Fig.
10-1

Azure APIs

Telephony Interactions

Corporate Line of
application business
End
users Company’s
Contacts Activities Master relationships
A accounts
Cases Other

Microsoft
Microsoft Master data integration
USD Dynamics
365 Master Employees
B

Email Exchange
Power server-sync
Applications Master Products
C
External
knowledge Knowledge
Customers base base
self-service Master Product
articles
D alerts

SharePoint SharePoint
document
library

Financials
Corporate
accounting
application

Types of enterprise data

There are several distinct types of physical data to be aware of. This
type of data typically falls into one of four categories: master data,
configuration data, transactional data, and inferred data.

Master data
Master data is your source of common business data and represents a

Implementation Guide: Success by Design: Guide 10, Data management 237

non transactional, consistent, and uniform set of tables. The details are
maintained and controlled for accuracy. For example:
▪ Company data
▪ Contact data
▪ Account holdings
▪ Customers
▪ Vendors
▪ Products

Configuration data
Configuration data is data about different setups needed to prepare
your production environment for operational use. Besides importing
master data like customers and products, you need to import the
configurations data. For example:
▪ Currencies
▪ Tax codes
▪ Modes of payment
▪ Address (countries, states, postal codes, etc.)

Transactional data
This type of data is generally high in volume due to the nature of its use.
Transactional data typically refers to events or activity related to the
master data tables. The information is either created automatically or
recorded by a user. The actual information could be a statement in fact
(like in banking), or it could be a user interpretation, like the sentiment of
a customer during a recent sales call. Here are a few other examples:
▪ Communication history
▪ Banking transactions
▪ IoT transactions
▪ ERP transactions (purchase orders, receipts, production orders, etc.)

Inferred data
Inferred data is information not collected by the business or users.
Typically, this information is automatically generated based on other
external factors, which adds a level of uncertainty. For example:
▪ Social media posts
▪ Credit score
▪ Segmentation

238
 Governance
Data modeling
Architecture
With a base understanding of the data governance and data archi-
Modeling tecture, we can now focus our attention on how we plan to store the
information we collect. Data modeling should always be completed
Storage before any configuration begins. Let us start with a basic understand-
ing of data modeling, recommended practices, and how it is related to
Migration
a project.
Integration
According to the DMBoK 2, “The process of discovering, analyzing,
Quality representing and communicating data requirements in a precise form
is called the data model.” While data architecture is at an elevated level
and involves a data architect looking at the business requirements
broadly, data modeling is at a lower level and deals with defining and
designing the data flows in precise detail.

Data architecture and data modeling need to work together when de-
signing a solution for a specific business problem. For example, if your
organization wants to implement an e-commerce solution, you cannot
do that unless somebody knows and has defined the existing data ar-
chitecture and data models, different integrations in play, existing data
imports and exports, how customer and sales data is currently flowing,
what kind of design patterns can be supported, and which platform is a
better fit into the existing architecture.

Likewise, if your organization wants to implement a data lake, some-

body needs to answer questions like what the existing data flows are
both in and out of the business applications, what is their format, and
what are the existing data warehouses we need to migrate to data lake?

It is also recommended that the data model conforms to the Common

Data Model standard, which does not deviate from or repurpose
existing out-of-the-box tables, to ensure cross application compatibil-
ity and future readiness. Common Data Model provides a shared data
language for business and analytical applications to use and makes it
possible for data and its meaning to be shared across applications and
business processes such as Microsoft Power Apps, Power BI, Dynamics

Implementation Guide: Success by Design: Guide 10, Data management 239

 365, and Azure. For more information, refer to Common Data Model.

What is data modeling?

A data model is a visual model showing how data flows through your
Fig.
10-2 system and how different tables relate to each other. Data models
define the relationship types
between tables and abstract a da-
1 Technical contact 0
Customer tabase to a visual representation
Customer
contact that is easy to understand.
1 Account manager 0
1
If your project requires creation
To technical
contact of new tables to support storing
data for some specific business
0
processes, it is recommended that
Monthly 1 Attached 1 Market you create a data model to define
report email report PDF
and understand the schema and
data flow. But if yours is largely
a vanilla implementation or has
minimal customizations, you may
Fig.
10-3 not need to do this. It is recom-
mended to discuss with your
Supplier Delivery Order detail
delivery system integrator and solution
Delivery ID Delivery ID architect to see what works best
Delivery ID
Delivery date Delivery date
Delivery date for your case.
Supplier ID Supplier ID
Supplier ID

There are multiple types and

Product Order detail Order

standards for data modeling,
including Unified Modeling
Product ID Order detaill ID Order ID
Language (UML), IDEF1X, logical
Supplier ID Product ID Order date
Order ID Headquarters ID data model (Figure 10-2), and
Product quantity physical data model (Figure
10-3). A common type of physical
data model is an entity relation-
ship diagram (ERD).
Branch Headquarters

Branch ID Headquarters ID
Branch ID

Implementation Guide: Success by Design: Guide 10, Data management 240

 Governance
Data storage
Architecture
Next, we need to talk about data storage and what you can do to
Modeling manage the growth and costs involved. The reality is no matter where
your data is stored, there is always a cost involved. Whether you are
Storage managing data on-premise or in Microsoft’s cloud, you should be
aware that storing and moving your data can impact the overall cost of
Migration
the solution.
Integration
There are several factors to consider when defining a data storage strategy.
Quality

Forecasting capacity
The capacity your environment consumes should be part of the over-
all environment strategy. Every environment should be assessed for
storage requirements based on usage. For example, the development
environment does not require as much storage as a production envi-
ronment to operate.

First, you need to calculate the approximate volume and size of the
data to properly come up with an estimate. You can gather this detail
Fig. from your existing data sources. You should come up with a size either
10-4
in megabytes (MB) or gigabytes (GB) needed in production.
Based on the estimated size for
% of
Environment Use production, you then need to
Product
allocate a percentage for each en-
Production Contains all data required for release 100 vironment. Figure 10-4 provides
an example.
Contains a sampling of data for
Training 15
training
The best practice is to build a data
storage forecast for a minimum of
QA Contains a sampling of data for testing 15
three years, including an average
increased annual volume. We
SIT Contains a sampling of data for testing 15 recommend that when designing
and configuring your different en-
Contains limited data only for vironments you discuss with your
DEV 5
development needs
system integrator, and Microsoft,

Implementation Guide: Success by Design: Guide 10, Data management 241

 the amount of storage you anticipate you need and when you need
it by. The business needs to track the data growth to make sure the
overall size stays within your allotted storage entitlements.

The reality is no Environments should be managed with a well-planned environment

strategy. For example, a developer should not be able to simply go
matter where your ahead and create new environments on the fly. Doing so could create
data is stored, there opportunities to go over entitlements, which leads to enforcement,
is always a cost which in turn could limit your ability to deliver on the application

involved. lifecycle management (ALM) strategies.

Read more about environment and ALM strategies in Chapter 9,

“Environment strategy,” and Chapter 11, “Application lifecycle manage-
ment,” respectively.

Governance
Configuration data and
data migration
Architecture

Modeling
Once you have a good understanding of your business requirements,
Storage
project scope, and have set up proper data governance, data ar-
chitecture, and data models, you need to start the preparations for
Migration
importing data to prepare for your solution to be ready for go live(s).
Integration
There are two types of data we are talking about: Configuration data
Quality
and migrated data.

Managing these two data imports is often the most challenging part
of Dynamics 365 implementations. We often see customers underes-
timating the efforts required and projects suffering as issues surface
later. In this section, we talk about plans for configurations, plans for
data migration, different environments, identifying data sources, ETL
(Extract, Transform, Load) or ELT (Extract, Load, Transform) processes,
and finally, staffing the right resources who can manage these efforts.

Implementation Guide: Success by Design: Guide 10, Data management 242

 Configuration data
Configuration data is data about different setups needed to prepare
your production environment for operational use, for example, curren-
cies and tax codes. Besides importing master data like customers and
products, you need to import configuration data. Managing configura-
tions is an activity that needs to happen throughout the lifecycle of the
implementation. As the team understands the business requirements
and designs required, they understand different setups and parameters
needed to enable the business processes required by the business.

Since the Dynamics 365 offerings are highly configurable, you need
tight control over the different configurations required to be set. With
so many development and testing environments and different team
members, there is a high chance of incorrect configurations being set,
leading to errors in processes. In the case of Dynamics 365 Finance,
Supply Chain Management, and Commerce, we recommend setting
up a dedicated golden configuration environment to capture and
maintain the correct configuration data. A gold environment can
help ease the pain in tracking, storing, updating, and controlling the
configuration data. Gold environment settings should be kept pristine,
have tight access control, and no one should be allowed to create any
transactions. You can use the golden configuration database to restore
other environments for testing, data migration, and even initial go live.

Use a configuration plan

Next, it is important to discuss the idea of using a configuration plan. This
is especially important in a phased rollout scenario when you need to plan
to setup a number of environments, legal entities, or business units.

You can use a golden configuration environment to restore into pro-

duction for the first phase of go live, but for every subsequent phase
of go lives you do not have that luxury. The double whammy of a large
number of setups and slight but crucial variations of each go live can
be highly daunting, so doing it manually is fraught with danger. We
highly recommend that organizations use a configuration plan.

Implementation Guide: Success by Design: Guide 10, Data management 243

Configuration plans are a structured list of different master and con-
figuration data that you want to import for go live and any subsequent
go lives. It is organized in the sequence required by the system and
manages the dependencies between entities. As an example, if you
need to enable purchase orders you need vendors first, if you need
projects, you need accounts first, etc.

This plan also represents the functional scope for the solution. As
described in Chapter 7, “Process-focused solution,” you start with your
process definition and scope for the implementation. This process re-
quires enabling functionality to be enabled in the application, and this
functionality requires configuration and master data and eventually
some transactional data in connection to the first use of the applica-
tion, like some opening balances. The configuration plan represents
this scope in terms of system enablement and the correlation between
all of them.

Another benefit of the configuration plan is to provide early visibility

of how those configurations and master data play a role between the
different business units (legal entities) or apps that share the same
database foundation. You can determine if that data is natively shared,
or if it is unique for an entity. Having this visibility prevents you from
having redundant data replication and improves the speed to enable
the solution.

For example, flagging the setups that only need to be imported once
as they are shared across business units or legal entities versus set-
ups that need to be imported for each phase as they are not shared.
A configuration plan can help you be organized and consider the
requirements for all the subsequent business units or legal entities.
Another example, in the case of Dynamics 365 Finance, Supply Chain
Management, and Commerce, can be whether you want to use data
sharing features to copy the configurations across companies instead
of manually importing those configurations into each company.

The configuration plan is your configuration playbook that shows you

scope, dependencies, and data redundancy and on top of that, it can
help you to track the progress of enabling the solution in terms of

244
 configurations and master data required to be loaded.

Note that managing a configuration plan is an iterative process. As

you make progress in the project and discover new requirements, and
thus new setups, you should come back to the plan and update it. The
list is your one-stop shop to tell you what you are going to do for each
go live for each configuration, what data is already available, and what
data needs to be imported. Just like your golden configuration envi-
ronment, it’s crucial to keep your configuration plan up to date.

A proper well-maintained configuration plan makes sure the team

considers all the important configurations for all the planned rollouts
and does not miss them causing unwanted errors in production after
go live.

This plan also helps the team be prepared for cutover and calculates
the required downtime window to import all necessary data. The
team should test the configuration plan a number of times, note the
time it takes for all the configurations, and confirm that the total time
required is within the cutover window.

We strongly recommend that teams put some thought in and create

a configuration plan to manage all their imports as the benefits of
reduced errors and faster imports far outweigh the efforts needed to
create and manage the plan.

Since the configuration plan is essentially a list, many customers

prefer to use Excel sheets as most organizations are familiar with the
program. There are also customers who prefer to maintain the list in
DevOps. For example, you can have one work item for currencies and
another work item for tax codes with exported data entities attached
to the work items. It does not matter which option you choose, as long
as your team members can access and update the list when necessary.

A sample plan for managing configuration imports for Dynamics 365

Finance, Supply Chain Management, and Commerce can be down-
loaded from Sample Configuration Plan using Data Entities.

Implementation Guide: Success by Design: Guide 10, Data management 245

 Data migration
The second type of data we are dealing with is migrated data. Data
migration, in simple terms, is the process of moving data from one
data model to a different data model for future use. In the case of
Dynamics 365, data migration refers to moving data from a legacy
system to a Dynamics 365 application. Whether your legacy system
was another application or disparate Excel spreadsheets, you may have
been capturing data to run your business and you need that data when
you move to your new Dynamics 365 application.

Migrated data is either master data such as customers, products, and

vendors, or open transactions such as open sales orders, open pur-
chase orders, on hand stock, and open balances.

Migration planning
When deploying a solution, data is particularly important. Different
departments within the organization can find it challenging to support
certain activities when there is no data in the system.

Typically, when replacing another business application, relevant data is

Fig.
10-5 migrated into the new system during the deployment so the users can
see relevant business information when they start
using the application. It is especially important to

Discover manage the timeline of data migration activities in

the project plan so you can allocate time and people
to the activities. For example, you need to track
activities like migration planning, identifying data
Verify Analyze
Data sources, data mapping, ETL process, importing data
migration for UAT, and importing data for go live.
lifecycle
Explore the parts of the data migration lifecycle in
Figure 10-5. Take time to discover the mandatory
Implement Plan data required by your solution and analyze the data
types and systems that source the data to be imported.

Migrating data is a time-consuming and expensive

Implementation Guide: Success by Design: Guide 10, Data management 246

process that often lacks full visibility and understanding of the use
cases. The use cases help define how the data is used within your
solutions processes.

When building a plan, keep in mind that data migration activities can be a
disruptive task and should not co-exist with other testing activities, so it is
advised to procure a dedicated high tier data migration environment.

It is recommended to implement at least one round of system integra-

tion testing (SIT) and user acceptance testing (UAT) with migrated data.
During testing, developers and users are able to verify the migrated
data and ensure it meets the requirements.

Defining a migration strategy

Data migration is a major undertaking and is often run as a parallel
activity to the development and testing activities and spans multiple
phases. Several decisions need to be made regarding the scope of
the migration, environments, source databases, and the tools used to
extract and transform.

There needs to be clear business requirements defining the data migration

scope, timelines, and expected cutover plans before a project kickoff.
You should raise concerns immediately if these details are not clarified.

▪ Identify the source for where the data needs to be pulled

and migrated.
▪ Define the environments that are needed (staging environment).
▪ Is there a mapping document for the tables/fields between old
and new systems?
▪ Do you have the right analysts and developers resourced?

Multiple sources data

The source data can come in all shapes and sizes. From a salesperson’s
Excel spreadsheet to a legacy systems SQL database. Knowing all the
sources allows time to analyze the options for extracting the information.

247
 ▪ SQL database
▪ Third party database
▪ External webservices
▪ Access
▪ Flat files/Excel

In advance and during the project, it is recommended that the LOB

begin the process of identifying, consolidating, deduping, scrubbing,
and cleansing the data. The goal is to limit the amount of transforma-
tion needed during migration. The more logic required to transform,
the slower the migration takes to load.

Environments
Another missed factor is sizing of the import and staging databases
required for running migration tooling and cleansing of data. You need
to make sure environments used are sized appropriately to handle the
volumes of data in scope. It is recommended practice to have all data-
bases and environments running under the same location and region.
During environment planning, the appropriate environments should
be sourced for data migration.

Data mapping
The process of data mapping can start once the solutions for data
modeling have been defined. You should be organized to keep the
process systematic and simple.

▪ Use an Excel spreadsheet with each tab representing a system table.

▪ Capture all the fields being mapped together along with any
transformation requirements.
▪ Pay close attention to the target system you are loading data into.
For example, data types should match between systems or trans-
formed accordingly.

Extract, transform, load (ETL) tooling

Once you have identified the sources, set up the environments, and

Implementation Guide: Success by Design: Guide 10, Data management 248

 Fig.
documented the data mappings, then the logical next step is to ex-
10-6
tract the data from the source, transform the data for target, and
load the data into the target, as illustrated in Figure 10-6. You can
Extract
use many types of tools and approaches when migrating data to your
Dynamics 365 solution. Some of the very standard options include:
Source ▪ Data import/export wizards
▪ Azure Data Factory or Azure Synapse Analytics
database ▪ SQL Server Integration Services (SSIS)
▪ Third-party integration products

Refer to the product-specific section later in this chapter for a list of

some of the most common options.

nsform
Tra Roles and responsibilities
Customers and partners should staff the project team with the right
Staging resources who understand data and the tools in Dynamics 365.

database Examples of some important roles and responsibilities in Figure 10-7.

Fig.
10-7

Role Responsibility

Assist in designing, planning, and managing the data

Load migration process.
Data migration
analyst Work with subject matter experts and project team to
identify, define, collate, document, and communicate the
Target data migration requirements.

database Maintain data and manage it according to data properties

as required by administration.
Data steward
Coordinate with stakeholders and provide all definitions
for data.

Design and develop the environments required for

migration.
Data migration
architect/ Develop data packages as required to move the data
developer between systems.

Provide initial testing and validation.

Implementation Guide: Success by Design: Guide 10, Data management 249

 Data integration
Integration is the connecting of one or more parts or components of a
system to create a more unified experience or to ensure a more con-
sistent outcome of a process. Integration allows leveraging of existing
services both internal and external without having to rebuild or mi-
grate existing functionality.

Data integration is done to bring data into the system or out to other
systems. Typically, this happens through an event or in a batch on a
schedule. For example, when a record is created or updated it would
be event driven and the nightly scheduled transmission of data would
be a batch.

Refer to Chapter 16, “Integrate with other solutions” for more details.

Data quality
Once data is migrated and integrated to your solution, it is critical that
information remains accurate, complete, reliable, and, most importantly,
up to date.

You can use various techniques to manage the quality of the data,
including data profiling, data cleansing, and data validation. The most
important aspect of data quality to understand is that data quality is
the responsibility of everybody in an organization. We see many cus-
tomers who overestimate the quality of their data and underestimate
the effort it takes to get it into shape. Keeping data quality top notch
requires following all the principles we highlighted in this chapter and
equally strong leadership to drive the habit of managing data on an
ongoing basis.

Make your data quality a habit!

Implementation Guide: Success by Design: Guide 10, Data management 250

Conclusion
Data has taken center stage in all enterprise transformation projects.
Businesses want to be data driven. Having the right data flowing in
your business means you have set up a well-functioning business in
which you have the most up-to-date and accurate information about
your customers and products and can provide a superior experience to
your staff and customers.

The benefits of data do not just stop there. Having the right data
means you can start leveraging ML and AI today to predict what your
customers need tomorrow.

Having proper data governance, data integration, data security, and

data quality are not just best practices; they are necessary for your
business’s survival because every business wants to be data-driven
and compete for customers’ attention. These concepts can help lay the
foundation on which you can start tapping into data to transform your
processes, gain insights, and predict customer behavior.

Data management covers so many different areas, it is easy to get

overwhelmed. By starting with the right people, process, and data, you
cannot go wrong.

Product-specific guidance
Up to this point in the chapter, our data related guidance has applied
to Dynamics 365 Finance, Supply Chain Management, Commerce, as
well as Customer Engagement application projects. While both appli-
cations live in the Dynamics 365 ecosystem and customers frequently
adopt both systems, often simultaneously, there are differences
between the two. This can mean differences in how each application
should manage its data.

In this section we highlight some of the different approaches for both

applications.

251
 Customer Engagement
This section includes a number of recommendations and resources
provided in Customer Engagement to help manage modeling, storage,
migration, and archival.

Data modeling
Data modeling is a science, and there are data modeling professionals
and established standards for data modeling. To be effective with
Dynamics 365 data modeling, you do not have to be a professional
data modeler or use any special tools. Popular tools like Microsoft Visio
can be used to quickly create a basic ERD diagram that visualizes the
relationships and flow of data between tables. In this section, we dis-
cuss some general best practices for data modeling for Dynamics 365
deployments.

▪ Data models should be updated continuously during a de-

ployment. It is common for a data model to be designed at the
beginning of a project, but it is very important that it does not
stop there. As you go through the deployment, new fields and ta-
bles are added—it is important to capture these in the data model
to make it a living data model. Recommend to customers that they
continue to update it in order to enhance the system.
▪ Do not include every table. Some core tables, such as activities,
notes, and users (record owners), are related to nearly every entity
in Dynamics 365. If you include every relationship with these tables
in your data model, the result is unreadable. The best practice is to
include only the primary tables used in your configuration in your
data model diagram and include only custom relationships with
the user and activity tables to maximize readability.
▪ Data models should include tables outside of Dataverse. If you are in-
tegrating with other systems via Dataverse data connectors or virtual
tables, or if data flows outside of the Dataverse via an integration, this
data should also be represented in your data model diagram.
▪ Start simple with the standard tables, then add custom entity
relationships to your data model.

Implementation Guide: Success by Design: Guide 10, Data management 252

 ▪ User experience should influence your data model. Sometimes it is
easy to over normalize your data, but the process could make the
application more cumbersome to use.

Start with what you need now but design the data model in a way that
supports what you are going to be doing in the future. For example,
if you know that down the road you need to store additional details
about sales territories, using a text field for territory now makes it more
difficult to implement than if you use the territory entity relationship.
Plan for what is coming.

Data storage
This section provides product-specific guidance you can use while
implementing or maintaining your solution.

Storage capacity
The storage capacity is a standard calculation within the Power
Platform that is easily managed by the system administrator. The Power
Platform admin center is the tool you should use to maintain visibility
of storage and consumption. Within the Power Platform admin center,
go to Resources > Capacity > Dataverse for more details about your
capacity entitlements, as shown in Figure 10-8. You can access this by
going to the Power Platform admin center.
Fig.
10-8

Implementation Guide: Success by Design: Guide 10, Data management 253

Storage entitlements
Dataverse capacity (database, file, log, and add-ons) is pooled across
the tenant and shared among all environments and workloads. The
first subscription provides a one-time default capacity entitlement for
the tenant. Learn more about the Dataverse storage capacity.

Storage segmentation
To better understand how capacity is calculated within Customer
Engagement, the following provides the breakout based on storage
type and database tables.

Dataverse database: All database tables are counted for your data-
base except for logs and files below.

Dataverse files: The following tables store data in file and database
storage:
▪ Attachment
▪ AnnotationBase
▪ Any custom or out-of-the-box entity that has fields of datatype file
or image (full size)
▪ Any table that is used by one or more installed Insights applications
and ends in “-analytics”
▪ AsyncOperation
▪ Solution
▪ WebResourceBase
▪ RibbonClientMetadataBase

Dataverse logs: The following tables are used:

▪ AuditBase
▪ PlugInTraceLogBase

Data migration
This section provides product-specific guidance you can use while
implementing or maintaining your solution.

Approaches and tools

You can use many types of tools when migrating data to your

254
 Dynamics 365 solution. The following is not an exhaustive list, but it
includes some of the most common options.

Do not migrate data  It is often assumed that large amounts of data

must be migrated into Dynamics 365 to provide a complete view of the
customer. However, just because you need to see data in Dynamics 365
does not mean that the data must live in the Dynamics 365 Dataverse
database. Considerations should be taken for the value of the data
weighed against the quality of data, the volume of the data, the
amount of time required to load the data, and potential storage costs
before migrating data. Not physically migrating data can frequently be
replaced with embedded Power BI reports, virtual entities, embedded
canvas apps with connectors to the source data, and other options. For
example, the email auto-capture option in Dynamics 365 Sales Insights
displays activities to salespeople from their Exchange inbox rather
than having to load the data to Dynamics 365. This can be a viable
option to show activity data rather than loading large amounts of data.
Virtually displaying data can also significantly reduce the time and cost
of deployment, as you do not have to take time to load data and test
your data migration.

Power Apps Excel add-in  This add-in can be used to open entities
directly in Excel and create and update records. Records are updated
or created directly in Dataverse. Not all entities support updates from
Excel, and lookup fields must be manually edited to correctly match.

Get data from Excel At make.powerapps.com, you can choose “Get

data” when viewing an entity and import data from an Excel or .csv
file. This option gives additional control over the mapping of data than
the Excel add-in, and lets you import data from one file into multiple
entities. To help ensure that the format of fields and mapping is cor-
rect, you can export a template from a Dataverse entity, populate your
data, then import it. Alternatively, you can import a custom file and
provide mapping. Lookup fields must include the primary key or entity
alternate key values to match correctly.

Power Platform dataflows At make.powerapps.com, you can select

“Dataflows” from under the Data menu and configure an import from

Implementation Guide: Success by Design: Guide 10, Data management 255

several data sources. Data sources includes common external services
file sources as well as generic web APIs. Data from these data sources
can be transformed prior to import using Power Query.

Legacy Dynamics 365 data import utility  You can import data to
Dynamics 365 entities from CSV, .xls, .xml, and .zip. While the Dataverse
API Get Data option is recommended for most flat file imports, the
legacy data import option has several unique capabilities that might
be useful in some cases.

▪ Legacy data import can be used to create new entities and fields
and option set values in Dynamics 365. While this is convenient,
the best practice is to add these items from a solution rather than
from data import.
▪ Legacy data import can import multiple files at the same time
when multiple files are added a zip file.
▪ Legacy data import can resolve lookup fields using values not
contained in the primary or alternate keys.

Dynamics 365 Configuration Data Migration Utility  This is a

tool provided by Microsoft that is designed to move data between
Dynamics 365 environments. If you are migrating data from sandbox
environments to production, this tool is very useful as it allows you to
keep the record IDs consistent between environments. Consider this
when you have entities like country and specific records are referenced
in workflows or flows, if the ID changes between environments, the
process does not work. Configuration Data Migration Utility was up-
dated in early 2020, so you can now provide filters to tailor the records
included in the Configuration Data Migration Utility data packages.

Extract, transform, and load (ETL) software  For more complex mi-
grations, such as migrating an entire legacy CRM environment’s data,
manual import from flat files is not efficient and can be error prone.
For more complex migrations, commercially available ETL tools like
SSIS, Azure Data Factory, or a number of third-party ISPs offer services
and tools that can be used to create a migration data transformation
specification (DTS) that can migrate legacy data to Dynamics 365. This
approach can have the following benefits:

256
 ▪ Reusability of migration, allowing the data migration developer to
test and refine the migration multiple times before go live.
▪ Delta sync loads when moving the data from an in-production
system to a new system. Users still use the legacy system until the
switchover to Dynamics 365 happens. If the data is loaded into
production several days before go live, there is additional data cre-
ated in the legacy system after the initial load is finished. ETL tools
allow the data loaded to be filtered to only include data changed
since the last load, ensuring that the migrated data is up to date
when the users start using the new system.
▪ Consistency with data integration. Sometimes data is migrated in
an initial load, and then updated via an ongoing integration. In
these cases, it is optimal to use the same tooling that you use for
the ongoing integration to also migrate the initial data load, and
in those cases the same DTS and field mappings may be used for
the data migration as well.
▪ More complex data transformation. When moving data via flat
files, you can choose what data is mapped to the new system, but
if you wish to change or transform the data, it must be done man-
ually in the source flat files. With an ETL based migration, you have
flexibility to transform the data during the migration. For example,
say there are five different types of accounts in the source data,
but you wish to consolidate them to three, ETL tools allow for this
transformation in the data translation specification.
▪ Updates and upserts. The flat file and Excel imports support
updating records that match record keys, but sometimes your
matching logic is much more complex. Say you want to both
insert and update records (upsert). This is complex and tedious
to do with the out-of-the-box data import options, as you don’t
always have the record IDs. ETL tools allow the data migration
developer to define record matching logic and update, insert, or
upsert records based on any matching criteria. This is also helpful
for making sure that duplicate records are not being created in the
target.
▪ More flexibility in mapping lookup fields. Lookups to other en-
tities can be challenging for data imports, especially fields like
“customer” that are polymorphic, or fields like “owner” that have
special properties. If your legacy data has owning users that have

Implementation Guide: Success by Design: Guide 10, Data management 257

 a different format, the ETL tool can transform the values into a
format that can be imported.
▪ Many-to-many (N:N) relationships. ETL based imports can easily im-
port data to N:N relationships and entities like marketing list member,
which is not available from some of the flat file import options.
▪ Faster import of large data sets. Good ETL tools can use multi-
threading to import data quickly into the common data service.

Power Automate and Azure Logic apps  These can be used to im-
port data and provide connectors to over 300 services. These options
provide many of the same benefits as an ETL tool. In addition, since
Power Automate includes more than 300 connectors to many leading
services and application platforms, Power Automate can be a great
option to migrate data from these services. Although we have these
options, they are not really meant to perform large data migration
jobs, they could be considered a great option for low-volume or low-
rate delta migrations that are not too demanding on throughput.

Custom API migrations  These can be used to migrate data to

Dynamics 365. This option gives you more control over the migration
and does not require any extra tools, but it requires that you have a
developer to develop and support the data migration.

Data archival and retention

The challenge for most organizations is not necessarily defining a
strategy but rather implementing one. Depending on your solution,
there is most often a need to be a technical solution required to satisfy
the strategy of moving the data.

As mentioned earlier, over time the volume of your data grows as well
as the cost. Implementing a data archival and retention strategy allows
regular shifting of data from one storage location to another. The fact
is, over time your data grows and the SQL server comes under strain.
This impacts SQL server’s performance and thus degrades user experi-
ence. Even though Dynamics 365 is a SaaS application, and you do not
bother with managing SQL server, it is a clever idea to have periodic
data maintenance routines in place to archive and delete unwanted
data to keep database nimble.

Implementation Guide: Success by Design: Guide 10, Data management 258

 For example, you can set up data export schedules to replicate data to
Azure Data Lake, which is comparatively cheaper than Dataverse. From
a best practice’s perspective, you do not want to keep old historical
data in Dynamics 365, data that the business does not need for day-
to-day operations. Once data is moved to the data lake, you can setup
Azure Data Factory to create data flows, transform your data, and
run analysis and you can use Power BI to create business reports and
produce analytics.

Figure 10-9 shows a modern data warehouse architecture by using

Azure Data Lake Storage.
Fig.
10-9

Other services or Ingest Model and serve Visualize

customer business apps Azure Synapse Analytics
(structured) (data warehouse)
Up to14x faster and costs 94% less
than other cloud providers
Media
(unstructured) Prep and train
Azure Data Factory Power BI
Logs and files
Code-free data transformation and integration Azure Databricks Leader in the magic quadrant for business
from 90+ data integration connectors intelligence and analytics platforms
(unstructured)

Store

Azure Data Lake Storage

Microsoft High-performance low-cost data lake
Dynamics 365 available in 54 Azure regions

For more information, read Ingest Microsoft Dataverse data with

Azure Data Factory and Analyze Microsoft Dataverse exported to
Azure Data Lake Storage Gen2 data with Power BI.

More information about reporting strategy can be found in Chapter 13,

“Business intelligence, reporting, and analytics.”

Operations
Dynamics 365 Finance, Supply Chain Management, and Commerce
have a number of data management tool sets and data maintenance
cleanup schedules, which we describe in more detail here.

Implementation Guide: Success by Design: Guide 10, Data management 259

 Data maintenance
When it comes to data storage, an important factor to consider is
that not all data needs storing. You have lots of logs and staging data
generated that can be safely truncated after a certain time.

This is where you need to set up a plan for data maintenance.

In Dynamics 365 Finance, Supply Chain Management, and Commerce,

cleanup routines are available across various modules, which can keep
the tables tidy. It is important to note that these cleanup routines
should be completed only after detailed analysis and confirmation
Fig.
10-10
from the business that this data is no longer needed. Always test each
LCS
lifecycle routine in a test environment prior to running it in production. This
services article provides an overview on what is available today: Cleanup rou-
tines in Dynamics 365 for Finance and Operations.

Data
management Best practices with handling PII (Personally identifiable
import/ information) data
export ▪ Avoid storing unnecessary PII data in Dynamics 365 Finance and
Operations apps if possible.
• Copy legal entity
• Templates
▪ Identify, tag, and classify PII data that you need to store.
▪ Dynamics 365 Finance, Supply Chain Management, and

Database Commerce uses Azure SQL database that allows data encryption at
operations rest and transport.
Backup / restore ▪ X++ APIs/patterns to encrypt and decrypt data at columns level
/ refresh / point in
time restore for added security.
▪ Build Edge applications/integration to help store and mitigate
• Product to sandbox data residency and PII requirements.
• Sandbox to product
• Tier 1 to sandbox (new)
Data management toolsets
Dynamics 365 Finance, Supply Chain Management, and Commerce

Data sharing has a rich toolset and processes available to support customers’ data
framework movement and migration requirements. Customers can use the fea-
tures in app and LCS to combine different approaches. Figure 10-10
highlights the options.

Data management workspace

In the case of Dynamics 365 Finance, Supply Chain Management,a

Implementation Guide: Success by Design: Guide 10, Data management 260

 and Commerce, first and foremost is the data management frame-
work. This is the primary tool, and you access it through the data
management workspace.

Here you have an administrator managing all data-related activities

through data projects using concepts like data entities, data templates,
and data packages.

You can use this workspace to run a number of data-related scenarios

like copying configurations between environments or loading data as
part of data migration.

For more information, refer to:

▪ Data entities overview
▪ Configuration data projects
▪ Configuration data templates
▪ Configuration data packages

Database operations
Though you can use data entities and data packages to move small
configurations, this may not be practical always. You may often find it
handy to move entire databases.

Database movement operations are a suite of self-service actions

that can be used as part of data application lifecycle management
(DataALM). These actions provide structured processes for common
implementation scenarios such as golden configuration promotion,
debugging/diagnostics, destructive testing, and general refresh for
training purposes.

You can use database movement operations to perform refresh, ex-

port, import, and point-in-time restore. For example, once your golden
configuration environment is ready and all testing cycles have been
completed and signed off in SIT and UAT, you can choose “Sandbox to
Production” type database request to restore this database to produc-
tion for go live.

Implementation Guide: Success by Design: Guide 10, Data management 261

 For more info go to:

References
▪ Database movement operations home page
▪ Submit service requests to the Dynamics 365 Service
Engineering team
Data Management/Data
Warehousing information, news
Data sharing framework
and tips - SearchDataManagement
(techtarget.com) Cross-company sharing is a mechanism for sharing reference and
group data among companies in a Dynamics 365 Finance, Supply
Insights-Driven Businesses Set The
Pace For Global Growth
Chain Management, and Commerce deployment.
(forrester.com)
This framework is introduced to allow sharing setups and master data
DMBoK - Data Management Body
of Knowledge (dama.org) across multiple legal entities. This facilitates master data management
when you are dealing with multiple legal entities and want to designate
one legal entity as master for some setups and parameters data. For ex-
ample, tax codes may not change from company to company so you can
set up in one legal entity and use cross company data sharing framework
and its policies to replicate the data across rest of the legal entities.

For more information, go to Cross-company data sharing.

Copy company configuration within an existing environment

If the deployment is a phased rollout involving multiple companies
within one Dynamics 365 instance, a very good strategy can be to use
the “Copy into legal entity” feature. This allows you to designate one
base company as the source template from which you can seed multi-
ple other companies or legal entities. All the setups are automatically
copied into those destination companies. This way you can templatize
your rollouts, allowing you to quickly stand up companies when need-
ed. This strategy needs some upfront planning and efforts in doing
the setups, but depending on the scale of a company’s rollouts can be
quite an effective strategy savings time and reducing errors.

For more information, see Copy configuration data between compa-

nies or legal entities overview.

262
 Configuration data and
Checklist data migration
Create, maintain, update, and test a configuration plan
throughout the project lifetime. It accounts for all the
Data governance required configuration data you import to support go live.

and architecture Ensure the data migration analyst, data migration archi-
Establish data governance principles to ensure data tect, and data steward create a plan for data migration
quality throughout the business processes lifecycle, that includes identifying data sources, data mapping,
focusing on the data’s availability, usability, integrity, environments, ETL, testing, and cutover planning.
security, and compliance.
Focus on maximizing the data throughput during
Appoint a data steward to ensure data governance migration by following Customer Engagement apps best
principles are applied. practices.

Define proper use cases and make data available to Optimize for network latency by staging in the cloud
support the business processes. and batching requests.

Define proper rules and select applications for master

data management. Data integration
Define a proper data architecture that depicts a holistic Only store necessary data in the app for key processes
view of the data repositories, their relationships with that actively interact with it and choose the right type of
each other, and ownership. Capture the data owners, data store based on the usage.
systems, and conceptual flow between systems during
your design and analysis phases.
Data quality
Data modeling Create a realistic view of your data quality and estimate
the efforts required to perform the necessary cleanup.
Define, clearly document, and keep a data model up to
Ensure that the apps have the necessary validations
date to serve as a blueprint.
and controls to enforce data quality and that you have
Conform the data model to the Common Data Model processes to measure and report on it.
standard without deviations to ensure cross-application
Maintain high-quality data by following the principles
compatibility and future readiness.
in this chapter, and have leadership drive the habit of
managing data on an ongoing basis.
Data storage
Estimate and forecast data storage needs across different
environments and types of data stores used in the solution.

Implementation Guide: Success by Design: Guide 10, Data management 263

 Case study
Building a rock-solid data
foundation, step by step
A global travel company with thousands of staff in the UK, Dubai, and
ANZ had been a user of a legacy homegrown IT system that had ex-
panded over the years.

The system was hard to maintain and fraught with issues as customer
data was divided in a number of applications and databases.

In order to target productivity, achieve compliance, and reduce costs,

the business started a project to move to Dynamics 365 Finance,
Supply Chain Management, and Customer Engagement.

The first phase of the project targeted the UK business and involved
moving over 150 of their business processes to Dynamics 365.

Since the travel business handles people’s personal data, including

sensitive data like passport information and medical certificates, they
had to be GDPR (The General Data Protection Regulation) compliant.
The team worked with an experienced system integrator and took
a proactive approach to data governance and data security starting
with charting out the architectural blueprint of their “to be” design.
They wanted to be on top of all data outflows and inflows to make
sure there is consistency, integrity, and security of data. There is one
source of truth, so they defined apps that held the master data and a
companywide policy formulation to let every department know what

Implementation Guide: Success by Design: Guide 10, Data management 264

 to find where and the exact process to request, update, and delete a
customer’s record.

They also built in workflows to manage data retention responsibly, like

disposing of information when requested by a customer.

The previous systems were disjointed with data existing in silos.

Different departments had sometimes widely different views of cus-
tomers, which caused confusion and poor service for customers. Call
centers were constantly busy as a customer’s call was routed from
department to department as no one person knew the whole picture.

One of the senior managers remarked that taking a proactive approach

to data governance and coming up with a design that put data at the
heart of every decision, helped the company to become truly data driven.

With the first phase of the UK deployment successful, the company is

now actively working on going live for Dubai and ANZ businesses.

The system integrator also recommended an approach of managing

configuration data for deployments. With businesses in three countries,
there is lot of commonalities in setups in the three legal entities that
could be reused. But at the same time, there are peculiarities that need
to be maintained separately. Instead of managing all this complexity
manually, system integrators had good experience using data templates
and packages in Finance and Operations apps. They worked with the
company and came up with a configuration plan checklist that identified
each of the three businesses’ specific setup requirements. This planning
exercise paid back multi-fold when the business discovered how smooth,
quick, and error free their subsequent deployments became.

As part of the project, they also moved their historical databases of

different applications to Azure Data Lake and connected with Power BI
to provide analytical reporting to the end users.

The company now has a unified view of their customers, which helps
them provide better customer service and allows marketing efforts to
be more targeted.

Implementation Guide: Success by Design: Guide 10, Data management 265

11
Guide
Application
lifecyle
management
Implementation Guide: Success by Design: Guide 11, Application lifecycle management 266
Success starts with
self-discipline.

Introduction
Generally, everything has a life and its own lifecycle.
Your solution goes through the same process, starting
with conception, moving through implementation,
then continuous operation, and finally to transition.
A solid application lifecycle management (ALM) strategy brings a
successful solution to customers with complete visibility, less manual
interaction with automation, improved delivery, and future planning.

With ALM, you have defined processes and In this chapter, we talk about ALM in the context of the overall solution
practices, a structured team, and tools at lifecycle. Your solution may use one or more of the Dynamics 365
your disposal.
Business Applications such as Finance, Supply Chain Management,
ALM is the management of your lifecycle (from
Sales, Field Service, or Commerce.
conception to operation) of your solution.

ALM includes disciplines that span throughout

the entire lifecycle of a solution, such as
Your entire solution lifecycle may go through several iterations. Your
governance (decision-making), project application lifecycle may also be part of a larger solution lifecycle. ALM
management, requirement management,
architecture, development, test management, provides defined guidance for people, processes, and tools. It sets the
maintenance, support, change management,
stage for a successful solution outcome. Additionally, innovation and
and release management.
automation are essential for improving the lifecycle of your solution.

What is ALM?
ALM is managing the end-to-end lifecycle of your solution, starting
from procuring the Dynamics 365 license, to mapping your business

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 267
 requirements in your application, designing your solution, extending
your custom requirements, validating and testing the solution consid-
ering business requirements, deploying the solution to business, and
maintaining it through the lifetime of your solution (Figure 11-1).

Planning
Fig.
11-1

Maintenance Requirement

Deployment Design

Testing Configuration

Development

The high-level lifecycle of your solution could be one or many imple-

mentations of your solution.

ALM focuses on a few key concepts:

▪ Defining best practices, processes, and templates for implementation
▪ Collaboration between customer, partner, and Microsoft team
members
▪ Multiple phase rollouts of your entire solution
▪ Innovation and automation using tools

Having well-documented processes connected with teams using

various tools will result in team confidence, a complete picture of the
implementation, and ultimately the success of your solution.

It’s very important that the implementation team has the appropri-
ate ALM tooling. ALM tools (such as Microsoft Azure DevOps) are

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 268
required to manage all aspects of the solution, including application
governance, requirement management, configuration, application
development, testing, deployment, and support. The ALM tool should
be well connected with all team members as well as all processes. For
example, when a developer checks in the code, they should mark the
changes with a particular work item, which connects the development
work item with the business requirement work item to the requirement
validation work item.

Application vs. development lifecycle

management
Some people may define ALM as improving development processes
such as version control, build automation, release deployments.
Although this is part of ALM, we refer to this part of the lifecycle as the
software development lifecycle (SDLC), or Dev ALM.

ALM isn’t just about the development lifecycle, in which we look at

the solution’s development activities including design, develop, test,
and deploy. Beyond the development lifecycle, ALM covers the entire
lifecycle of your implementation, including decision-making, project
management, requirement management, the application development
lifecycle, test management, and solution maintenance.

Why have
an ALM strategy?
From the time you conceptualize your Dynamics 365 solution, you
start the application lifecycle: from the project Initiate phase, to the
Implement phase, Prepare phase, and finally the Operate phase
(Figure 11-2).

During the lifecycle of the Dynamics 365 solution, you identify partner
teams, require project management, gather and map business process-
es, develop new processes, perform testing, deploy code, and finally
maintain the solution in production.

269
 Fig.
11-2

Initiate Implement Prepare Operate

Kickoff Configuration UAT Go live

Requirements Development Mock go live Support

Testing Cutover Bug tracking

Application Lifecycle Management

Think about taking a trip. You set dates, book flights and hotels, and plan
places to visit. All that planning will likely result in a great vacation.

In the same way, a well-planned ALM will lead to a solution that grows
your business. With ALM recommended practices, you’re set for success
in your solution implementation. You gain visibility into several areas:
▪ The current work items (requirement, development, testing)
▪ The work items completed, in progress, or planned
▪ The teams that worked, are working, or will work on specific tasks
▪ Issues and risks associated with solution implementation
▪ Development best practices
▪ Code history or code version control
▪ Build and release automation
▪ A testing plan, test cases, and test results against requirements

Your ALM may not be perfect right from the start. But it’s a foundation.
You can refine your ALM practices over time.

Implementation without defined ALM

ALM is the lifeline of your implementation. Let’s return to the planned
trip. You planned your dates and booked your flights, but you didn’t
book lodging or decide on any sightseeing excursions or activities.
Your trip could be a disappointment, and you might end up spending
more time planning when you should be out having fun.

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 270
 Like a poorly planned trip, if you don’t have effective ALM, you can
expect a negative impact on your implementation, solution quality,
and business satisfaction.

Without effective ALM, you may have poor decision-making, recom-

mended practices may not be followed, and teams are disjointed in the
implementation, which can cause delays.

Is ALM only for complex implementations?

You might think that implementing well-defined processes and track-
ing work items, individuals, and using automation tools is too much
work for a small or straightforward implementation.

However, every implementation no matter its size should follow ap-

plication lifecycle components, such as having a project plan, defined
scope, business requirements, and documentation. Your implementa-
tion might not require development or integration, but it will still have
configurations and application version updates.

With effective ALM and well-defined practices, you can keep your
solution healthy and up to date with the latest application releases.

Steps to successful ALM

You should use ALM from solution conception through operation. In
this section, we dive into some of the areas of the ALM process, includ-
ing before and after implementation, project management, process
management, other configurations, development, testing, and finally
maintenance and support.

During implementation
While you’re implementing the solution, you go through multiple
phases: Initiate, Implement, and Prepare. ALM applies to all these
aspects in your implementation:

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 271
 ▪ Project management
▪ Business process management
▪ Application configuration
▪ Development
▪ Testing
▪ Bug tracking
▪ Ideas, issues, risks, and documents
▪ Release management

After implementation
When you’re live in production, you’re in the Operate phase. ALM
continues with the following aspects:
▪ Continuous updates
▪ Independent software vendor (ISV) updates
▪ Maintenance and support
▪ New features
▪ Next phase

Project management
Efficient ALM requires having documented processes, templates, and
tools for all project management activities, such as project planning,
cost management, team management, ideas for improvements or
solutions, issues, and risk management. A project manager performs
their responsibilities efficiently when these areas are defined and
documented. For example, you should have a template for the project
plan that the project manager can apply in the Implement phase.

Chapter 8, “Project governance,” discusses project management in

more detail.
Efficient ALM
requires having
documented Business process management
processes, ALM processes and tools should include managing business require-
ments and processes efficiently and effectively. After you define your
templates, templates and tools, provide a framework for functional team mem-
and tools. bers to work together to define requirements and business processes

272
 efficiently. Teams can track business requirements for business process-
es and connect further with configuration and development.

For example, a functional team member can gather and define business
requirements in a given template and track them in Azure DevOps. They
can have it reviewed by assigning a work item to business users and then
store the business requirement document in a repository.

Chapter 7, “Process-focused solution,” covers requirement management

in detail.

Application configuration
ALM processes and tools should also include managing application
configurations. Before configuring an application, the business should
define and review business processes and requirements. Functional
team members should have defined processes with tools to perform fit
gap analysis, and a defined configuration required in the application.

It’s important to track your configuration setup and changes using

tools such as Azure DevOps. Typically, configuration is a one-time
activity without documentation and tracking. This leads to confusion,
no visibility, and questions about why particular decisions were made.
You should always maintain your documentation for application con-
figuration during testing, user acceptance testing (UAT), cutover, and
even after go live during maintenance and support.

Development
Having an efficient development lifecycle is one of the mandatory
aspects of ALM. In general, the development lifecycle consists of the fol-
lowing: Design, Develop, Build, Test, and Deploy. Continuous integration
and continuous deployment (CI/CD) is one of the best and latest practic-
es to enable delivering code changes more frequently and reliably.

After the business requirements are defined and identified, the devel-
opment team should get involved to work on any gaps in the solution.
The development team analyzes the requirements, reviews the

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 273
 functional design, prepares the technical design, and gets the technical
design reviewed. The development team should use version control,
create development tasks, link check-ins with work items, and prepare
a unit-testing document.

Finally, use build definitions and pipelines for automation and to

identify and fix any build issues. Build automation is vital for continu-
ous integration or gated check-ins. Manual builds are error-prone and
time-consuming. The build process, either manual or automated, is
mandatory for development.

The following example of a CI/CD approach (Figure 11-3) has a central-

ized version control in place, a build agent using build definitions, and
release pipelines to push code to a sandbox or other environments. In
this scenario, Dev A develops Solution A and Dev B develops Solution
B. Dev A gets Solution B (Dev B code) using version control and after a
successful build automation. When the build is successful, the code is
released to the sandbox environment using the release pipelines. After
successful testing, the code is released to the production environment.

Fig.
11-3
Developer A Sandbox
Developer A Code Solution A

Developer B Code Solution B

Solution A Version Control Build Agent

Build
Solution B

Successful Product
Developer B testing in Yes Solution A
Developer B Code
Developer B Code sandbox Solution B

Testing
Testing is an integral part of ALM. Under ALM, test management
processes and tools should be defined with templates to help manage
test preparation, implementation, and reporting. It should also include

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 274
 what tools to use for these steps. Chapter 14, “Testing strategy,” pro-
vides more information about test management.

Maintenance and support

In addition to the steps taken for planning, developing, and deploying
your solution, as part of effective ALM, you should have a plan to make
sure your solution continues to be optimal and healthy.

Chapter 20, “Service the solution,” and Chapter 21, “Transition to sup-
port,” cover maintaining the solution and the support process in detail.

ALM strategy for

Dynamics 365
ALM can improve every aspect of your Dynamics 365 solution.
Dynamics 365 offers various tools to achieve a successful ALM in
your implementation.

Team responsibility
Various team members are involved in an implementation, with sep-
arate roles and responsibilities. Multiple teams such as the customer,
partner, and ISV work are involved and must work together.

Every team member, either directly or indirectly involved, should own

the ALM maintenance of the processes and tools they’re using. For
example, the project manager owns and maintains the project plan,
and the tester owns and maintains the test plan. You should also have
cross-team collaboration for maintaining ALM practices.

Azure DevOps
Microsoft recommends using Azure DevOps as the tool for managing
or maintaining your ALM practices and processes. For some areas of
Dynamics 365 (such as Finance and Operations apps), Azure DevOps is

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 275
 the only version control tool.

In addition to the standard templates in Azure DevOps, you can build

your own custom templates according to your business needs. Azure
DevOps lets you store all your work items, including features, user
stories, requirements, tasks, bugs, and more.

A standard lifecycle is available for each work item, and you can build
your own custom states and rules for each type of work item.

Dynamics 365 also recommends Azure DevOps as the version control

management tool for your code. You can achieve build automation
and release pipelines using build definitions and release definitions.

You can also use Azure DevOps for project management, test manage-
Take some time to learn more about
ment, bug tracking, release management, and many more aspects of
DevOps tools on Azure.
your implementation.

Finance and Operations apps

In this section, we provide some of the Finance and Operations apps
recommendations to achieve an efficient ALM in your solution. We’ll
cover things such as Microsoft Dynamics Lifecycle Services, business
process modeling, development, and version control.

Microsoft Dynamics Lifecycle Services

Microsoft Dynamics Lifecycle Services (LCS) is a cloud-based solution in
which all participants (customer, partner, and Microsoft) work collabo-
ratively, and you manage your implementation from pre-sales, through
implementation, and finally the Operate phase. LCS provides various
checklists and tools to help you manage your project efficiently:
▪ A single location for all your projects
▪ Project onboarding
▪ Project methodologies
▪ Azure DevOps configuration
▪ Business process modeler (BPM)
▪ Cloud-hosted environments (CHE)
▪ Data application lifecycle management (DataALM)

276
 ▪ Management of Microsoft support
▪ Customization analysis
▪ Subscription estimator
▪ Issue search
▪ Continuous updates
▪ Service requests to Dynamics Service Engineering (DSE)
▪ Environment monitoring and diagnostics
▪ Asset library

The goal of LCS is to deliver the right information, at the right time, to
the right people, and to help ensure repeatable, predictable success
with each rollout of an implementation, update, or upgrade.

Business process modeler

Business process modeler (BPM) provides abstract explanations of your
solution’s business processes. Business processes are the underlying
foundation for application configuration and application development.

Effective ALM provides defined, documented, and templated processes

to follow in your implementation. You can model your business pro-
cesses in Dynamics 365 Finance and Operations apps Management
using BPM under LCS.

You can use the BPM tool to define, view, and edit the Finance and
Operations apps out-of-box business processes (in the form of librar-
ies), which you can use in future implementations. The tool helps you
review your processes, track the progress of your project, and sync your
business processes and requirements to Azure DevOps.

Every business is different in some ways; this tool helps you align your
business processes with your industry-specific business processes and
best practices. BPM libraries provide a foundation for your business
process, and you can add, remove, and edit your processes according
to your solution requirements.

Development
In Finance and Operations apps, Microsoft Visual Studio is used as the
development environment. The development lifecycle includes the

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 277
 following steps (Figure 11-4):
▪ Each developer uses their own development environment
▪ Developers write source code and check in their code to
Azure DevOps
▪ Developers also sync code from Azure DevOps to get the source
code from other developers
▪ Source code can be merged to respective branches depending on
the branching strategy
▪ The build takes the source code from Azure DevOps, uses the build
definition, and creates a deployable package
▪ The build pipeline also pushes the deployable package to the LCS
asset library
▪ Azure release pipelines work with Visual Studio to simplify deploy-
ing packages to UAT
▪ When UAT is complete, the deployable package is marked as a
release candidate to deploy to production
▪ The Dynamics Service Engineering (DSE) team deploys it to pro-
Fig. duction using a service request from LCS
11-4

Environment
Developer Developer
Build User acceptance Product
virtual machine virtual machine
testing
1 2

Source code Source code Source code Application Application Release Release
deployable deployable candidate candidate
package package

Azure DevOps Lifecycle Services

shared assets

Tool

Version control
The primary purpose of version control is storing and maintaining
source code for customizations, as well as ISV solutions. You develop
against local, XML-based files (not the online database), which are

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 278
 stored in version control tools such as Azure DevOps. The following are
Fig.
recommendations for version control branching:
11-5
▪ Consider using minimum branching option
▪ Consider the following recommended branching strategy:
Release Release XX ▫ Development Developer check-in and testing with develop-
branch branch
ment data (Trunk/Dev)
▫ Test Deploying to the tier 2+ and testing with current produc-
tion data (Trunk/Main)

Test (main) branch ▫ Release or Release XX Retesting in the tier 2+ and deploying to
production (Trunk/Release) or v-next (Trunk/Release XX)
▪ Run Customization Analysis Report (CAR) tool to check best prac-
tices and resolve errors and warnings
Code ▪ Use the shelve command or suspend changes to keep work safe
Development upgrade
branch branch ▪ Request a code review to ensure code quality
▪ Check in code when a feature is complete, and include changes
Arrows show the direction
when creating the branch
from one feature in each changeset
▪ Merge each feature in a separate changeset
▪ Don’t check in code directly to the test or release branches
▪ Don’t check in changes for more than one feature in a single
changeset
▪ Don’t mark deployable packages from the development and test
branches as release candidates
For more information, take a look at ▪ Don’t merge untested features into the release branch
our document on how to develop and
customize your home page.
Figure 11-5 illustrates a recommended branching strategy.

Build automation
Speed is essential for rapid implementation, and build automation is
the key to achieving this.

The build process is mandatory for any code to run. This process
involves compiling the source code and producing binary files (assem-
blies). A database sync also requires a build first because the schema is
retrieved from the assemblies (and not the XML files).

The Azure DevOps build system provides the following triggers for builds:
▪ Scheduled builds, such as nightly at 6 PM
▪ Continuous integration, such as:

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 279
 ▫ Starting a build as soon as code is checked in
▫ Gated check-in
▪ Manual builds (on demand)

Here’s a checklist to ensure your build goes smoothly:

▪ Make sure the code compiles without errors
▪ Don’t use build environments for development activities
▪ Create a build definition for each branch and for each build trigger
▪ Consider using continuous integration or gated check-in build triggers
To learn more about build automation, ▪ Perform automated testing after successful builds
check out our guides on how to:
▪ Create a single deployable package that contains all packages
▪ Deploy and use a continuous build
and test automation environment ▪ Don’t use Visual Studio to create a deployable package, for the
▪ Create a build automation that uses
Microsoft-hosted agents and Azure
following reasons:
Pipelines ▫ Deployable packages contain all the code on the development
▪ Update model versions in an
automated build machine, including experimental code and incomplete features
▪ Create deployable packages in Azure ▫ Deployable packages may not include all the modified pack-
Pipelines - Finance & Operations |
Dynamics 365 | Microsoft Docs. ages, which means they’re not independent and testing results
could vary
▪ Create a deployable package with a build machine so it’s self-con-
tained and contains all applicable modules
▪ Make sure to keep service updates on the build machine less than
Fig. or equal to production
11-6
▪ Follow a consistent naming convention for your deployable package

Author tests
Automated testing
Unit tests As part of development ALM, testing automation should be in place
Task recorder tests to achieve fast-moving code from development to deployment. In
Component tests Finance and Operations apps, you can integrate testing and validation
two different ways:

Integrate tests ▪ Unit and component level testing using SysTest framework
▪ Automated testing using Task recorder and the Regression suite
Test module setup automation tool (RSAT)
Integrate with
build system
To keep up with innovation and constant changes in your solution,
it’s critical to invest and build in continuous validation. Achieving this
Run tests
requires different components (Figure 11-6).
Continuous
validation The RSAT significantly reduces the time and cost of UAT for Finance

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 280
 and Operations apps. RSAT lets functional super users record business
tasks by using Task recorder and converting the recordings into a suite
of automated tests, without having to write source code.
For more information about automated
testing, we offer tutorials on testing and
validations, using the regression suite Deployment
automation tool, and acceptance test
library resources. Let’s review the key concepts for Finance and Operations
apps deployment:
▪ Deployable package  A unit of deployment that can be applied
in an environment
▪ Deployment runbook  A series of steps that are generated to
apply the deployable package to the target environment
▪ AX Installer  Creates a runbook that enables installing a package

You can deploy a package manually or through automated deploy-

ment (which isn’t applicable to production). We recommend the
following when deploying Finance and Operations apps:
▪ Consider using automated deployment to achieve continuous
deployment.
▪ Make sure that the deployable package that should be applied is valid.
▪ Make sure that the package is applied in a sandbox environment
For more information about deployment, before it’s applied in the production environment.
we offer tutorials on:
▪ If you want to apply multiple packages, create a merged all-in-one
▪ Download assets by using Azure
Pipelines - Finance & Operations | package that can be applied first in a sandbox environment and
Dynamics 365 | Microsoft Docs
▪ All-in-one deployable packages -
then in the production environment.
Finance & Operations | Dynamics 365 | ▪ Production deployment is manual. After successful UAT, mark the
Microsoft Docs
▪ Deploy assets by using Azure Pipelines - deployable package as a release candidate and schedule a service
Finance & Operations | Dynamics 365 |
request with the DSE team.
Microsoft Docs
▪ Clean up the old deployable package periodically from the LCS
asset library.

One Version (Continuous Updates)

Microsoft is innovating and investing in Dynamics 365 for Finance and
Operations apps to ensure that this can be achieved with the right
attention to quality and stability demanded by customers.

Prior to Finance and Operations apps, there was a major concern in

regards to keeping the environment up to date with the latest features
and bug fixes. Customers were using different versions, and upgrading

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 281
 to latest version was a hefty task. Once you have gone live with Finance
and Operations apps, the days where customers need to have large
upgrade implementations are gone.
To learn more about One Version or
Continuous Updates, see One Version
service updates overview - Finance & With One Version, all customers benefit from being on the same ver-
Operations | Dynamics 365 | Microsoft Docs.
sion with access to the latest capabilities available, and no one is stuck
You can find many of your questions
on less capable older versions.
answered in One Version service updates
FAQ - Finance & Operations | Dynamics
365 | Microsoft Docs.

Customer Engagement apps

When organizations implement their software as a service (SaaS) cloud
solution, some level of customization and extensibility typically exists,
with the goal to provide additional value and adjust the functionalities
to specific business or industry needs.

The following concepts are important to understanding ALM when

using the Power Platform.

Solutions
Solutions are the mechanism for implementing Dev ALM in Power
Platform apps. They’re the vehicle that distributes components across
the different environments.

A component represents what can be customized within the solution,

such as site maps, applications, the components of the data model,
forms, charts, or plug-ins.

Solutions have a publisher associated with them. The publisher defines

the prefix associated with newly created components and allows you to
easily recognize who developed the components. One custom publish-
er should be used for all components and should have a meaningful
name such as the company’s initials.

Solutions can either be managed or unmanaged. You should use

unmanaged solutions in development environments while changes are
still underway. Managed solutions are used to deploy to any environ-
ment that isn’t a development environment for that solution.

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 282
 Use connection references and environment variables to define
information about a connector or to store parameter keys and val-
ues. This streamlines the process of migrating the solutions between
For more information about deployment,
environments.
refer to our guides on how to:
▪ Organize your customizations with
solutions
▪ Create and deploy solutions
You should have a good understanding of the solution concepts in
▪ Use tools to migrate you solutions the Power Platform—the solution is the baseline for the components
between environments
▪ Support team development across their lifecycle. For more information, refer to our overview of
solution concepts.

Tools
Several tools are available to help automate the process of managing
and shipping solutions, which can help increase efficiency, reduce
manual labor, and reduce human error when working with solutions
throughout their lifecycle.

Consider deploying the ALM Accelerator for Makers or Advanced

Makers in order to expedite your ALM setup time. These accelerators
are pre-built applications that provide a starting point for automating
your solution version control and migrations between environments.

Applications like Azure DevOps and GitHub provide developer ser-

vices that support the activities required to plan, store, collaborate,
build, and deploy. This includes synchronization of solution metadata
between development environments and your version control system,
generating build artifacts, deploying to downstream environments,
provisioning or de-provisioning environments, and performing static
analysis checks against your solution by using the Power Apps checker
service. Consider using the Power Apps Build tools or GitHub actions
as building blocks in creating your pipelines.

Consider using the following tools:

▪ Version control system  This category of software tools helps
record changes to files by keeping track of changes committed to
software code. A version control system is a database of changes,
which contains all the edits and historical versions of a software
project. Version control systems allow you to maintain a single
source of truth and recall specific versions when needed. Git and

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 283
 GitHub are a popular example of version control applications.
▪ The Configuration Migration tool  This tool enables you to
move configuration and reference data across environments.
Configuration and reference data is data that is stored as records
in Dataverse but that supports various aspects of the application
such as USD, portal configuration, and lookup values.
▪ Package deployer  The package deployer lets administrators or
developers deploy comprehensive packages of relevant assets to
Dataverse environments. Packages can consist of not only solution
files, but also flat files, custom code, and HTML files. You have the
ability to automate the execution of the package deployer with
DevOps build tools task Deploy package.
▪ Solution packager  This tool can unpack or pack a compressed
solution file into multiple XML files and other files or vice versa,
so they can be easily managed by a source control system or be
deployed to an environment from source control. The execution of
this can be automated with DevOps build tools tasks.
▪ The Power Apps CLI  The Power Apps CLI is a simple, single-stop
developer command-line interface that empowers developers
and app makers to create code components. You can also use this
interface to automate the deployment of portal configurations.
▪ PowerShell cmdlets  The PowerShell cmdlets for administrators,
app makers, and developers allow automation of monitoring,
management, and quality assurance tasks that are possible
through the Power Apps admin center user interface.
▪ Test Studio   Use Test Studio to ensure the quality of your
canvas apps.
Check out our additional resources for:
▪ Microsoft Power Platform Build ▪ Easy Repro   Use Easy Repro or Playwright to generate test
Tools for Azure DevOps
▪ Power Apps build tools for Azure
scripts that can be used to regression test changes to your model
DevOps driven apps.
▪ Key concepts for new Azure
Pipelines users ▪ Performance Insights  Use performance insights to analyze
runtime user data and provide recommendations.

ALM workshop strategy

The implementation team should plan for an ALM workshop before
project initiation or before project kickoff. The customer and partner

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 284
 should gather the required processes, tools, and teams before project
kickoff, such as:
▪ Project plan template
▪ Business requirement document template
▪ Functional and technical spec templates
▪ Test plan and test suite template
▪ Defined tools such as DevOps template
▪ Defined roles required for partner and customer teams

In addition to this workshop, the implementation team should plan

for a Dev ALM workshop before starting development activities. Our
FastTrack ALM workshop focuses on your development lifecycle and
provides guidance and best practices such as development environ-
ments, version control, build automation, and release pipelines. It
underlines the importance of having fast builds, quick testing, and fast
deployments using automation, which helps ensure a rapid develop-
ment lifecycle, especially for agile development methodology.

Workshop scope
FastTrack ALM workshops are designed for implementers who want to
make sure that their development approach meets the requirements
of the implementation and is aligned with typical best practices. The
workshop could cover several topics:
▪ Development work management  Focuses on high-level,
day-to-day developer activities. It reviews that the team has devel-
opment guidelines, development best practices, and work items
such as requirement, tasks, and bugs management.
▪ Code management  Reviews your version control, branching,
code merging, and code reviews strategy. Branching can be simple
or complex depending on the project and phases. It also looks at
how customization is checked in, such as gated check-in.
▪ Build management  Looks at your build strategies, such as
manual build or automated build. It mainly reviews your build

An ALM workshop plan for build definitions, different variables, and build triggers for
your implementation. It also reviews whether you’re using a build
provides guidance environment or Azure-hosted builds.
and best practices.
285
 ▪ Release management  Assesses your deployment plan for

References customizations, ISVs, and continuous (one version) updates. It also

reviews how hotfixes and your code release, or one version updates,
are applied to production and nonproduction environments.
DevOps and
Azure Pipelines Timing
It’s important to ensure ALM workshops are conducted at the right time.
DevOps solutions on Azure
Ideally, the ALM workshop should be before project kickoff, and the Dev
Release pipelines
ALM workshop should be before starting development or sooner.
Sign up for Azure Pipelines

You should complete both workshops either in the Initiate phase or

Finance and when starting the Implement phase. If you plan ALM workshops too far

Operations apps into implementation, any findings and recommendations could cause
significant rework.
Implement application lifecycle management
in Finance and Operations apps - Learn |
Microsoft Docs

Business process modeler (BPM) in Lifecycle

Services (LCS)
Conclusion
Dynamics 365 Finance and Operations Tools - ALM is the management of your solution lifecycle from conception to
Visual Studio Marketplace
operation. It includes governance (decision-making), project manage-

Customer ment, managing requirements, solution architecture, development,

CI/CD, test management, maintenance and support, change manage-
Engagement apps ment, release management, and many more areas.
Application lifecycle management (ALM) with
Microsoft Power Platform
Your solution lifecycle may go through several evolutions. Each evo-
Microsoft Dynamics CRM SDK Templates
lution may use the same SDLC methodology. Basically, SDLC is part of
Solution Lifecycle Management: Dynamics 365
ALM, ALM is far bigger than Dev ALM, and Dev ALM is part of SDLC.
for Customer Engagement apps

286
 Checklist

Establish the ALM at the onset and align it to the

environments, source control, implementation method-
ology, and automation strategy.

Ensure the ALM provides an efficient, reliable, robust,

and automated mechanism to smoothly deploy the
configurations and customizations across different envi-
ronments, including peripheral solution components.

Ensure ALM strategy enforces a source control-centric

approach that enables creation of new environments
with the latest configurations without the need to
replicate existing environments.

Use a build automation and CI/CD pipeline for the project.

Consider the structure of the implementation team to

enable effective collaboration while providing exclusive
control to developers where needed.

Follow documented best practices in general and

specifically to each app.

For Customer Engagement projects, download the

whitepaper Solution Lifecycle Management: Dynamics
365 for Customer Engagement and complete the
self-assessment.

Implementation Guide: Success by Design: Guide 11, Application lifecycle management 287
 Case study
Global transport systems
company finds ALM strategy
to be a cornerstone of
implementation
A global transport systems company embarked on a multiple-phase,
multiple-region implementation of Dynamics 365 Sales, Customer
Service, and Field Service with an aggressive schedule for achieving a
minimum viable product (MVP) go live.

To give business stakeholders a taste of the solution before the official

project kickoff, the customer asked their Dynamics 365 implementa-
tion partner to lead a rapid prototyping phase, which the customer
referred to as “Phase 0.” The Phase 0 team included technical salespeo-
ple, as well as consultants who would leave when the project kicked off.

Considering Phase 0 to be a rapid prototype, the implementation team

didn’t see a need to focus on governance and quality. To meet the
deadlines, the team made several quick decisions without assessing the
impact on the overall application management:
▪ Staffing Phase 0 with resources who wouldn’t remain part of the
project team.
▪ Developing the prototype using separate, unmanaged solutions with
different publishers for Sales, Customer Service, and Field Service.
▪ Pivoting to managed solutions for test and production

Implementation Guides: Success by Design: Guide 11, Application lifecycle management 288
 environments too late in the design phase, after unmanaged
solutions were already built and deployed to these environments
during Phase 0.
▪ Merging the Sales, Customer Service, and Field Service solutions
into a single “workstream” solution during the build phase of the
implementation lifecycle.
▪ Resolving errors too quickly during solution deployments because
of pressure from project leadership, instead of taking the time
to understand each problem’s root cause. (Such pressure led to
developers customizing directly within the test environment.)

As the project team prepared for the go live, the seemingly indepen-
dent decisions made during the initial phases resulted in deployment
issues that eventually stalled the go live. Investigations by Microsoft
Support confirmed that there was no ALM strategy in place, and iden-
tified the key issues:
▪ Unmanaged solutions caused solution-layering issues for testing and
production, affecting the sanctity and integrity of the environments
▪ A prototype solution employed for production purposes introduced
quality issues
▪ Failure to use ALM practices such as DevOps caused traceability
issues and prompted developers to build functionality that wasn’t
aligned with customer requirements
▪ Suboptimal code was implemented because tools such as solution
checker weren’t used to enforce code quality
▪ Testing without traceability was insufficient, and buggy code was
deployed to other environments

As the old saying goes, “By failing to prepare, you are preparing to fail.”
The MVP go-live date was delayed by 12 weeks and Microsoft worked
alongside the project team to determine the root cause of each issue.
The team eventually acknowledged that a series of seemingly unre-
lated decisions affected the MVP go live, and they sent every team
member working in a technical role to an ALM refresher training. The
project team also confirmed plans that should have been solidified at
the beginning of the project, including an environment strategy and a
solution management and ALM approach.

Implementation Guides: Success by Design: Guide 11, Application lifecycle management 289
 After the refresher training, the project team started with a “crawl to
walk” approach. During the “crawl” stage, they implemented mandatory
ALM practices with these governance elements:
▪ Cleaning up their production and test environments, and moving
from unmanaged to managed solutions
▪ Implementing a build process to deploy builds from development
to test to production
▪ Establishing a governance process to restrict developer access to
production and test environments
▪ Adding a bug-triaging process that allowed the development team
to troubleshoot and fix issues in development environments and
use the build process to deploy fixes in higher-tier environments
▪ Mandating the generation of solution checker reports

Once these practices were in place, the implementation team moved

toward a partially automated ALM process that included setting up
DevOps pipelines to automate build deployments and auto-generate
solution checker reports. As this case study illustrates, an ALM strategy is
a cornerstone of the entire implementation—whether it’s a fresh imple-
mentation or a rapid prototype elevated to a production-use solution.

Implementation Guides: Success by Design: Guide 11, Application lifecycle management 290
12
Implementation Guide: Success by Design: Guide 12, Security
Guide
Security
291
“Businesses and users are going
to embrace technology only if
they can trust it.”
– Satya Nadella, Chief Executive Officer of Microsoft

Introduction
In this chapter, we look at the fundamental
security principles applicable to Microsoft
Dynamics 365 implementations.
Next, we discuss in more detail how some of these principles apply
differently to Dynamics 365 Customer Engagement and Dynamics 365
Finance and Operations applications. We then address the importance
of making security a priority from day one, with specific
examples from each product that build upon the concepts we’ve
discussed. Finally, we look at how to avoid common mistakes by
examining some key anti-patterns.

Security overview
Security is the protection of IT systems and networks from theft or
damage to their hardware, software, or data and disruption of the
service. Dynamics 365 is a software as a service (SaaS) offering from
Microsoft. In a SaaS service, data and applications are hosted with a
provider (in this case, Microsoft) and accessed over the internet. In this
deployment model, the customer maintains ownership of the data,
but shares application control with the provider. Therefore, security,
compliance, privacy, and data protection are shared responsibilities
between provider and customer.

Implementation Guide: Success by Design: Guide 12, Security 292

 Figure 12-1 illustrates the areas of responsibility between customer
Fig.
12-1
and Microsoft based on deployment type.

Shared responsibility model

Physical datacenter SaaS PaaS IaaS On-
Prem
Information and data
Devices (mobile and PCs) Responsibility always retained by customer
Accounts and identities
Identity and directory infrastructure
Applications
Network controls
Responsibility varies by service type
Operating system
Physical hosts
Physical network Responsibility transfers to cloud provider
Physical datacenter

Microsoft Customer

Fig.
12-2 Microsoft takes its commitment seriously to safeguard customers’
data, to protect their right to make decisions about that data, and to
Security Implement
be transparent about what happens to that data. On our mission to
strong security
measures to
empower everyone to achieve more, we partner with organizations,
safeguard your empowering them to achieve their vision on a trusted platform. The
data Microsoft Trusted Cloud was built on the foundational principles of
security, privacy, compliance, and transparency, and these four key
Privacy Provide you with principles guide the way we do business in the cloud (Figure 12-2). We
control over your apply these principals to your data as follows:
data to help keep ▪ Security  Implement strong security measures to safeguard
it private
your data
▪ Privacy  Protect your right to control and make decisions about
Compliance your data to help keep it private
Help you meet
▪ Compliance  Manage your data in compliance with the law and
your specific
help you meet your compliance needs
compliance needs
▪ Transparency  Be transparent about our enterprise cloud services
and explain what we do with your data in clear, plain language

Transparency Explain what we

do with your
Microsoft has been defending against threats and providing security
data in clear, protections for our online services since 1994, and we invest over $1
plain language billion dollars per year to continue our commitment to protecting our
customers. For these reasons, we say that our products and services

Implementation Guide: Success by Design: Guide 12, Security 293

 run on trust. You can place your trust in our expertise. This confidence
allows you to focus on running your business. Ninety-five percent of
Fortune 500 businesses run on the trusted Microsoft Cloud.

Fig.
12-3
Compliance
Every organization must comply with the legal and regulatory standards
Compliance goals of the industry and region they operate in, and many are also subject
Define and document standard to additional contractual requirements and corporate policies. Figure
operating procedures that meet 12-3 lists some standard compliance goals and their implementation in
multiple certification requirements
Dynamics 365.
Run the service in a compliant
fashion and collect evidence Microsoft is responsible for the platform, including the services it

Control and audit access to

offers, and provides a cloud service that can meet or exceed the se-
environments as well as actions curity, privacy, and compliance needs of your organization. Microsoft
to data complies with data protection and privacy laws applicable to cloud
services, and our compliance with world-class industry standards is
verified. Detailed information about compliance for our cloud services
Implementation details and solutions that help organizations meet regulatory requirements
for data security is available at our online Microsoft Trust Center.
Adhere to strict privacy and security
practices when building features
and when operating the service As noted earlier, compliance is a shared responsibility. To comply with
laws and regulations, cloud service providers and their customers
Pass internal and external audits
enter a shared responsibility to ensure that each does their part. Tools
Prioritize cross-industry available at the Trust Center include compliance offerings that help
certifications
you comply with national, regional, and industry-specific requirements
Authentication and authorization governing the collection and use of data, and audit reports that help
you verify technical compliance and control requirements. Specific
tools you can access at the Trust Center include:
▪ Compliance Manager, a cross-Microsoft cloud services solution
designed to help organizations meet complex compliance ob-
ligations like the General Data Protection Regulation (GDPR). It
performs real-time risk assessment that reflects compliance pos-
ture against data protection regulations when you use Microsoft
cloud services. It also provides recommended actions and step-by-
step guidance.
▪ Service Trust Portal, which contains details and documents such as

Implementation Guide: Success by Design: Guide 12, Security 294

 whitepapers, ISO reports, and other resources detailing Microsoft’s
implementation of controls and processes that protect cloud
services and customer data.

Customer responsibility
As a customer, you’re responsible for the environment after the service
has been provisioned. You must identify which controls apply to your
Refer to Microsoft compliance
offerings for more information business and understand how to implement and configure them to
about regulatory compliance
standards and Microsoft products.
manage security and compliance within the applicable regulatory
requirements of your nation, region, and industry.

Privacy
You are the owner of your data; we don’t mine your data for advertising.
Fig. If you ever choose to end the service, you can take your data with you.
12-4

Figure 12-4 lists some standard privacy goals and their implementa-
Privacy goals tion in Dynamics 365.
You own your data

You know where your data is located How we use your data
You control your customer data Your data is your business, and you can access, modify, or delete it at
any time. Microsoft will not use your data without your agreement, and
when we have your agreement, we use your data to provide only the
services you have chosen. We only process your data based on your
Implementation details
agreement and in accordance with the strict policies and procedures
All data is classified that we have contractually agreed to. We don’t share your data with

Role-based security puts the

advertiser-supported services, nor do we mine it for any purposes like
customer in charge marketing research or advertising. Learn more about how Microsoft
categorizes data in the delivery of online services.
Your own logically isolated data
repository helps maximize the
security and integrity of your data We believe you should have control over your data. The Trust Center
can tell you more about how we handle data requests from government
Documented data-handling
practices in the Microsoft Online and law enforcement agencies.
Services in the Trust Center

Implementation Guide: Success by Design: Guide 12, Security 295

 Customer responsibility
As a customer, you’re responsible for data classification, identity man-
agement, and assigning appropriate security roles to secure the data.

Security
Security is a shared responsibility in a SaaS deployment. This means
that some aspects of security are shared by both the customer and the
provider, other aspects are the responsibility of the customer, and others
are the responsibility of the provider. For Dynamics 365 deployments,
Microsoft as the cloud provider is responsible for security aspects includ-
ing physical datacenter security, the operating system, network controls,
and providing a secure application framework (Figure 12-5).
Fig.
12-5

Microsoft
protecting you
Intelligent
Security
Graph
Industry
Partners
Antivirus
Network
Certifications ...
Cyber Defense
Opperations Center

Malware protection center Cyber hunting teams Security response center Digital crimes unit ...

Conditional
access
Cloud app
security
Event
management
Rights
management
Key
vault
Security
center
... Active
protection
service
Windows
update
Microsoft 365
advanced threat
protection
Smart
screen

PaaS IaaS

Advanced
threat Terminal
analytics app

Azure Active Microsoft SaaS Windows Outlook

Directory Dynamics
365

Identity Applications and data Infrastructure Device

Implementation Guide: Success by Design: Guide 12, Security 296

 Fig.
Figure 12-6 lists some standard compliance goals and their implemen-
12-6
tation in Dynamics 365.

Security goals The following is a list of core security controls available in Dynamics 365:
Safeguard data using state-of- ▪ Security Development Lifecycle
the-art security technology and ▪ Datacenter security
processes of the industry
▪ Data segregation
Use the same identity platform as ▪ Encryption
Microsoft 365, so users have the ▪ Secure Identity
same username and password for all
▪ Authorization
Control who can access what data ▪ Auditing and monitoring

Customer data is isolated from

other tenants
Security Development Lifecycle
The Security Development Lifecycle (SDL) is a process that helps de-
velopers build more secure software and address security compliance
Implementation details requirements by introducing security and privacy considerations
State-of-the-art physical security throughout all phases of the development process. It consists of the
measures following phases: training, requirements, design, implementation,

24/7 incident response teams to

verification, release, and finally response (Figure 12-7).
mitigate threats and attacks

Encryption of data transferred Datacenter security

between our datacenters and at rest
Microsoft designs, builds, and operates datacenters in a way that
Single sign-on federated through
strictly controls physical access to the areas where your data is stored.
Azure Active Directory
Microsoft understands the importance of protecting your data and is
Role-based security committed to helping secure the datacenters that contain your data.
We have an entire division at Microsoft devoted to designing, building,
Fig.
12-7

Training Requirements Design Implementation Verification Release Response

Core security Establish security Establish design Use approved tools Perform dynamic Create an incident Implement incident
training requirements requirements analysis response plan response plan

Create quality gates Perform attack Desperate unsafe Perform fuzz Conduct final
and bug bars surface analysis functions testing security review
and reduction

Perform security Use threat Perform Conduct attack Certify release

and privacy rich modeling static analysis surface review and archive
assessments

Implementation Guide: Success by Design: Guide 12, Security 297

 and operating the physical facilities supporting Azure, which delivers
Dynamics 365 (Figure 12-8). This team is invested in maintaining state-

Fig.
of-the-art physical security.
12-8

Barriers Fencing
Perimeter

Seismic Security 24/7 Days of backup

bracing operations center security staff power

Building

Two-factor access control:

Cameras Alarms Biometric readers and card readers
Computer
room

DDoS defense system

A distributed denial of service (DDoS) is an attack in which multiple
compromised computer systems attack a target, such as a server,
website, or other network resource, and cause a denial of service for
users of the targeted resource.

Azure has a defense system against DDoS attacks on its platform services.
It uses standard detection and mitigation techniques, and is designed
to withstand attacks generated from outside and inside the platform.

Data segregation
Dynamics 365 runs on Azure, so it’s inherently a multi-tenant service,
meaning that multiple customers’ deployments and virtual machines
are stored on the same physical hardware. Azure uses logical isolation
to segregate each customer’s data from others. This provides the scale
and economic benefits of multi-tenant services while rigorously pre-
venting customers from accessing one another’s data.

Implementation Guide: Success by Design: Guide 12, Security 298

 Encryption
Data is an organization’s most valuable and irreplaceable assets,
and encryption serves as the last and strongest line of defense in a
multi-layered data security strategy. Microsoft business cloud services
and products use encryption in transit and at rest to safeguard custom-
er data and help maintain control over it.

At-rest data protection

At minimum, Dynamics 365 environment databases use SQL TDE
(Transparent Data Encryption, compliant with FIPS 140-2) to provide
real-time I/O encryption and decryption of the data and log files for
Fig.
data encryption at rest (Figure 12-9).
12-9

Full database encryption using SQL Transparent Database Encryption

Data protection
Windows OS level Capabilities
API (DPAPI)
DPAPI encrypts SMK
SQL TDE performs real-time I/O encryption and encryption of
the data and log files to provide data encryption at rest

SQL instance level Service master key All data transmitted between servers within the datacenter for
mirroring is encypted as well
SMK encrypts the DMK
for master DB Microsoft manages the keys and handles management of
encryption

Master DB level Database master key Encrypts the entrie SQL database

No CMROL code changes required

DMK creates cert in
master DB
Master keys protect database-level keys

Master DB level Certificate Can change the database master key without having to change
the actual database encryption

Certificate encrypts DEK

in content DB

Availability
Database
Content DB level
encryptions key
Available now for all Dynamics 365 online environments
Certificate encrypts DEK
in content DB

Organization DB

Implementation Guide: Success by Design: Guide 12, Security 299

 By default, Microsoft stores and manages the database encryption keys
for your Dynamics 365 deployments. Finance and Operations apps use
server-side encryption using service-managed keys. All key manage-
ment aspects such as key issuance, rotation, and backup are handled
by Microsoft. For Customer Engagement apps, the optional customer
managed encryption key capability provides administrators the ability
to self-manage the database encryption key for all the instances in
their tenant.

In-transit data protection

Your data in transit is protected in several ways (Figure 12-10).
Dynamics 365 uses HTTPS encryption so that you can enable encryp-
tion for traffic between Dynamics 365 and end users. Azure protects
data in transit to or from outside components, as well as data in transit
internally, such as between two virtual networks. Azure uses industry
standard transport protocols such as TLS between user devices and
Microsoft datacenters, and within datacenters themselves.
Fig.
12-10

End-to-end
Data in transit Data in transit
encryption of
between a user between
communications
and the service datacenters
between users

Protects user from Protects from bulk Protects from

interception of their interception of data interception or loss of
communication and data in transit
helps ensure transaction between users
integrity

Implementation Guide: Success by Design: Guide 12, Security 300

 Secure identity
Identity and access management is critical to every organization. Azure
Active Directory (Azure AD) is a complete identity and access manage-
ment solution with integrated security that connects 425 million people
to their apps, devices, and data each month. Dynamics 365 applications
are safeguarded using Azure AD as a seamless identity solution.

Authentication: Users
Authentication is the process of proving an identity. The Microsoft identity
platform uses the OpenID Connect protocol for handling authentication.
By default, only authenticated users can access Dynamics 365.

Azure AD is used as a centralized identity provider in the cloud. To

access the system, users must be provisioned into a Dynamics 365
environment and should have a valid Azure AD account in an autho-
rized tenant. Azure AD is built to work for apps in the cloud, on mobile,
or on-premises, and delegating authentication and authorization to
enable scenarios such as the following (Figure 12-11):
Fig.
12-11

Access control: Authentication for online services

PC and device

Express route

Multifactor
authentication

Azure Active Directory

Azure Active
Other directory services Microsoft 365 Dynamics 365 Directory
Azure Active Directory

Implementation Guide: Success by Design: Guide 12, Security 301

 ▪ Conditional access policies that require a user to be in a specific
location
▪ Multifactor authentication (MFA)
▪ Single sign-on (SSO), in which a user can sign in once and then
be automatically signed into all the web apps that share the same
centralized directory

Azure AD provides fully secured identity federation with Active

Directory on-premises. Federation with Azure AD or Microsoft 365
enables users to authenticate using on-premises credentials and access
all resources in the cloud.

Authentication: Conditional access

Users can access cloud applications from anywhere and from any de-
vice such as mobile phones, tablets, or laptops from the office or home.
As a result, there are many scenarios in which access control decisions
shouldn’t be made based solely on who can access a resource; you also
need to consider how a resource is accessed. With Azure AD condition-
al access, you can address this requirement.

With conditional access, you can implement automated access control

decisions for accessing your cloud apps that are based on conditions
(Figure 12-12).

Fig.
12-12

Conditions Actions
Cloud
User / group Allow access applications

Cloud application
Device state On-premises
Enforce MFA
Location (IP range)
User per user/per
Client application application
Sign-in risk

Block access

Authorization
Authorization is the control of access to the Dynamics 365 applications.

Implementation Guide: Success by Design: Guide 12, Security 302

 It allows you to control the functionality and data accessible to the
users. Dynamics 365 uses security roles for authorization. Once a
user is authenticated, the security roles assigned to a user (or groups
of users) authorizes the users for access to data, services, menus, or
other Dynamics 365 features and capabilities. Note that the concept
of security roles is different for Customer Engagement apps (based
on Microsoft Dataverse and Power Apps) and Finance and Operations
apps. We examine this difference later in this chapter.

Auditing and monitoring

All Dynamic 365 applications provide audit functionality. Auditing
provides a log of events related to activity for secure datastores, logins,
or other important events. As with authorization, there are differences
related to auditing in Customer Engagement apps and Finance and
Operations apps, which we examine later in the chapter.

Customer responsibility
As a customer, you’re responsible for:
▪ Account and identity management
▪ Creating and configuring conditional access policies
▪ Creating and assigning security roles
▪ Enabling and configuring auditing and monitoring
▪ Authentication and security of components of the solutions other
than Dynamics 365

Security for Dynamics 365 integrations

Most Dynamic 365s implementations have integrations with a range of
Azure services, on-premises systems, and cloud systems (Figure 12-13).

Consider the out-of-the-box integrations before building custom solu-

tions. The native integration with the other Microsoft services is a tried,
tested, and recommended solution for these scenarios. The following
are examples of native integration options available in Dynamics 365:
▪ Dynamics 365 and Microsoft SharePoint integration
▪ Dynamics 365 and Microsoft Exchange Online integration

Implementation Guide: Success by Design: Guide 12, Security 303

 ▪ Dual-write for integration between Customer Engagement apps
Fig.
and Finance and Operations apps.
12-13

Dynamics 365 (SaaS) Azure (PaaS) Office 365 (SaaS)

(Customer’s subscription)

Dynamics 365 Sales SharePoint Online

Dynamics 365 Customer Service Azure SQL for use with the Data Export Service Exchange Online
Azure Service Bus
Azure BLOB Storage

Azure Services Azure (IaaS)

Used by Dynamics 365 internally (Customer’s subscription)

Full text search Virtual machines used for OP2OL migration

Azure SQL

On-premises systems Other cloud systems

(Customer’s system) (Customer’s subscription)

Use server-to-server (S2S) authentication to securely and seamlessly

integrate Dynamics 365 with your custom websites and services. Some
of the native integrations we mentioned such as Dynamics 365 and
SharePoint also use S2S authentication.

Regarding your on-premises systems, Microsoft recommends you

deploy an on-premises data gateway connected to an Azure hybrid
connection where possible.

Azure Service Bus is a fully managed enterprise message broker with

message queues and publish-subscribe topics. Service Bus is used to
decouple applications and services from each other. In many cases,
we recommend integrating line-of business applications with cloud
services by using Service Bus, which provides a secure channel for
communicating between different on-premises or cloud-based line of
business applications.

Implementation Guide: Success by Design: Guide 12, Security 304

 Many customers implement an allowlist for IP ranges that are per-
mitted for outbound connections. Modern cloud services such as
Dynamics 365 don’t have static and specific IP ranges for Dynamics 365
solutions. We recommend you include the whole Azure region on the
IP allowlist for the system to work properly. Note that the list of IPs to
allow is considerably shorter for Azure Logic Apps. In addition, the use
of the on-premises data gateway previously mentioned can reduce or
eliminate this need.

Transparency
Microsoft is transparent about where your data is located. You know
where your data is stored, who can access it, and under what condi-
tions. Dynamics 365 customers can specify the Azure datacenter region
Fig. where their customer data will be stored. Microsoft may replicate
12-14
customer data to other regions available within the same geography
for data durability, except in specific scenarios, such as the following:
Transparency goals ▪ Azure AD, which may store AD data globally
Choose where your data is stored ▪ Azure multifactor authentication, which may store MFA data globally
▪ Customer data collected during the onboarding process by the
Transparent about how we respond
to government requests for your data
Microsoft 365 admin center

Microsoft imposes carefully defined requirements on government

and law enforcement requests for customer data. As described at
Implementation details the Microsoft Privacy Principles, if Microsoft receives a demand for
You select the region where your a customer’s data, we direct the requesting party to seek the data
Dynamics 365 data is stored directly from the customer. If compelled to disclose or give access to
any customer’s data, Microsoft promptly notifies the customer and
No third-party (including law
enforcement, other government provides a copy of the demand unless legally prohibited from doing so.
entity, or civil litigant) is given direct
or unfettered access to customer
Figure 12-14 lists some standard transparency goals and their imple-
data except as you direct
mentation in Dynamics 365.

Implementation Guide: Success by Design: Guide 12, Security 305

 Customer Engagement
apps security
The Customer Engagement apps (Dynamics 365 Sales, Dynamics
365 Customer Service, Dynamics 365 Field Service, Dynamics 365
Marketing, and Dynamics 365 Project Service Automation) use the
Microsoft Dataverse to provide a rich security model that can adapt to
many business scenarios.

The goals of Dataverse security model are as follows:

▪ Provide users with access only to the appropriate levels of infor-
mation that is required to do their jobs
▪ Categorize users by role and restrict access based on those roles
▪ Support data sharing so that users and teams can be granted access
to records that they don’t own for a specified collaborative effort
▪ Prevent a user’s access to records the user doesn’t own or share

Security features
We use three main categories of security features to provide appro-
priate end-user access (Figure 12-15): fundamental security controls,
additional security controls, and manual sharing. Most of the security
requirements should be addressed using fundamental security con-
trols; other options should be used to manage the exceptions and
edge scenarios.

Record ownership
Dataverse supports two types of record ownership:
▪ Organization owned  When a record is assigned to
Organization, everyone in the environment can access the record
▪ User or Team owned  If not assigned to the organization, a record
is assigned to Business Unit, Child Business Unit, Team, or User

Some out of the box tables are exceptions to the above 2 types such as
system user record is owned by a Business Unit.

Implementation Guide: Success by Design: Guide 12, Security 306

 Fig.
12-15

Manually sharing records Manually handles exceptions to the model

Additional controls
Hierarchy security Handles exceptions to the
Field-level security fundamental security
Access teams controls more easily
Table relationships behavior

Fundamental security controls

Business units hierarchy Generally covers
Security roles most requirements
Users and teams
Record ownership

Environment-level security group configuration

Business units
Business units are a security modeling building block that helps in
managing users and the data they can access. The name “business
unit” can be misleading because the term doesn’t necessarily have any
direct relationship with an organization’s operating business units. In
Dataverse, business units provide a framework to define the organiza-
tional structure of users, teams, and records. Business units group users
and teams by organizational hierarchy and can work in conjunction
with security roles to grant or restrict access to data.

The real power of business units comes from their hierarchical nature.
Users can be given access to records just in their business unit, or their
business unit and the business units under their unit. For example, the
hierarchical nature of business units can allow you to limit access to re-
cords at the site, district, region, and corporate levels. Business units are
useful to segment data into ownership containers for access control.

Implementation Guide: Success by Design: Guide 12, Security 307

 New changes in Business units
Recently, the Dynamics 365 team has introduced some changes to
modernize the business unit’s security model. These changes are
scheduled for general availability at the end of September 2021. Until
now, only the security roles from the user’s business unit could be
assigned to a user. The new changes will allow security roles from
different business units to be assigned to a user. This allows a user to
access data from different business units based on their security roles.

This change will address some of the following limitations in the

current model.
▪ Users are required to move to the appropriate business unit (BU)
upon creation in Dataverse (requires special coding/handling).
▪ To allow users to access cross-BU data outside of their BU, owner
Teams are required.
▪ User-owned records follow the user to the new BU when a user
moves from one BU to another BU.
▪ When an AAD group team is used to manage user access, it cannot
As a best practice, minimize the number
be used across different BUs.
of business units in your implementation
based on access control requirements ▪ Existing BU is focused on hierarchical organization structure which
instead of mapping the organizational
chart into business units.
makes matrix organization data access structure challenging to
implement.

Security roles
A privilege is permission to perform an action in Dynamics 365. A
security role is a set of privileges that defines a set of actions that can
be performed by a user. Some privileges apply in general (such as the
ability to use the export to a Microsoft Excel feature) and some to a
specific table (such as the ability to read all accounts).

To access Dynamics 365, users must be assigned one or more pre-

defined or custom security role, either assigned directly to their user or

Customer Engagement apps come with

inherited from a team they’re a member of.
several out-of-the-box security roles. If
you need additional or different security
roles, start by copying existing ones. For example, the salesperson role is assigned a set of privileges that are
relevant to the performance of the tasks defined for that role.

Implementation Guide: Success by Design: Guide 12, Security 308

 Teams
Teams provide a straightforward way to share records, collaborate with
other people across business units, and assign security roles. While
a team belongs to one business unit, it can include users from other
business units. You can associate a user with more than one team, which
is a convenient way to grant rights to a user that crosses business units.

Multiple types of teams available are in Dataverse:

▪ Owner teams  You can make an owner team the owner of a record,
which is a useful way to link a record to a specific business unit. Owner
teams can also share records and be assigned to security roles.
▪ Azure AD security group teams and Azure AD office group
teams  These are special types of owner teams. Their member-
ship can’t be managed within Dynamics 365, instead they’re linked
to an Azure AD security group or Microsoft 365 groups. Provided
that the users are appropriately licensed and part of the environ-
ment security group (optional), users who are added to the Azure
AD group are automatically enabled in the system and added to
the team when they connect to the environment. Using Azure
AD is particularly useful because the groups and permissions also
Use owner teams when:
▪ Team-level record ownership is extend to Azure AD-enabled apps outside of Dynamics 365.
required, rather than user ▪ Access teams  These teams are unique because they can’t own
▪ You need to assign security roles to
users through teams records and can’t have a security role association. However, just like
▪ Reporting on team progress is required
owner teams, access teams can have records shared with them. When
Use access teams when:
▪ The team members need different enabled at the table level, access teams can grant specific record-level
permissions on individual records than
the same permission on the record type
permissions to the members of a record’s access team. This is an
alternative to manually sharing the record with a user or a team.

Field-level security
You can use field-level security to restrict access to high business im-
pact fields to specific users or teams. For example, you can enable only
certain users to read or update the credit score of a business customer.

Field-level security is available for the default fields on most out-of-the-

box tables, custom fields, and custom fields on custom tables. Field-level
security is managed by the security profiles. The scope of field-level
security is organization-wide and applies to all data access requests.

Implementation Guide: Success by Design: Guide 12, Security 309

Sharing
Record sharing lets a user give access to a table record to other users or
team. The user must have a share privilege on the record to be able to
share it. Sharing should be seen as a way for users to manually manage
exceptions to the default security model.

Keep in mind the following recommended practices:

▪ Share only the necessary records with the smallest set of users or
teams possible
▪ Grant the minimum access required for users to do their jobs
▪ Disable share rights in security roles where possible
▪ Extensive sharing can cause performance issues at scale
▪ Microsoft recommends that you don’t automate sharing

Hierarchy security
You can use a hierarchy security model for accessing data from a user
or position hierarchy perspective. With this additional security, you
gain more granular access to records, for example by allowing man-
agers to access the records owned by their reports for approval or to
perform tasks on reports’ behalf.

Summary of different access control mechanisms

Figure 12-16 summarizes the different access control mechanisms
discussed in this section. It highlights the ideal scenarios, benefits, and
limitations of each option.

Environment security group

You can associate an Azure AD security group with a Dataverse en-
vironment to control user access. Unless users have specific, highly
privileged Microsoft 365 admin roles, a user in Azure AD can’t access
any information in Dynamics 365 even with a valid authentication and
security role assigned unless they are also a member of the correct
environment security group in Dynamics 365.

It’s a best practice to associate a security group with an environment.

This prevents all the Azure AD-eligible (appropriately licensed) users
from appearing as active users in the Dynamics 365 environment.

310
 Fig.
12-16

Bypasses need for

For open access
Organization information to all
security checking, so
optimal performance

Can be used for

Great for large volumes Reduces management and
Allows for optimal maintenance
Business unit of data accessed by large
groups of users
access checking overhead
oversight access with
more granular access
at lower levels

Ideal for large Also drives business

Optimized access Limited to one
volumes of overall unit position so
checks combine type of access i.e.
User ownership data but individual
memory cache and owned by single enables combination
granular access to business unit
record data directly user or team
smaller subset access

Ideal for large Isolate users from

Optimized access Reduces cascading
volumes of overall record business unit
checks combine impact as user
Team ownership data but group memory cache and involvement
structure, enables
granular access to user movement
record data directly changes
smaller subset without data impact

But comes with cost,

Ideal for modeling Provide very checking lots of rules
exceptions over specific unique expensive in perfor- Sharing with teams
Sharing standard model using permissions per mance, storage, and mitigates cost
other features record maintenance
overhead

Use to support Not intended to be

Builds on ownership immediate replacement for
Inheritance and sharing models managers in chains buisness unit access
above granular at more senior or
indivual access broad scope levels

Audit
Auditing helps you comply with internal policies, government regula-
tions, and consumer demands for better control over confidential data.
Organizations audit various aspects of their business systems to verify
that system and data access controls operate effectively and identify
suspicious or non-compliant activity.

Dataverse audit
Audit logs are provided to ensure the data integrity of the system and
to meet certain security and compliance requirements. The auditing

Implementation Guide: Success by Design: Guide 12, Security 311

 feature logs the changes that are made to records and user access so
the activity can be reviewed later.

Auditing can be enabled on:

▪ Table level (you can select specific fields in a table for auditing)
▪ User access

Data auditing can be viewed:

▪ Contextually, at the record level, on the form
▪ Globally, from the audit summary view
▪ From the APIs

Don’t enable auditing for all tables and columns. Do your due dili-
gence to determine which tables and fields are required for auditing.
Excessive auditing can affect performance and consume large volumes
of log storage.

Microsoft 365 audit log

Activity logging records user and admin activities across Microsoft 365
and Dynamics 365 apps. Data is stored in a Microsoft 365 unified audit
log. It has a smaller footprint on system resources compared to the
previous Dataverse audit functionality. Microsoft 365 audit logs have
several additional benefits:
▪ Management in a central location on the Microsoft 365 security
and compliance portals  Microsoft 365 admins can manage
settings and access activity reporting for all environments within the
Microsoft 365 security and compliance portals. Dataverse auditing is
set up and stored separately within each Dynamics 365 environment.
▪ All data in the system is logged  This functionality logs all data
transactions, including read operations, plug-in operations, table
operations, bulk operations, user sign in and sign out sessions, and
even Microsoft Support operations.
▪ Configurable alert policies  You can set up the system to notify
administrators or compliance officers of certain events according
to configurable settings.
▪ Audit log search capability  Administrators can easily query
audit logs via predefined or custom filters.
▪ Analyze suspicious behavior with Security Information and

Implementation Guide: Success by Design: Guide 12, Security 312

 Event Management (SIEM)  Use SIEM functions in near-real
time to analyze and alert administrators of possible suspicious
behavior within the system and provide actions to address these
events.
▪ SIEM vendor integration  Dynamics 365 now provides out-of-
the-box integration with multiple SIEM vendors such as ArcLight,
Microsoft OMS, Okta, SumoLogic, BetterCloud, and standard CEF
format integration for many others. Microsoft has also released a
connector to integrate activity logs into Azure Sentinel.

By default, activity logging data is retained for 90 days or 1 year based

on the Microsoft 365 license type. We recommend changing the reten-
tion policy or moving these logs to external stores if you need to retain
the data for longer periods.

Power Apps Portals

Security
Power Apps Portals allow internal and external users access to the
Dataverse data through external facing websites. As of the 1st of
October 2019, Dynamics 365 portals are referred to as Power Portals.
Power Apps Portal supports Dynamics 365 apps along with standalone
power apps. Portal data can be exposed anonymously or only to
authenticated users. An example of anonymous access is an unauthen-
ticated landing page or home page. Security in the Power Apps portal
has two parts: authentication and authorization.

Authentication
For users to access any Power Apps portal, they must exist in Dataverse
as contacts. This applies to both internal and external users. Power
Apps portals support Azure AD, Azure B2C, ADFS, and other third-par-
ty providers such as LinkedIn, Facebook, and Google. You can find
Authentication configuration details and the complete list of identity
providers at Get started with configuring your portal authentication -
Power Apps | Microsoft Docs.

313
 Sign-up
There are two common ways to control sign-ups for the Power App portals:
▪ Azure B2C is the preferred ▪ Open Registration is the least restrictive option to sign up for the
authentication provider for portals.
portal access. Configuration portal allows a user account to be
It separates authentication from
authorization. registered by providing a user identity. A new contact is created in
▪ Azure B2C supports third-party
authentication providers such as
Dataverse on sign-up.
LinkedIn, Facebook, Google, and ▪ The alternative option is by invitation. The invitation feature of
many more with custom policies. Use
Azure B2C as a bridge to other Identity portals allows you to invite contacts to your portal through auto-
providers as it will support more options mated email(s) created in your Microsoft Dataverse. The people
and Microsoft won’t be duplicating
these investments in the portal. you invite receive an email, fully customizable by you, with a link
▪ Local Authentication is deprecated but
to your portal and an invitation code.
not removed yet. It cannot be extended
to other systems like Azure B2C and
Microsoft is not investing in new local
authentication features. Its use is limited
Authorization
and short-lived. Authorization is a control to provide access to data and web pages in the
▪ For B2B scenarios, consider guest users
with Azure AD authentication. What Power Apps portal. The authorization is managed through web roles.
is B2B collaboration in Azure Active
Directory? | Microsoft Docs

Web Roles
Web roles allow portal users to perform special actions and access
protected Dataverse contents. It’s similar to security roles in Dynamics
CE apps. A contact can have multiple web roles.

A Power Apps portal website can have multiple web roles, but can
only have one default role for authenticated users and one for anon-
ymous users. Web roles control the following:
▪ Dataverse table permissions allow access to individual records in
the Dataverse tables. It allows you to set a scope for access such as
global, contact level, account level, parental level, etc. Customers
can control the granular access on a record such as read, write,
delete, append, and append to.
▪ Page permissions allow access to portal webpages. For example,
you can allow pages to be available anonymously for public ac-
cess, or restrict access to users who have specific roles.
▪ Website access permissions allow portal users to manage front
side editing of some portal contents, such as content snippets and
weblink sets.

Implementation Guide: Success by Design: Guide 12, Security 314

 Figure 12-17 depicts portal security using Azure B2C authentication
Fig. for external users.
12-17

Dynamics 365 Power Apps Portal/

Dynamics 365 Portal Authentication
Dataverse
Contoso Portal

Contacts Contoso employees Contoso Azure AD

Portal users Azure AD users “Contoso employees”
Azure AD
Stores portal authentication App registration
identification information (automatic)

BSB users Used by Office 365 users,

Dynamics 365 users,
Authentication

Guest Azure AD users

B2B guest users

Web roles Azure B2C Directory

External users
(External users)
Authenticated users (default) Azure BSc log-on
Azure B2C
Custom role 1 App registration Login options:
(manual)
Custome role 2 • Azure B2C users (local accounts
with name, email address, etc.)
More roles...
• LinkedIn, Facebook, and other
third-party providers
Sign up Sign up • My work Azure AD accounts
Invitation redemption (example: [email protected])
Invitation code
(Sent through mails) Open registration • Custom Identity Provider (SAML,
ADFS, etc.) using custom policies

Best Practice
Use Portal Checker in your portal deployments. It lists forms, entity
lists, and OData feeds exposed anonymously. Make sure no data is
exposed to anonymous users by mistake.

Other security options

▪ Azure AD and Azure B2C supports conditional access policy. You
can use conditional access to control the access to the portal.
▪ Power Apps portals allow the administrations to setup IP restric-
tions to the portal. Restrict portal access by IP address - Power
Apps | Microsoft Docs

Implementation Guide: Success by Design: Guide 12, Security 315

 ▪ Power Apps portal also allow customers to use Web Application
Firewall (WAF) to enhance the perimeter security.

Finance and Operations

apps security
Finance and Operations applications use Azure AD as a primary identi-
ty provider and utilize the security capabilities of Azure. To access these
applications, users must have a valid Azure AD account in an autho-
rized tenant, and must be provisioned into a Finance and Operations
application instance. Once authenticated, these applications use role-
based security to authorize user access to individual elements of the
applications. Additionally, you can use extensible data security policies
to restrict access to a subset of data.

The security architecture of Finance and Operations apps consists of

the following components (Figure 12-18):
▪ Authentication  Only authenticated users with user rights in
Finance and Operations apps can connect
▪ Authorization  Security roles are comprised of duties, duties are
comprised of privileges, and privileges are comprised of permissions
▪ Security roles  Roles grant access to Finance and Operations apps
▪ Duties  Duties allow access to parts of business processes
▪ Privileges  Privileges allow access to individual tasks
▪ Permissions  Permissions grant access to individual securable
objects
▪ Data security  You use data security to deny access to tables
fields, and rows in the database (extensible data security framework)
▪ Auditing  The system logs when a user signs in or out of the
application

Role-based security
In Finance and Operations apps, role-based security is aligned with
the structure of the business. Users are assigned to security roles based
on their responsibilities in the organization and their participation in

316
 business processes. Because you can set up rules for automatic role
assignment, the administrator doesn’t have to be involved every time
a user’s responsibilities change. After business managers set up secu-
rity roles and rules, they can control day-to-day user access based on
business data. A user who is assigned to a security role has access to
the set of duties that are associated with that role, which is comprised
of various granular privileges. A user who isn’t assigned to any role has
no privileges. Privileges are composed of permissions and represent
access to tasks, such as cancelling payments and processing deposits.
Duties are composed of privileges and represent parts of a business
process, such as maintaining bank transactions. Roles are assigned to
users in either all legal entities or within specific legal entities to grant
Fig.
access to Finance and Operations application functionality.
12-18

Authentication (Azure Active Directory)

Users

Authorization
Security roles

Duties Privileges

Permissions for application elements

User interface
Tables and fields SSRS reports Service operations
elements

Data security Data security Table permissions

policies framework

Database
Database

Implementation Guide: Success by Design: Guide 12, Security 317

 By default, approximately 100 standard security roles are provided
within the solution. All functionality is associated with at least one of
the standard security roles. The administrator can duplicate and mod-
We recommend duplicating the sample
security roles and using these for role
ify the sample security roles to fit the needs of the business or create
assignments for all users before you start entirely new security roles.
creating custom roles.

Security diagnostics for task recordings

The Task recorder for Finance and Operations apps is a utility that lets
users record business processes for several different use cases. This is
an invaluable tool for many purposes, including process definition,
training, testing, and security. The security diagnostics tool for the
Task recorder allows implementers to analyze and manage security
The standard security roles included in permission requirements in any task recording. When you use the Task
Finance and Operations apps incorporate
segregation of duties. If a user needs full recorder to define and document business processes for testing or
access, they often need a combination training purposes, the security requirements for implementing those
of roles (such as an HR assistant or HR
manager for full access to HR features). tasks are also captured. Implementers can then use this tool to tailor
duplicated copies of the sample security roles to better meet their needs.

Segregation of duties
Some organizations may require some tasks to be separated and per-
formed by different users. For example, you might not want the same
person to maintain vendor information, acknowledge the receipt of
goods, and process payments to the vendor. This concept is sometimes
referred to as separation of function, and the standard security roles
included in Finance and Operations applications have been created
with this separation in mind. To cite the previous example, different
security role assignments are required in Dynamics 365 Finance to
maintain vendor information, acknowledge the receipt of goods, and
process payments to the vendor.

In addition to the separation of functions described above, Finance

and Operations applications also facilitate the segregation of duties.
Segregation of duties helps you reduce the risk of fraud, enforce
internal control policies, and detect errors or irregularities by identi-
fying specific duties in your organization that must not be performed
by the same personnel and creating rules to detect and prevent the

Implementation Guide: Success by Design: Guide 12, Security 318

 assignment of security roles that would conflict with those rules.

When an organization defines rules for segregation of duties, you should

evaluate existing role definitions and role assignments for compliance,
thereby preventing role definitions or role assignments that don’t com-
ply. Future role assignments that result in conflicts with existing rules are
logged and must be allowed or denied by an administrator.

Security reports
Finance and Operations applications provide a set of rich security
reports to help you understand the set of security roles running in your
environment and the set of users assigned to each role. In addition to
the security reports, developers can generate a workbook containing
all user security privileges for all roles.

User role assignments report

Generates a view of the current user role assignments in your system.
By default, the report includes all users with roles assigned. You can
optionally limit the report to a specific set of users by entering a list of
users when generating the report. In the User role assignments pa-
rameters pane, under Records, choose Filter. From here you can add or
remove filters to the list of users the report will be generated for.

Role to user assignment report

Provides an aggregation of role assignments. Expanding a role in the
report shows the list of users assigned to the role, and expanding the
username shows any restrictions the role has applied. You can apply
the same method for filtering the set of users to this report as de-
scribed for the User role assignments report.

Security role access report

Provides a view of the effective permissions for each security role. This
report provides a flattened list of permissions grouped by type across
all sub-roles, duties, and privileges contained in the role.

Security duty assignments report

Provides a view of all the duties contained within a role. You can

Implementation Guide: Success by Design: Guide 12, Security 319

 configure this report to run on any collection of roles to ensure that
segregation of duties is maintained between roles. By default, the
report includes all the roles. To limit the roles included, use the filtering
Familiarize yourself with these reports
during implementation and engage your
provided in the Records to include section.
compliance team early. There is no benefit
to waiting for an audit to use these reports
for the first time. To aid compliance with GDPR requirements, you can use the per-
son search report when a data subject (the individual EU citizens or
residents that share personal data) approaches the controller (the
organization collecting and managing personal data) to request a copy
of their personal data.

System log
System administrators can use the User log page to keep an audit
log of users who have logged on to the system. Knowing who has
logged in can help protect your organization’s data. The user logging
capability allows administrators to identify roles that provide access to
sensitive data. The sensitive data identifier enhances the user logging
experience by letting your organization produce audit logs that show
who in your system has access to sensitive data. This capability is help-
ful for organizations that might have multiple roles that have varying
degrees of access to certain data. It can also be helpful for organiza-
tions that want a detailed level of auditing to track users who have had
access to data that has been identified as sensitive data.

Extensible data security policies

Extensible Data Security (XDS) policies in Finance and Operations apps
allow developers to supplement role-based security by restricting ac-
cess to table records based on security policies. Policies can be created
that contain a query to select information and a table or tables that
will be constrained by the policy. The query in the policy applies a filter
and only records that satisfy the conditions of the filter are accessible
from the restricted tables. XDS provides a powerful data security tool
for advanced requirements, but you should use it judiciously to avoid
unanticipated impacts on performance or usability.

Implementation Guide: Success by Design: Guide 12, Security 320

 Database logging
Database logging provides a way to track specific types of changes to the
tables and fields in Finance and Operations apps, including insert, update,
delete, and rename key operations. When you configure logging for a
table or field, a record of every change to that table or field is stored in the
database log table (sysdatabaselog) in the environment database.

You can use database logging to create an auditable record of changes

to specific tables that contain sensitive information, or to monitor the
use of electronic signatures. Note that this feature is in addition to
audit trail capabilities already available in many modules. Database
logging is intended to track individual changes to specific, sensitive
information or configurations. It’s not intended to track automated
transactions that are run in batch jobs and is not recommended for
transactions. Although database logging can be valuable from a busi-
ness perspective, it can be expensive with regard to resource use and
management, so it too should be used judiciously.

Microsoft recommends the following when implementing database

logging:
▪ Create a plan for reviewing and using your database logs
▪ Limit log entries and help improve performance by selecting
specific fields to log instead of whole tables
▪ Create a plan for how long you will retain logged data and use the
included cleanup job

Make security
a day one priority
Security should be at the forefront when you start a Dynamics 365
project, not an afterthought. Negligence of security requirements can
lead to significant legal, financial, and business risks. It can also impact
the overall scalability and performance of the solution. In this section,
we discuss some of the security impacts on Dynamics 365.

Implementation Guide: Success by Design: Guide 12, Security 321

 Impact on scalability
Security design can have a significant impact on scalability. Security
design should consider scalability especially for multi-phased and
global deployments.

Customer Engagement examples

In multi-phased global deployments, short-sighted security design can
affect the rollout to other regions. The different regions can have their
own compliance, business processes, and security requirements. The same
model may not work for all the regions. Therefore, you should think about
the security model holistically at the beginning of the project.

One such example can be a global deployment where in the first

region (Phase 1) you create all users in the root business unit with
business unit-level access in security roles. Because Phase 1 users were
placed in the root business unit, when you roll out the solution to the
next region (Phase 2), Phase 2 users have access to the users in the first
region. This may cause some regulatory or compliance issues.

Several businesses use Organization owned entities for reference

data tables, which means security roles don’t apply to these entities.
In multi-phased deployments, some departments may not want their
reference data to be seen by other departments, or simply only want
to see their reference data appearing when creating records. This
segmentation is hard to achieve using Organization owned entities
and much simpler to achieve when User or Team owned entities have
Business Unit access. Entity ownership is a choice that happens at the
time the entity is created and can’t be changed. This assignment con-
trols who can access the entity. If you want to change who has access
to these entities, the only option you have is to delete these entities
Security should and recreate them in an owner (Organization, Team, Business Unit, and

be at the forefront so on) that that is properly scoped.

when you start Finance and Operations examples

a Dynamics 365 Similarly, for Finance and Operations implementations, different
project, not an legal entities may have distinct functions and purposes, be subject to
different security constraints, and have completely different security
afterthought.
322
 requirements. Even if legal entities, subsidiaries, or regions have a high
degree of standardization and centralization and are good candidates
for streamlined security implementation, failing to source require-
ments from the right stakeholders, design security as modular building
blocks, or otherwise fail to take a long-term view of security needs can
result in subsequent rollouts struggling with ill-fitting global templates
or a need to start over from scratch.

Design of role-based access control should follow best practices, or

subsequent rollouts will be unable to use and further build upon
established security roles. Security roles should be designed based on
business processes and incorporate segregation of duties, not individ-
ual personas. We recommend that you avoid creating a single role for
each user and avoid creating nested security roles. The resulting mod-
ular security roles can be assigned to users as they cumulatively give
the access needed for a users’ duties, and no more. Native tools, such
as the Task recorder and Regression suite automation tool, support the
testing of individual user permissions, which facilitate this approach.
We also recommend following recommended practices for the initial
rollout (such as copying standard roles rather than directly assigning
them). This gives subsequent rollouts the flexibility of building upon
existing role structures, while retaining access to standard roles for
comparison. This prevents the subsequent rollouts from being forced
to choose between modifying roles already in use elsewhere or starting
from scratch because existing roles are too specialized.

Impact on performance
Security design can impact system performance. Dynamics 365 has mul-
tiple access control mechanisms, all with their own pros and cons. The
ideal security design is a combination of one or more of these controls
based on your requirements and the granularity, simplicity, and scalabili-
ty they provide. The wrong choice can lead to poor performance.

Customer Engagement examples

Sharing provides particular and unique permissions on a specific
record, but sharing for large volumes of records comes at a cost.
To check access for a user, the system checks user security roles,

Implementation Guide: Success by Design: Guide 12, Security 323

 team memberships, and records shared with a user and their teams.
Checking lots of rules is expensive in terms of performance, storage,
and maintenance overhead.

Sharing can also trigger cascading rules that can be expensive if not set
appropriately. Sharing should be for edge cases and exception scenarios.

You can get more information on this topic from the Scalable Security
Modeling with Microsoft Dynamics CRM whitepaper. It’s an older
document, but still relevant for Dynamics 365 security.

The volume of data will impact performance in the long run. You
should have a well-defined archival strategy. The archived data can be
made available through some other channels.

Finance and Operations examples

Database logging in Finance and Operations can be valuable from a
business perspective, but it can be expensive regarding resource use
and management. Here are some of the performance implications of
database logging:
▪ The database log table can grow quickly and increase the size of
the database. The amount of growth depends on the amount of
logged data that you retain.
▪ When logging is turned on for a transaction type, each instance of
that transaction type causes multiple records to be written for the
Microsoft SQL Server transaction log file. Specifically, one record
is written for the initial transaction, and one record logs the trans-
action in the database log table. Therefore, the transaction log file
grows more quickly and might require additional maintenance.
▪ Database logging can adversely affect long-running automated
processes, such as inventory close, calculations for bills of materi-
als, master planning, and long-running data imports.
▪ When logging is turned on for a table, all set-based database op-
erations are downgraded to row-based operations. For example, if
you’re logging inserts for a table, each insert a row-based insert.

When you’re implementing database logging, be sure to have a plan

for how long you’ll retain logged data, and how you’ll limit log entries

Implementation Guide: Success by Design: Guide 12, Security 324

 and help improve performance by selecting specific fields to log in-
stead of whole tables.

Extensible Data Security policies allow developers to supplement role-

based security by restricting access to table records based on security
policies. The query in the policy (the policy query) applies a filter so
only records that satisfy the conditions of the filter are accessible from
the restricted tables. The policy query is added to the WHERE clause, or
ON clause, on SELECT, UPDATE, DELETE and INSERT operations involv-
ing the specified constrained tables.

Unless carefully designed and tested, policy queries can have a signif-
icant performance impact. Therefore, make sure to follow the simple
but important guidelines when developing an XDS policy. Performance
can be significantly impacted by poorly performing queries. As a best
practice, avoid numerous joins, consider adding lookup tables to im-
prove performance, carefully tune queries for optimal performance, and
carefully and thoroughly test any XDS policies when using this feature.

Impact on regulation and compliance

Several different security controls are required to meet regulations and
compliance based on industry and region. These controls are a combi-
nation of duty separation, data categorization, identity management,
access control, log management, and auditing. Missing any of these
controls can make your solution noncompliant.

Governance takes time to understand, roll out, and adopt. The earlier
Governance you begin the process, the easier compliance becomes.

takes time to
Customer Engagement examples
understand, roll The activity logging data is retained between 90 days to 1 year based
out, and adopt. The on the Microsoft 365 license type. For compliance reasons, if you need
earlier you begin to keep these records for longer, you should move this data to an
external store.
the process, the
easier compliance
becomes.
325
 Impact on rollout
In many projects, security design is created late in the implementation
process. It can lead to a lot of issues, ranging from system design to in-
adequate testing. Keep the following recommended practices in mind:
▪ Test the system using the right security roles from the very beginning.
▪ Use the correct security profile during test cycles. Don’t provide an
administrator role to everyone.
▪ When training people, make sure they have access to make all the
required actions. They shouldn’t have access to more information
than required to complete those actions.
▪ Validate security before and after data migration occurs.

Finance and Operations examples

Design your security roles with both compliance and scalability in
mind. By engaging your security and compliance team early in your
implementation, you can use the segregation of duties framework to
identify conflicts and create rules that enforce separation of duties that
should be separated.

For example, a global company has made efforts to standardize

business processes and is planning to roll out a new functionality in
multiple phases to globally. Rules that segregate duties in your large
legal entities prevent assignments that may result in conflicts.

However, in smaller legal entities, users may need to accomplish the

same tasks with a smaller staff, which makes it challenging to construct
declarative rules that enforce segregation of duties. In this case, modu-
lar security roles allow you to selectively enforce the same segregation
of duties rules when assigning multiple roles to some users by applying
documented overrides to the detected conflicts. In this manner, your
global rollout can reuse modular security components, and you have
a robust segregation of duties framework in place for entities of many
sizes, allowing them to grow without introducing the need to redesign
the security implementation.

Implementation Guide: Success by Design: Guide 12, Security 326

 Impact on reporting
Security design can have a significant impact on reporting. For example,
your security design identifies sensitive data that requires restricted
access to be enforced and monitored. You need to create auditing
reports related to access and changes on sensitive data, and designated
personnel responsible for oversight need to review them periodically.

In addition to built-in reporting capabilities, you have multiple options

for exporting Dynamics 365 data into external stores for reporting and
analytics. The two native options are exporting to a data lake or using
the Data Export Service (not available in Finance and Operations apps).
When data is in external stores, it doesn’t honor Dynamics 365 security
permissions (nor would it be useful to do so in many cases). You need
to apply your own security mechanism for reporting. We recommend
having a data management strategy for all data that is stored outside
of a core application. This strategy should include security require-
ments (encryption, access controls, row level, personally identifiable
information) as well as data residency and retention requirements. The
strategy should consider the future state of Dynamics 365 data in the
analytical platform as well as its legacy and other applications.

You can use Tabular Data Stream (TDS) endpoints with Customer
Engagement for direct query scenarios, which honor the Dynamics 365
security context.

Operational aspect of security

Operation aspects can have a big impact on the security design. Operation
aspects can decide which access mechanisms are the best fit for the
solution, such as in cases of organizational changes or maintenance.

Organizational changes
What happens if a user moves to another team or business unit? Do
you need to need to reassign the records owned by the user? What
should be the process?

Do you reassign all records or just active record? What is the effect

Implementation Guide: Success by Design: Guide 12, Security 327

 of changing the ownership on the child records, especially closed
activities, closed cases, and won or lost opportunities? Based on these
questions, you can come up with a process and cascading rules in
Dynamics 365.

Reassigning a large volume of records can also take a long time and affect
the system’s performance. Learn more about assigning or sharing rows.

Maintenance overhead
Plan out how to assign security roles to new users. Are you going to
automate a process? For example, you can use Azure AD group teams
in Dynamics 365 to assign roles to new team members. This can make
it very easy to assign license and security roles to new team members.

Security anti-patterns
An anti-pattern is a frequently implemented, ineffective response to a
problem. Several security anti-patterns should be avoided for scaling,
performance, and security reasons. We shared a few in the previous
section, such as using Organization owned entities for reference data
tables in Customer Engagement, or indiscriminately logging transac-
tional tables in Finance and Operations apps. In this section, we discuss
a few more anti-patterns to avoid.

Security through obscurity

Project teams may try to implement security through obscurity. One
Azure Key Vault allows you to securely store example is hiding the fields from forms and views through custom
and manage application credentials such
as secrets, keys, and certificates in a central code. This is not security, because anyone with a proper security role
and secure cloud repository. Key Vault
can access those fields in several ways such as with the use of Advance
eliminates the need to store credentials in
your applications. Your applications can Find, Power Automate, and APIs.
authenticate to Key Vault at run time to
retrieve credentials.
Always consider all the ramifications of avoiding default security controls.

Credentials stored in Dynamics 365

This is also one of the more common anti-patterns. Projects often use

Implementation Guide: Success by Design: Guide 12, Security 328

Resources calls to external systems through plugins and JavaScript. Calling exter-
nal services through plugins is recommended, but the credentials to
General call these services are stored in the code of some configuration entities
Microsoft Dynamics CRM Online in the system. This practice creates a significant security risk because
security and compliance planning the credentials could be discovered by numerous roles not directly
guide
related to the use of the service. In addition, if the credentials change,
Trusting the Cloud the call to the external service fails. As a best practice, don’t store any
credentials in the system.

Compliance
An Introduction to Cloud
Computing for Legal and Conclusion
Compliance Professionals
In this chapter, we introduced how the Trusted Cloud is built on the
Achieving Trust and Compliance
foundational principles of security, privacy, compliance, and trans-
in the Cloud
parency, and outlined basic concepts of information security as they
apply to Dynamics 365. We then took a closer, product-specific look at
Privacy how these security concepts are applied to Customer Engagement and
Privacy at Microsoft Finance and Operations applications. With that as a foundation, we

Protecting Data and Privacy in the

then examined why it’s crucial to the success of your implementation
Cloud to make security a priority from day one, citing some specific examples
and considering the impact of security on performance. Finally, we
wrapped up with a handful of anti-patterns to avoid.
Security
Dynamics 365 Security Assessment Equipped with this information, you can be confident in the security
Security Incident Management in considerations and capabilities of Dynamics 365 products and better
Microsoft Dynamics 365 ensure the security of your implementation.
Encryption in the Microsoft Cloud

Security in Microsoft Dataverse

Security architecture

Plan and implement security in

Finance and Operations apps

Review the security model for your

Dynamics 365 solutions

329
 Checklist

Privacy and compliance Application security

Understand the responsibilities of the service provider as Understand the app-specific security constructs and
a data processer versus the customer responsibilities as model your security based on the native access control
the owner and data controller to ensure compliance on mechanisms rather than customizing the build.
both sides.
Understand that hiding information from the view
Refer to the Dynamics 365 cloud service agreements doesn’t necessarily remove access, there are alternate
and compliance documentation to understand the mechanisms to access and extract information.
policies, procedures for handling data, disaster recovery
Understand the impact of losing the security context
strategy, data residency, encryption, and more.
when the data is exported.

Identity and access Ensure the security model is optimized to perform and
provides the foundation for further expansion and scale
Establish an identity management strategy covering user by following the security model best practices.
access, service accounts, application users, along with
Have a process to map changes in the organization
federation requirements for SSO and conditional access
structure to the security model in Dynamics 365. This
policies.
needs to be done cautiously in a serial way due to the
Establish the administrative access policies targeting downstream cascading effect.
different admin roles on the platform, such as service
admin and global admin.

Enforce relevant DLP policies and procedures to make

changes or request exceptions.

Have necessary controls to manage access to specific

environments.

Implementation Guide: Success by Design: Guide 12, Security 330

13
Guide
Business
intelligence,
reporting,
and analytics
 Introduction
In a world where so much comes down to reporting
and analytics, business intelligence has become the
differentiator—it’s all about actions and outcomes,
measuring what matters the most, and empowering
users with the right insights at the right time.
Business intelligence solutions have evolved to keep up with an ev-
In this chapter, we discuss how data
helps businesses make informed
er-increasing number of data sources, as organizations seek ways to
decisions. We also cover the evolution of expand their global reach to attract more customers, and give their
business intelligence solutions, and how
organizations can build a data estate current customer base access to more content, purchasing channels,
based on modern components, on top and brand options. In this context, turning collected data into compet-
of solutions that support augmented
analytics and embedded intelligence. itive business knowledge is more important than ever.

Business intelligence solutions strengthen

organizations, prepare them for challenges While AI and machine learning keep evolving—and are now used natively
and market changes, and guide them
toward the best insights and actions.
by apps—your organization’s intelligence strategy should be based on a
modern data estate, with a focus on how to strengthen the organization.

The importance of data

As James Phillips, president of Microsoft’s Digital Transformation
Platform Group, wrote in the Microsoft Dynamics 365 blog, “There is

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 332
 a fundamental change occurring across industries and organizations:
Data now comes from everywhere and everything.” As an essential
part of a business solution, data represents a sizable portion of each
user’s engagement. The solution processes and analyzes the data to
produce information, which allows an organization to make informed
decisions, and determines actions that can come from it.

As a part of the day-to-day usage of any application, users generate

significant amounts of process-related and customer-related data—
structured and unstructured. For example, customers generate data by
visiting a website or using a specific product. In turn, businesses gen-
erate data about their products or services, as well as how customers
interact with those offerings.

In addition, the pace of data growth is increasing dramatically. Not

only does user activity create data, but new data sources such as the
Internet of Things (IoT) need to be captured, rationalized, and secured.

The key role of data is to generate information that helps businesses

and customers make informed decisions. For businesses seeking to
stay competitive, the data also must trigger actions demonstrating that
they understand and value their customers.

Resilience—the ability to adapt quickly to changing conditions—has

become a hallmark for business success today. Competitive, resilient
organizations invest in turning the data they collect into business
knowledge, build the right data estate to generate a 360-degree view
of their users, and use business intelligence to determine the best
actions and outcomes.

Organizations that have already embarked on their transformation

journey to improve the ways they use data are better prepared to
respond and adapt to change.

Break the silos and use the digital

feedback loop
Organizations are challenged by how to deal with ever-increasing

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 333
 amounts of data—and how to generate user value with it—especially
when that data becomes siloed by the systems that generate or collect
it. Data silos make the goal of having a 360-degree view of each user
even more challenging. Successful organizations are able to digitally
connect every facet of their business. Data from one system can
be used to optimize the outcomes or processes within another. By
establishing a digital feedback loop (Figure 13-1) that puts data, AI,
and intelligent systems and experiences at the core, organizations can
transform, become resilient, and unlock new values for users.

Fig.
13-1

Engage Empower
customers people

l ig
Intel ent

Data & AI
s
sy

te
ce

ms n
ie
s

an d ex per

Optimize Transform
operations products

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 334
 Evolution of business
intelligence, reporting,
and analytics
Practices for gathering, analyzing, and acting on data have evolved
significantly over time. Traditional methods of standardizing and
generating static reports no longer give businesses the agility to adapt
to change. New technology and secure, highly reliable cloud services—
which are poised to meet the needs of organizations that must quickly
manage increasing amounts of data—have given rise to a new era of
digital intelligence reporting.

Traditional reporting
The early generations of business intelligence solutions were typically
centralized in IT departments. Most organizations had multiple data
repositories in different formats and locations that were later com-
bined into a single repository using extract, transform, and load (ETL)
tools, or they would generate reports within siloed sources and merge
them to provide a holistic view of the business.

Once the data was unified, the next steps were deduplication and
standardization, so the data could be structured and prepared for re-
porting. Business users who lacked the expertise to perform these tasks
would have to rely on the IT team or specialized vendors. Eventually,
the business would receive static intelligence reports, but the entire
process could take days or even weeks, depending on the complexity
of the data and maturity of the processes in place. Data would then
undergo further manipulation, when required, and be shared across
different channels, which could result in the creation of multiple ver-
sions that would be difficult to track.

Such an approach, which also heavily depended on the business

owner’s analysis to generate business insights and actions, was often a
time-consuming and error-prone process. And if an error was found, it

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 335
had to be fixed in the central repository. The process to produce the re-
port would need to be executed again—and repeated multiple times if
the same fix wasn’t applied at the source. This method of reporting was
referred to as “closing activities,” because it typically occurred at the
end of a month, quarter, or year, which meant that organizations were
slow to respond to opportunities because they waited until the end of
a given period to analyze their data and make decisions.

Self-service reporting
The evolution to a more agile approach favored self-service capabilities
to empower users. More user-friendly solutions reduced the IT depen-
dency and focused on providing business users with access to data and
visualization tools so they could do their own reporting and analytics.
This accelerated the speed of data analysis and helped companies
make data-driven decisions in competitive environments. However, in
this model, reporting was unmanaged—increasing the number of ver-
sions of reports, as well as different representations of the data—which
sometimes prevented organizations from using a standardized method
to analyze data and inhibited effective decision-making.

This updated approach didn’t entirely replace the IT-centric model,

as many organizations started using a combination of traditional and
self-service reporting, but it did provide quicker access to information, so
organizations could react faster to business challenges and opportunities.

Reporting from anywhere

The growth of IT infrastructures, networks, and business usage of
devices such as mobile phones and tablets launched a digital trans-
formation from legacy systems into more modern solutions for most
organizations. Business intelligence apps allowed reporting from
anywhere at any time, giving users access to data while they were out
of the office or on business trips. These apps improved how organiza-
tions could respond to customers, and gave a 360-degree view of each
customer interaction. They also provided more succinct visualizations
of data, with better features to interpret it.

336
 The new era of business
intelligence solutions
With data now coming from everywhere and everything, organizations
must be prepared to convert that data into business knowledge so
users can make informed decisions and trigger actions. Many organi-
zations employ highly skilled workers, such as data scientists, who are
responsible for analyzing the data and producing the necessary in-
sights that can affect the business. This approach is expensive and adds
dependency on specialized roles to perform tasks that usually can’t be
completed by typical business users.

An approach to reduce this dependency while simultaneously increas-

ing ready access to data is to use augmented analytics, a feature of
modern business intelligence solutions. According to Gartner, Inc.,
a research and advisory firm, “Augmented analytics is the use of en-
abling technologies such as machine learning and AI to assist with data
preparation, insight generation, and insight explanation.”

Embedded intelligence solutions that make use of augmented analyt-

ics allow processes, services, or products to deliver their own insights
to increase quality, efficiency, and customer satisfaction. These are new
types of reporting and analytic experiences where the intelligence is
by-design, and the solution itself analyzes the data to provide insights.
While the data may not yet be unified with such an approach, cus-
tomers and organizations can use insights provided by products and
services from the earliest phases of usage or production, which increas-
es their return on investment (ROI).

Reporting
and analytics strategy
To be a market leader, an organization’s products and services must
evolve continuously and exceed customer expectations. The informa-
tion used to improve a product or service is based on data that comes
to the surface via reporting. Reporting requirements can be as simple

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 337
 as determining the status of an order or as complex as a cash-flow
analysis for a global organization.

What is a reporting solution?

An effective reporting solution is one of the essential elements for
business success. A properly implemented reporting solution provides
a competitive advantage, as users have access to the exact information
they need exactly when they need it, which accelerates and improves
the decision-making process.

Define your analytics strategy

An analytics strategy determines how an organization will provide the
ideal level of accurate information to users for timely decision-making.

The strategy includes a list of tools and types of reporting (such as

analytics, static reports, and ad-hoc reporting), as well as how the
information will be secured and shared with users. Implementation of
the strategy forms a foundation for defining standards, processes, and
best practices.

Your analytics strategy can help transform data collected from different
processes and systems into knowledge to help you stay competitive in
the market.

A successful strategy caters to user requirements and how those re-

quirements must be fulfilled. For example, some business users may
require printed reports, which can be sent to customers and vendors.
Others may need data to be available for summarized, detailed, or
ad-hoc reporting. Business management may require financial or
operational reports to understand the health of the organization and
determine an overall strategy. Serving such varied user needs often
requires different tools to manage data coming from multiple sources.
For example, financial and ad-hoc reporting requirements may rely
on data extracted and stored in a staging location utilizing Microsoft
Azure data services such as Azure Synapse Analytics, as well as manu-
facturing or sales information in Dynamics 365.

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 338
 With a centralized data and business intelligence strategy, the IT de-
partment is usually responsible for data, ETL processes, and reporting
solutions. In a self-service approach, the IT team implements a data
Every organization needs people who
understand how to analyze and use data.
analytics platform that enables users without statistical, analytical, or
Training and communication are key to data-handling expertise to access and use data. Teams should under-
ensuring that business users know how
to use data and insights to optimize their stand their organization’s goals and strategy before implementing an
work, drive efficiency, and make informed enterprise data platform.
decisions.

Data and analytics processes change

Any reporting solution requires ongoing maintenance and updates
periodically to improve how information
is provided. Make sure users are aware of as underlying system requirements and business priorities change.
any upcoming change and understand
how it will affect them. Constant updates
The project team should clearly define the delivery, maintenance, and
without appropriate change management operational responsibilities. There should also be a plan for making
can confuse users, but effective change
management can improve the adoption historical data from legacy systems available for reporting, business
rate of analytics.
intelligence, or audit purposes.

Gain deeper insights from data

For most organizations, business intelligence and analytics are critical
to their growth and digital transformation. With technology such as
Azure Machine Learning and Adaptive Insights, organizations can gain
deeper insights from their data. Having an early understanding of
these technologies also can help organizations design their systems for
future-state data consumption.

Public, cloud-based business intelligence solutions are a good fit for

many organizations, as they are on-demand, self-service, and scalable
solutions. To choose the best technology to meet requirements today
and in the future, the implementation team must understand their or-
ganization’s current state and the future vision for enterprise business
intelligence and analytics.

Understand reporting requirements

Understanding reporting requirements is key to delivering successful
reports and helping improve the entire business. Reporting require-
ments must be examined in the context of the business process for
the purpose of the report—such as operational reporting, financial or
regulatory reporting, ad-hoc inquiries, and dashboard and analytical

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 339
 reporting—and should not be an afterthought.

As part of the business requirements gathering process, organizations

must address reporting requirements, and collect information about
how reports will be used, such as:
▪ Printed on paper and sent with a shipment on a truck
▪ Used to create a pivot table in Microsoft Excel and build another
report
▪ Submitted to government agencies at the end of the year
▪ Provided to a bank on a periodic basis for audit purposes

Organizations must also identify any critical reports that require data
mash-up with other sources, such as front-office apps or transportation
systems, to develop an appropriate data integration and solution strat-
egy. Data volume may help determine how reports will be designed
and shared with users.

If reports will be validated by a third party, organizations need to

determine how reports will be generated and shared, and what data
must be included in the report before it can be approved.

Requirements gathering also includes security policies—encryption,

access controls, and row and column level—and requirements related
to data retention, data residency, and personally identifiable informa-
tion (PII).

Document printing and electronic reporting

Document printing requirements are another factor to consider. To
comply with local laws, businesses typically must submit regulatory
and compliance documents in a pre-defined printed or electronic
format provided by government agencies. The data required for these
documents often resides in enterprise resource planning (ERP) systems.

The Finance and Operations apps have an Electronic reporting (ER)

tool to help organizations adopt new regulatory requirements and
generate documents in the required format to electronically exchange
information with government bodies, banks, and other parties.
Targeted at business users instead of developers, the ER tool lets users

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 340
configure different document formats instead of writing code, making
the processes for creating and adjusting formats for electronic docu-
ments faster and easier. It works with the TEXT, XML, Microsoft Word
document, and OPENXML worksheet formats, and an extension inter-
face provides support for additional formats.

Customer engagement apps allow users to use Word templates

to create standardized documents. The Dynamics 365 Customer
Engagement apps include system reports for sales insights, while the
Dynamics 365 Sales app lets you individually print quotes, invoices, or
other records and create PDF files from sales records.

Business document management

Business document management is built on top of the ER framework,
and enables business users to edit document templates by using
Microsoft 365 or a Microsoft Office desktop application. Edits might
include changing business document designs and adding placeholders
for additional data without source code changes and new deploy-
ments. No knowledge of the ER framework is required to update
templates of business documents.

Financial reporting
Finance and business professionals use financial reporting to create,
maintain, deploy, and view financial statements. Financial reporting
capabilities in Dynamics 365 Finance and Operations apps support
currency translation and financial dimensions to efficiently design
financial reports.

Financial reporting in the Finance and Operations apps provides de-

fault financial reports that organizations can use as is or as a starting
point for their financial reporting needs. In addition to traditional
financial reports such as income statements and balance sheets, the
app’s default reports include examples of the many types of financial
reports you can create and customize.

Dashboards, charts, and analytical workspaces

The Dynamics 365 Sales app uses dashboards to provide an overview
of actionable business data across an organization, and give insights on

341
 sales data and team performance. Sales representatives and managers
can use an out-of-the-box sales pipeline chart in Dynamics 365 Sales to
visualize the revenue for an opportunity based on each pipeline phase.

Dynamics 365 delivers rich and interactive reports that are seamlessly
integrated into application workspaces. By using infographics and vi-
suals supported by Microsoft Power BI, analytical workspaces let users
explore the data by selecting or touching elements on the page. They
also can identify cause and effect, perform simple what-if operations,
and discover hidden trends—all without leaving the workspace. Power
BI workspaces complement operational views with analytical insights
based on near-real-time information. Users also can customize embed-
ded reports.

Reporting categories
Reporting needs for an organization can be classified into two catego-
ries: operational reporting and business reporting.

Operational reporting
Examples of operational reporting include orders received in a day,
delayed orders, and inventory adjustments in a warehouse. This kind
of reporting supports the detailed, day-to-day activities of the organi-
zation. It is typically limited to a short duration, uses real-time granular
data, and supports quick decision-making to improve efficiency. It also
helps organizations identify their issues and achievements, as well as
future strategies and actions that may affect the business. Operational
reports can empower businesses to determine their next steps for
improving organizational performance. Organizations can fulfill oper-
ational reporting requirements using elements such as native controls,
SSRS reports, dashboards, and business documents.

Business reporting
Business reporting refers to reports detailing operating expenses and
financial key performance indicators (KPIs) to business stakeholders
so they can understand the organization’s overall health and make
more informed decisions. This kind of reporting delivers a view of
current performance to enable the stakeholders to identify growth

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 342
 opportunities and areas for improvement, and track business perfor-
mance against the planned goals for the organization.

The data estate and data

modernization
A data estate refers to the infrastructure that helps organizations man-
age all their data, regardless of where or how it is stored. The massive
growth in data and the increasing need for immediate insights demand
a fast, agile, accessible, secure, and highly scalable data platform.

Organizations typically used legacy applications to both store data

and serve as the interface for data collection. Data modernization
restructures how data is collected, stored, and managed to take advan-
tage of innovative technology, and is backed by AI and data analytics
solutions that optimize the data-review process. Organizations seeking
improved business performance must design and manage their data
estate based on a modern platform that breaks the data silos and
unleashes the data’s full potential.

Build a data strategy

For business solutions to deliver the intended results, an organization’s
data strategy must articulate a clear vision that effectively aligns busi-
ness intelligence investments with business goals to maximize impact,
drive growth, and improve profitability.

A common risk for businesses is the quality To define your data strategy, start with the business questions you
of their data. If data is weak, unstructured,
unsecured, or difficult to use or access, need to answer to meet your organization’s goal of using the data to
additional work will be required to move,
make more effective decisions that improve the business outcome.
scrub, protect, and improve the quality of
the data. Poor data quality also can lead to
lost opportunities, failure to consistently
attract customers, increased time and
With customers now having access to more information and channels,
expense for cleaning data or processing an organization’s data strategy should reflect the customer journey.
corrections, and inaccurate or inconsistent
KPI measurements. From a data management perspective, all channels and checkpoints
across the customer journey should be represented.

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 343
 Modernize your data estate
Your data estate modernization strategy should map the current data
estate, goals, business processes, and regulatory requirements to aid
gap analysis of the existing and desired future state. It should identify
key analytics and metrics to improve the business and effectively align
business intelligence investments with business goals.

Some organizations emphasize collecting Cloud-based solutions offer modern capabilities based on machine
data more than analyzing it to drive
insights. It’s also important to identify gaps
learning and AI to analyze portions of data and automate other pro-
and blockers for business objectives and cesses, such as identifying and solving data quality issues. In terms of
outcomes, and not focus solely on the data,
structure, analytics, tools, or technologies. analysis, prebuilt models can serve the most common use cases, elimi-
nating the need to build custom models.

The modern data estate and the digital

Implementing or investing in sustaining
feedback loop
legacy solutions brings the risk of a future Gartner, Inc., predicts that investments in data and analytics strategies
gap. AI and machine learning solutions take
a modern approach to data and analytics will critically affect businesses in the years ahead.
that grows as technology evolves, making
it possible for organizations to add new
features and capabilities in the future—and How organizations modernize their data estate is influenced by the
reduce the risk of being outperformed by
competitors.
four data signals found in the digital feedback loop (Figure 13-1).

Empower people
To do their jobs more efficiently, employees need tools and resources,
as well as timely access to information. Utilizing AI to further automate
business processes contributes to better and faster results, which then
empowers your people (Figure 13-2) to make the best decisions and
deliver value to customers.

Engage customers
Modern applications are already capable of delivering insights by using
AI and data analytics to optimize business processes and shed light on
customer activity. For example, Dynamics 365 can provide a customer
service agent with insights into a customer’s interests and purchasing
behavior information in real time, allowing the agent to make sugges-
tions tailored to the customer. These types of insights help businesses
intelligently engage customers (Figure 13-2) to provide a superior

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 344
 Fig.
13-2
2022
70%/60% OPTIMIZE OPERATIONS
70 percent of organizations will rigorously track data
quality levels via metrics, increasing data quality by 60
percent, and significantly reducing operational risks
and costs.

60% EMPOWER PEOPLE

2023
60 percent of organizations will utilize
packaged AI to automate processes in
multiple functional areas.

ENGAGE CUSTOMERS 30%

30 percent of customer service organizations will
deliver proactive customer services by using AI-enabled
process orchestration and continuous intelligence.
2024
30% OPTIMIZE OPERATIONS

OPTIMIZE OPERATIONS 30% 30 percent of organizations will invest in data and

analytics governance platforms, thus increasing
30 percent of organizations will exceed their
the business impact of trusted insights and new
data and analytics return on investment (ROI) by
efficiencies.
governing the least amount of data that matters
most to their strategic goals.

OPTIMIZE OPERATIONS 90% 2025

50%
90 percent of the world’s top 500 companies will have
converged analytics governance into broader data EMPOWER PEOPLE
and analytics governance initiatives. 50 percent of all enterprise business-to-business
(B2B) sales technology implementations will use

5x
customer-engagement analytics to optimize their
OPTIMIZE OPERATIONS sales processes.
Cloud-based AI will have increased five times from
2019, making AI one of the top cloud services.
50% TRANSFORM PRODUCTS
(AND SERVICES)

30%
More than 50 percent of equipment manufacturers
OPTIMIZE OPERATIONS will offer outcome-based service contracts that rely
30 percent of organizations will harness the on IoT-based connectivity (up from less than 15
collective intelligence of their analytics communities, percent in 2019).
outperforming competitors who rely solely on
centralized or self-service analytics.
50% ENGAGE CUSTOMERS
Digital adoption solutions will be white-labeled in
Source: “Predicts 2020: Data and Analytics Strategies – Invest, Influence 50 percent of customer-facing software as a service
and Impact.” Gartner.com. Gartner, Inc., December 6, 2019. (SaaS) a
 pplications, increasing customer satisfaction
and loyalty.

Implementation Guide: Success by Design: Guide 10, Business intelligence, reporting, and analytics 345
 customer service experience.

Transform products (and services)

Products and services have become sources of data that provide better
insights into lifecycle-related information that can be used to benefit
organizations and their customers. By analyzing those insights, organi-
zations can in turn transform their products and services (Figure 13-2)
to take advantage of opportunities and expand into new channels and
markets. For example, a smart air conditioner that allows you to set
an optimal temperature via an app can also apply what it learns about
your household’s daily routine to make sure the house is cool when
everyone gets home. The air conditioner’s manufacturer then uses
data collected from all users to enhance the product’s smart features,
leading to a continuous improvement cycle.

Optimize operations
Cloud-based AI usage is an increasingly common investment area for
most organizations—and not just for customer-facing technology.
For example, Dynamics 365 for Field Service can use AI to anticipate
hardware failures on a manufacturing floor and automatically dispatch
technicians before the malfunction. Organizations that optimize their
operations (Figure 13-2) with augmented analytics, AI, and embedded
intelligence will be more competitive in the marketplace.

Components of the
modern data estate
The Dynamics 365 platform can be an essential part of your modern
data estate architecture. Your business data is securely stored within
Dynamics 365 as a function of your day-to-day operations. In addition,
Dynamics 365 can export data to or ingest data from various sources
to be used in reporting, workflow, applications, or in any other way
that is required by your business.

You can also generate insights and analytics from data created
and managed inside each Dynamics 365 application. Apps using

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 346
 embedded intelligence, such as the audience insights capability inside
the Dynamics 365 Customer Insights app, enrich your data and allow
more informed decision-making.

Common Data Model

Dynamics 365 data estate components can ingest, store, prepare, mod-
el, and visualize data to produce insights that will support, boost, and
even transform operations (Figure 13-3). The Common Data Model
(CDM) is one of the key technologies enabling access to many kinds of
information from heterogeneous services and data stores.
Fig.
13-3

Components of a modern data estate

App 1 App 2 App N

Common
Data
Model

Data source 1 Data source 2 Data source M

The Common Data Model (CDM) The CDM enables quick app deployment
unifies data in a well-known and development, out-of-the-box
schema with semantic consistency. intelligence, and more.

There’s a growing collection of solutions that store data in CDM form:

Power BI Dynamics 365 Power Apps Azure

The CDM—a shared data language used by business and analytical

applications—offers a set of standardized, extensible data schemas to
enable consistency of data and its meaning across applications and
business processes.

The CDM specification defines out-of-the-box standard entities repre-

senting common concepts such as people, budgets, and campaigns.
Having your data in a common format for entities with the relevant
metadata makes it possible to build apps and services that light up
based on the data. The CDM format also makes it easier to consume
data in the Microsoft Power Platform, which uses Power BI to get

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 347
 insights from your data. Normalizing the data sets up your organization
to better identify opportunities and puts your data in a format that could
be utilized for future projects. Dynamics 365 uses Microsoft Dataverse—
which is structured according to the CDM—to store and secure app
data. The CDM structure is defined in an extensible schema, as shown
in Figure 13-4. This allows organizations to build or extend apps by
Fig.
13-4 using Power Apps and Dataverse directly against their business data.

CDM Schema

Core
Customer relationship management (CRM)
Account
Account Sales Service Solutions
Activity
Appointment Competitor Case
Marketing
Contact
Campaign Discount Contact Account

Currency
Contact Invoice Entitlement Contact

Email
Lead Opportunity Resource Event

Goal
Marketing list Order Service Marketing email

Letter
Phone call Order product Schedule group Marketing page

Note
Social activity Quote Task

Owner

Organization

Position Web Talent Healthcare

Link clicks Department Patient Risk assessment Procedure
Task

Web interactions Job position Practitioner Encounter

Webpage Worker Device Episode of care

Care plan Medication

Working with
industry and subject
matter experts, the
CDM is expanding to Account Attributes
include additional Description: A business that represents a • accountNumber • creditLimit • territoryId
business enterprises customer or potential customer. • accountRatingCode • openDeals • hotelGroup
and concepts. • createdOn • openRevenue • ...

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 348
 Data unification components
Services and applications that ingest data from multiple sources serve
a vital role in a modern data estate. Aggregation from data stores and
services provides users with business-critical information supplied in
dashboards and reports. The resulting data and events can also be
used to trigger workflows and act as inputs to the apps running on the
Dataverse platform. Many data unification components are built into
Dynamics 365 applications—and organizations can design their own
integrations to fit their business needs.

Customer Insights as a customer data platform

Dynamics 365 Customer Insights is a real-time customer data platform
that brings together transactional, behavioral, and demographics data
from various sources to create a 360-degree view of your customers.

The Customer Insights solution (Figure 13-5) offers prepackaged

customer-analytics capabilities such as segmentation, churn analysis,
and product recommendations, with Power BI embedded for easy
data exploration and visualization. It also incorporates AI and machine
learning models, so data scientists can create custom models in the
Fig. Azure Machine Learning platform. The addition of Azure Cognitive
13-5

Customer Insights solution

Internal External Process optimization Channels of
and automation engagement
Ingest Map and match
Contacts Partner data
mographic
De Mobile
Point of sale
Events
(POS)
Power BI Web
Cases Loyalty Analytics
io n al

Campaigns Geo
Beh

Social
act
av

Web Wi-Fi Power Automate

ra
ns

Tra
io

l AI and
Bots
Email Social machine
Conflate Enrich learning
Mobile LinkedIn Power Apps In-person

Data Unification Insights Action

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 349
 Services provides text, speech, image, and video analysis, and enriches
data via Microsoft Graph.

Dataverse applications
Building an app typically involves accessing data from more than one
source. Although it can sometimes be done at the application level,
there are cases where integrating this data into a common store creates
an easier app-building experience—and a single set of logic to maintain
and operate over the data. Dataverse allows data to be integrated from
multiple sources into a single store, which can then be used in Power
Apps, Power Automate, Power BI, and Power Virtual Agents, along with
data that’s already available from Dynamics 365.

Data export components

A key characteristic of a data estate is the ability to export data to
services as needed. Because data in the Dataverse conforms to the
CDM standards, businesses have the flexibility to export data from
Dynamics 365 to other cloud services and applications using an
industry-standard

format, greatly simplifying the standardization of exported and im-

ported data. To facilitate this business need, Dynamics 365 has several
built-in export services.

Export to Azure Data Lake

The Export to Azure Data Lake for Finance and Operations apps and
Azure Synapse Link for Dataverse services are the preferred option
to continuously export data from Dynamics 365 to Azure Data Lake
Storage Gen2. The service supports initial and incremental writes for
table data and metadata. Any data or metadata changes are automati-
cally pushed to the Azure Data Lake without any additional action.

Azure Data Lake Storage Gen2 is the foundation for building enterprise
data lakes on Azure and provides low-cost, tiered storage with high
availability and disaster recovery capabilities.

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 350
 Fig.
Azure Synapse Analytics is a limitless analytics service that brings
13-6
together data integration, enterprise data warehousing, and big data
analytics.
Microsoft Power
Platform Data Export Service
The Data Export Service replicates data from the Dataverse database
The low-code platform that spans
Office 365, Azure, Dynamics 365, to an external Azure SQL Database or an SQL Server on Azure virtual
and standalone applications machines. This service intelligently syncs all data initially, and thereaf-
ter syncs the data on a continuous basis as delta changes occur in the
system, enabling several analytics and reporting scenarios on top of
Azure data and analytics services.

Bring your own database (BYOD)

Power BI The bring your own database (BYOD) feature lets organizations export
Business analytics
data from the Finance and Operations apps into an external Azure
SQL database. BYOD helps support scenarios where organizations may
need to export data for unification or other purposes.

Power Apps Embedded intelligence

Application development
Dynamics 365 with embedded intelligence, such as Sales Insights and
Customer Service Insights, allow organizations to use AI without de-
pending on highly skilled resources. These apps continuously analyze
your data and generate insights to help you understand business
Power Automate relationships, evaluate activities and interactions with customers, and
Process automation
determine actions based on those insights.

Unified data results in a 360-degree view of customers to ensure

high-quality decision-making and predictions informed by the best,
most recent data.
Power Virtual Agent
Intelligent virtual agents

Power Platform
The Power Platform (Figure 13-6) enables organizations to analyze, act
on, and automate the data to digitally transform their businesses. The
Power Platform today comprises four products: Power BI, Power Apps,
Dataverse
Power Automate, and Power Virtual Agents.

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 351
 Power BI
Power BI is both part of the Power Platform and stands on its own
by bridging the gap between data and decision-making. Power BI
lets business analysts, IT professionals, and data scientists collaborate
seamlessly, providing a single version of data truth that delivers in-
sights across an organization.

Power BI helps you analyze your entire data estate within the Dynamics
365 or Azure platforms, or external sources. Power BI can connect
individually with siloed sources to provide reporting and analytics, or
it can connect with data stores within or outside of Dynamics 365. As
data can come from multiple sources, organizations should analyze
how Power BI will connect with those sources as a part of their data
estate pre-planning.

Making data available to all users is risky. Security in reporting and

analytics should be aligned with your business solution. Be sure to have
your data governance in place, and periodically review who needs
access to which information. This may be a time-consuming process,
but your data must be protected, secured, and only accessible to those
who truly need it.

Power Apps, Power Automate, and Power Virtual Agents

Built to work seamlessly with data in the CDM format, Power Apps
allows you to consume, create, and complement information to
contribute to scale and efficiency for the users consuming that data.
Organizations can also get accurate insights by adding low-code AI
tools to their process automation via Power Automate. Power Virtual
Agents help you create and manage powerful chatbots—without the
need for code or AI expertise—and monitor and improve chatbot
performance using AI and data-driven insights.

Microsoft Azure
With Dynamics 365 at the center of the data estate, Azure provides an
ideal platform (Figure 13-7) for hosting services for business work-
loads, services, and applications that can easily interact with Dynamics
365. Built-in services in Dynamics 365 let you export data as needed

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 352
 or scheduled. Power BI can aggregate information from Dynamics 365
and Azure sources into an integrated view, and Power Apps can access
both Dynamics and Azure sources for low-code, custom applications
Fig.
designed for business use.
13-7

Azure features and services with modern data warehouse architecture

Ingest Model and serve Visualize
On-premises
data source Azure Synapse
Analytics
Other data
source
Prep and train
Azure Data Factory Power BI
Logs and files Code-free data transformation Azure Databricks Leader in the Gartner Magic Quadrant for
(unstructured) and integration via built-in connectors Analytics and Business Intelligence Platforms

Store

Dynamics 365 Azure Data Lake Storage

High-performance, low-cost data lake
available in Azure regions

Azure cloud platform

For data estates that include Dynamics 365 and other platforms, Azure
can help with unification of data to generate insights. There are mul-
tiple options within the Azure cloud platform to ingest, store, prepare,
model, and visualize data (Figure 13-8), as well as build intelligence on
top of that data—whether it’s siloed or unified.

Azure Stack
Azure Stack is a portfolio of products that allows you to use embedded
intelligence to extend Azure services and capabilities to your environ-
ment of choice—from the datacenter to edge locations and remote
offices. You can also use it to build and deploy hybrid and edge comput-
ing applications, and run them consistently across location boundaries.

Data load, ingestion, and pipeline orchestration

Azure Data Factory is a hybrid data integration service for creating,
scheduling, and orchestrating ETL and extract, load, and transform
(ELT) workflows. Organizations can use Azure Data Factory to combine
CDM data and other data sources, and process that data for unique
insights. Azure Event Hubs can ingest data streams generated by a
client application.

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 353
 Fig.
13-8

Modern data platform reference architecture

Data Process Business user

Stream datasets and

real-time dashboard

Hotpath
Real-time analytics

Stream Load and ingest Azure Stream Analytics

V=Velocity
IoT devices, sensors,
gadgets Lambda architecture
(loosely typed)

Coldpath
History and
Nonstructured Azure Event Hubs trend analytics Azure Cognitive Services Power BI Premium Analytics
V=Variety Azure Machine Learning
images, video, audio, free
text
(no structure)

Semi-structured Azure Data Factory Azure Data Lake Storage Gen2 Azure Databricks Azure Cosmos DB Application
V=Volume
Scheduled/event-triggered Fast load data
images, video, audio, with PolyBase/ Integrated big data
data integration scenarios with traditional
free text ParquetDirect
data warehouse
(loosely typed)

Enterprise-grade
semantic model

Azure Synapses Analytics Power BI Premium Analytics

Relational database
(strongly typed, structured)

Store Server

Data store
Azure Blob Storage offers massively scalable object storage for any
type of unstructured data—such as images, videos, audio, and doc-
uments—while Azure Data Lake Storage eliminates data silos with a
single and secured storage platform.

Machine learning and AI

Azure Databricks provides a cloud-hosted Apache Spark cluster where
data engineers, data scientists, and business users can collaborate to
train their models and get analytics.

The Azure Machine Learning service gives developers and data sci-
entists a range of productive experiences to build, train, and deploy
machine learning models faster.

Azure Cognitive Services puts AI within reach of every developer—

without requiring machine-learning expertise. It only takes an API call

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 354
 to embed the ability to see, hear, speak, search, understand, and accel-
erate decision-making in your apps.

Model and serve

Azure Analysis Services employs enterprise-grade analytics as a ser-
vice (AaaS) to govern, deploy, test, and deliver business intelligence
solutions.

Azure Synapse Analytics is a fast, flexible, and trusted cloud data

warehouse that lets you scale, compute, and store data elastically and
independently with a massively parallel processing architecture. It can
scale across proprietary SQL and open-source databases and manage
analytics workloads to provide fast, cost-effective power over any
data—whether it is structured, semi-structured, run in real time, or
globally distributed.

Azure Cosmos DB is a fully managed NoSQL database service for

modern app development.

Synergy
Getting maximum value from your data requires a modern data
Due to budget, time, or skills constraints, estate based on a data strategy that includes infrastructure, processes,
some organizations decide to deliver
a business solution as a first step,
and people.
with a plan to improve insights and
analytics capabilities at some point in
the future. Insights and analytics should Your data can flow inside a cloud solution or via synergy with other
be addressed in the early phases of a components and platforms to provide an infrastructure and processes
business solution, even if it doesn’t
yet include all scenarios. Leaving these for analyzing data and producing actionable outcomes.
vital elements until later can affect
the business and the user experience,
eventually reducing adoption, increasing People are a big part of the process, and the data journey will often
costs, and giving a negative perception of
the value the solution has to offer.
start and finish with them. The insights and actionable outcomes will
allow them to make better decisions—for themselves and the business.

Conclusion
In this chapter, we discussed how organizations are becoming more
competitive, expanding their global reach to attract customers, and

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 355
 using business intelligence solutions to make the most of their data.

While seeking data on how customers interact with their products and
services, organizations are acting on that data to give customers access
to more content, new purchasing channels, and brand options. By
deeply understanding customer behavior, organizations can engage
customers proactively and make predictions about their future behavior.

Successful organizations will use data to empower their employees,

intelligently engage their customers, transform their products and
services, and optimize their operations. They’ll also use data to reduce
operational risks and costs, and to respond to opportunities and mar-
ket changes to meet their customers’ needs.

Business intelligence, reporting, and analytics should not be an af-

terthought. Focusing on building a unified and modern data estate
gives organizations access to augmented analytics and embedded
intelligence solutions, which will be the differentiator in the future.
By changing the ways that they use data to understand and engage
with their customers, organizations can reach each customer at the
right time, on the right channel, with the right message, for the right
outcome.

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 356
 Checklist

Reporting and Align the reporting and analytics to the broader master
data management strategy.
analytics strategy
Use customer data platform offerings such as customer
Align reporting and analytics strategy with overall
insights to unify the customer data from various siloed
implementation project plan.
data sources.
Map out the organizational data estate to develop a
Use modern data and BI platforms such as Azure Synapse
holistic view of different data sources, the type of data
Analytics and Power BI to build enterprise data solutions.
they hold, and the schema used.
Focus on not just delivering a report but aligning
Define your analytics strategy and the tools to support
reporting with business processes to add value to the
it. Ensure the approach meets the current and future
organization.
reporting requirements while considering the data
volumes and different sources of data.

Create an intelligence strategy that considers the report-

ing needs at various levels, such as strategic business
reporting, operational and process reporting, end-user
intelligence, and administrative reporting.

Take advantage of the out-of-the-box reporting and

embedded intelligence capabilities in the app.

Align the organization insights and analytics roadmap to

the business capabilities being delivered.

Align to the Common Data Model to take advantage

of the standardized business applications schema for
better inoperability across systems.

Understand the security implications when exporting

data from the app to external analytical data stores.

Implementation Guide: Success by Design: Guide 13, Business intelligence, reporting, and analytics 357
 Case study
Premier yacht brokerage cruises
into smarter marketing and
boosts sales by 70 percent with
Dynamics 365
To differentiate itself from the competition, a large company that sup-
plies luxury vessels decided to invest in reporting and intelligence. In
an industry where relationship-building efforts must be incredibly pre-
cise and personal, any insight into a customer’s mindset is invaluable.

The company implemented a solution based on sales, marketing, and

finance data, using Dynamics 365 and Power BI to help identify trigger
points that would help increase sales to charter clients, strengthen cus-
tomer relationships with improved marketing activities, and seamlessly
move customers through the sales cycle.

This solution allowed the company to build intelligence on top of a

centralized system running in the cloud, and facilitated a move from
legacy systems where data was siloed in different applications, chan-
nels, and departments.

Dynamics 365 made it possible to separate the features each de-

partment would use while accessing the same data. The marketing
and sales teams started customizing their content and campaigns to

Implementation Guide: Success by Design: Guide 10, Business intelligence, reporting, and analytics 358
 nurture leads toward a sale, and used Power BI reports to generate
insights that helped them identify the best prospects and create win-
ning proposals.

Integration with external applications brought in data from other

sources to improve the company’s bespoke offerings. For example, the
sales teams could collate even more information about their products
and send it to customers in just a few minutes—instead of spending
hours manually putting together the data.

After deploying the Dynamics 365 solution, the company experienced

a 70 percent increase in sales. They next brought in more departments
to create a central hub for all activities, simplifying every area of their
operation. For example, the sales team’s process integrated everything
that the accounting department needed to efficiently close a deal and
reduce the possibility of human error.

The company is also using Dynamics 365 and Power BI to uncover mar-
ket trends, as augmented analytics help the company learn about their
customers in ways that were not possible before—and build authentic
relationships that will last long after the initial sale.

Implementation Guide: Success by Design: Guide 10, Business intelligence, reporting, and analytics 359
14
Implementation Guide: Success by Design: Guide 14, Testing strategy
Guide
Testing
strategy
360
 Overview
During the implementation of a solution, one of the
fundamental objectives is to verify that the solution
meets the business needs and ensures that the
customer can operate their business successfully.
The solution needs to be tested by performing multiple business execu-
tion rehearsals before the operation begins and the system is taken live.

It is imperative that proper testing of the solution is completed before

deployment to production. This helps avoid costly challenges that
could delay the adoption, create negative perception of the final
product quality, and reduce the confidence to run the business on the
new solution.
Defining a comprehensive test strategy
for your implementation is a combination
of considering the project scope, testing
The goal of testing is to detect defects in the process in order to ready
planning with a test plan and test cycle
schedule in combination with the phases the solution, processes, and people to operate the business in produc-
and solution versions availability, and
selecting all the test types and how to
tion application efficiently and accurately as soon as possible. Testing
execute them. allows us to detect defects as early as possible in this process, in a
• Define scope based on your systematic way and at the recommended time per the test type.
processes and detailed test cases
• Create a test plan with the necessary
test cycles This chapter explores the components of:
• Select the adequate test types for
▪ The value of testing
your project
▪ Types of testing
We explore each of these areas in more
detail in this chapter.
▪ Testing at the right time with clear scope
▪ Defining the strategy

Implementation Guide: Success by Design: Guide 14, Testing strategy 361

 You can also find recommended practices for executing the testing,
how to deal with the outcomes, and finally, reaching a business sign off
that confirms the readiness of the solution.

We look forward to identifying the key components of your testing

strategy in this chapter.

Focus on quality and

keeping scope
Testing is a continuous task under application lifecycle management.
Not only is it critical during the implementation of the solution, but
also during the Operation phase. It is a continuous evolution that
keeps bringing fixes or extending the solution. In the beginning testing
is completed manually, but over time with automation we can make
the process far more efficient. Our objective is to ensure the quality of
the solution is always meeting customer expectations.

To align with the scope of the solution, testing needs to focus on vali-
dating the business processes in it to stay on track. Microsoft Dynamics
365 Business Applications are rich with features that can distract testers
easily during testing, so it is important to focus on the features that
add value to the business process. By only focusing on those features
and the requirements needed for them, we can get to the starting line
of initiating operations for the solution being built.

The approach for testing

Depending on the methodology you apply, agile, waterfall, or a hybrid
of the two, the testing frequency differs for each iteration. Regardless
which approach you follow, you need to look for a high-quality testing
approach to confirm readiness.

Implementation Guide: Success by Design: Guide 14, Testing strategy 362

 On top of your implementation methodology, you need to have a
clear strategy for testing, as illustrated in Figure 14-1. This strategy is
composed of:
▪ A clear testing scope defined by your processes and requirements
to validate.
▪ A comprehensive plan that brings clarity to when, why, who, how
much, and where to test.
▪ The different types of testing with descriptions of how the testing
is executed based on the complexity of the solution and minimum
quality gates allowing us to detect defects early in the process.
▪ A comprehensive tracking mechanism that allows you to monitor
overall progress, pass/fail rates, ownership, and issues tied to
specific test cases.
▪ A clearly defined process for adding or modifying test cases during
each individual test pass.

Fig.
14-1 Test strategy

Scope Plan Test type

What you When, who, where, How you

will test and how much you will test
will test

Explore each component to define the testing strategy for your

Dynamics 365 implementation.

Defining the
testing strategy
After observing thousands of Dynamics 365 deployments, we can see
that most customers do a reasonable level of testing before go live. But

Implementation Guide: Success by Design: Guide 14, Testing strategy 363

 the difference between each of those implementations is how thor-
ough the implementation team is at defining their strategy, and the
quality and depth of the testing.

In general, a good strategy does not change in approach between

implementations. What changes is the tailored artifacts used for each
implementation, for example, the process to define the scope can
be the same across projects, but the scope itself differs. Investing in
a good strategy and making sure the implementation teams use it
consistently increases the quality of the project outcome.

Fig.
14-2 Scope of the testing
rage of the tes
Cove t
The scope of the testing is defined early in the project after defining
the solution, and it is refined as the solution blueprint takes shape.
What changes throughout the project is the type of testing that comes
into focus as you move through the implementation lifecycle and the
Scope of
Processes ability to actually test the solution.
testing
Test cases
During the implementation, the number of test cases in the testing
scope keeps increasing as you progress in building the solution. After
go live, the scope is focused on maintaining the quality as you update
the solution, but testing can increase if you continue to expand.
Dee
pness of the test

Consider the testing journey as an incremental process, as illustrated

in Figure 14-2. Regardless of the timeline, you need to keep two
things in mind:
Make sure that every implementation ▪ A clear view of the end-to-end business processes.
you execute has a clear and solid
testing strategy.
▪ A clear definition of the depth and detail of the test case design,
but also the variations of testing, for example edge cases and
negative testing.

Testing connected to processes

To progressively build your solution you need to determine the imple-
mentation processes and the requirements connected to each of them.
As we build the solution and make it ready for validation, the scope
of the testing increases as we add more processes. During the project
progression, we go from a simple functional test to an end-to-end
testing scope during user acceptance testing (UAT). Always connect the

Implementation Guide: Success by Design: Guide 14, Testing strategy 364

 test to the process being validated since this is how the end user con-
firms if they can perform a connected action to execute the business
process in place.

This connection is primarily important for testing connected functional

requirements, there is testing that checks other aspects of the solution
out of the process, for example testing quality of data, testing latency
in the connectivity, etc.

Connect the test to the process being

Testing is not an artifact of the development process; it is an artifact
validated. This facilitates tracking, of the business process created when you implement business appli-
reporting, and final sign off on
readiness by the business team. cations. Plan your testing by connecting those tests to the processes
and include processes that are executed outside of the new application
being implemented.

Test case definition

An important part of the scope is defining the test. Proper test doc-
umentation is key, and test cases (or test scripts) are the artifact that
achieves it. The nature of a test case varies depending on the test type.
Clear and detailed test cases guide the tester to validate the expected
outcomes without distractions and align to the scope of the solution.

Writing a good test case requires a series of tools and tasks that help
you validate the outcome. Dynamics 365 provides some of the tools to
build test cases faster, for example, the task recorder in the Finance and
Operations apps where you can dynamically capture the steps needed
to execute a task in the app.

Test cases should reflect the actual business execution in the system
and represent how the end user ultimately operates the system, as
illustrated in Figure 14-3.
Azure DevOps is great tool for documenting
test cases and connecting them to the
solution development lifecycle. Having this
artifact in Azure DevOps allows for tracking
Test cases should be composed of, at minimum, the following:
any bug discovered during testing and ▪ The process and requirements that the test case is covering.
helps to trigger the fix and plan for the next
testing cycle while also determining the ▪ The prerequisite, or entry, criteria to execute the test, which can
coverage of your test; for example, process, be dependent on other tests or necessary data to produce the
user story, requirement, design, etc. This is
depicted in Figure 14-4. expected outcome.
▪ The reproduction steps.

Implementation Guide: Success by Design: Guide 14, Testing strategy 365

 ▪ The expected outcome, or exit, criteria that helps to confirm the
readiness.

Finally, when you design your test cases, design for what you expect to
happen based on the process objective but also on what it should not
do. Plan for tests that can break the process.

Fig.
14-3 Components of the test case

+ + +
What Get ready How Quality check
Linked business Prerequisites Reproduction steps Expected outcome
process requirements

Fig.
14-4

Process Requirement Prerequisites

Prospect to cash Create sales order based on servicing Customer with no credit line available
system part request Stock available

Description
Dealership request parts for recreational vehicles out of warranty

Test case ID Test steps Test data Expected results Actual results Pass/Fail
PC-20 1. Create sales order header Customer: Sales order cannot Sales order is Fail
2. Check credit line in order CU050 - Great Adventures be posted posted
3. Create order line Dealership
4. Validate and post sales order
Part:
P001 - All weather speakers

Qty: - 4 pieces
Tester notes P001 - All weather speakers

Credit check is not stopping the order to be posted. Customer setup with wrong credit limit. Data quality issue.

Implementation Guide: Success by Design: Guide 14, Testing strategy 366

Planning
We have defined what is needed to determine the scope of the testing
using the processes as a foundation. With this valuable information, we
can determine how to execute the testing. This is where the planning
becomes important.

Planning for testing is a fundamental part of the testing strategy. The

next section describes the minimal components needed to define a
recommended testing plan. This strategy can be used to implement
any of our business applications.

During the planning stage of the testing, you need to be able to an-
swer and provide clarity for the following questions. This helps design
and define the frequency of the testing based on the necessary itera-
tions or test cycles to secure a high-quality solution.
▪ When do you need to start testing?
▪ How do you control the versions of the solution to be tested based
on the progressive readiness of it?
▪ How do you document the outcome of the testing?
▪ Who participates in the testing?
▪ Where does the testing happen in terms of environments and type?
▪ How is the ownership of the outcome defined?
▪ How deep you are going to test?
▪ What types of testing do you execute based on the project needs
and solution complexity?

This helps to plan the quality control portions of the solution and must
happen at the start of the Initiate phase of the project.

The plan for testing must be documented and signed off on by the
business team prior to its execution. This is important because it leads
into other types of planning, like environment planning, that deter-
mines where the test is done.

Creating a test plan

The test plan is the blueprint for executing your tests. It establishes
the guard rails to stay on track in terms of quality assurance for the

367
 solution, schedule of testing, and resources needed, and it contains the
scope of the testing.

Depending on the complexity of the project, this artifact can be de-

veloped by the same implementation team. In larger, more complex
organizations, this is prepared in conjunction with the quality assur-
ance team to secure a neutrality at tracking and reporting the outcome
of the test, as illustrated in Figure 14-5.

Fig.
14-5 Components of the test plan

Test cycles Roles Resources

How many Different personas Environments
cycles involved , devices,
printers,
etc.

Schedule Version Ownership Documentation

When testing Solution version Who will own Test asset
happens or scope per the outcome library
cycle

Always create a test plan for your project regardless of if your imple-
mentation is simple, complex, or if you are implementing a standard
solution or extending the solution.

Always create a test plan at The test plan brings transparency and helps keep your goals and
implementation and obtain sign off
on it by business stakeholders. objectives on securing a quality solution. It contains important infor-
mation that guides the team on how to conduct the testing.

When does testing happen?

It’s simple: the testing should start as early as possible. It is never too
early to start testing. Your testing cycles are incremental in terms of the
scope. Your first test cycles could be to test the standard functionality
to validate how far are you from the overall desired outcome.

Implementation Guide: Success by Design: Guide 14, Testing strategy 368

 Do not wait too long, or until the end of the implementation when you
think you have built the entire solution, to start testing. Testing late
adds serious risk to your implementation by having issues and gaps
that are hard to fix in a short period of time. This becomes a constraint,
especially when you have reduced time available for time sensitive
projects tasks like preparing to go live. Always plan to avoid testing
close to your go live date since it leaves no time to resolve issues. Poor
planning or no testing can ruin a project.

Testing cycles or iterations

Testing cycles are the frequency of the testing aligned to the availabil-
ity of the solution in scope for the testing at a given period time in the
implementation. Test cycles act as comprehensive testing milestones.

The successful outcome of the test cycle confirms the readiness of the
solution to meet the requirements of the business. There are different
terms used to describe this testing event, for example conference room
pilots, testing iterations, testing cycle, etc. The important message here
is that the testing event is comprehensive.

To implement Dynamics 365, it is recommended that the scope in each

testing cycle is based on business processes. Test cycles by application
module, or organizational department in isolation, risk losing the big
picture and thereby limit the effectiveness of the test.

Every testing cycle represents a key milestone in building the solution.

Consider every test cycle as a mini go live, where you are rehearsing
the business operation at every test.

Solution versions and alignment to them

We discussed the importance of planning the testing cycles process as
part of the scope, this scope also triggers the readiness of the code and
data that supports the test cases connected to those processes. After
determining the planned scope of the test cycle, you should plan your
development efforts to cover those processes and the readiness of the
required data in terms of master data and configurations.

Your solution version needs to be controlled at every test cycle so you

Implementation Guide: Success by Design: Guide 14, Testing strategy 369

 can have a clearly defined milestone. For every new milestone, the
solution keeps expanding in scope and complexity.

The solution version is a combination of the version of your:

▪ Code
▪ Configuration
▪ Master data
▪ Migrated data

Different teams need to coordinate and plan how to combine their

efforts to prepare for the testing cycle.

Align your solution version to the Each component of the solution version is very important, and all of
testing cycles and coordinate the
them need to come together as part of the solution. This is especially
readiness of all the dependent parties to
provide their own artifacts. Missing one important with data. The earliest you bring migrated data into the mix
necessary artifact can cause delays in
the implementation, introducing risk.
the better. One of the most common causes of errors during the first
days of operation is poorly migrated data.

Ownership definition
Having clear expectations of who takes care of the testing outcome
is key. Testing can result in bugs, but also it can result in discovering
configuration issues, gaps, or new opportunities where improvements
can be applied in future cycles or solution releases. Based on the type
of outcome, we need to define who takes ownership of the fix and
what test plan is needed to account for that.

▪ Bugs go back to developers for resolution.

▪ Configuration issues go to the consultant connected to the test case.
▪ Gaps go to the project manager for further discussion with the
stakeholders and implementation team.
▪ Standard product issues go to Microsoft Support or any other
third-party solution provider connected to the issue.
▪ Conceptual design issues go to the architects on the implementa-
tion team.
▪ Process improvements go to the business Subject Matter Experts
(SMEs).

Documenting the outcome of the test cycle and assigning clear

Implementation Guide: Success by Design: Guide 14, Testing strategy 370

 ownership for the fix facilitates tracking and resolution. Plan for testing
thoroughly, even if you are implementing standard product.

One common pattern for implementations

The right environment for the right type of test
of standard functionality is the
project team challenging the need to Where to test is dependent on the type of test being executed and this
thoroughly test when the application
has not been modified. The reality is
definition impacts the environment planning discussed in Chapter 9,
that implementation teams are testing “Environment strategy.” Refer to this chapter for the different types of
the specific business solution being
implemented, not the standard software. environments available and their overall purpose. Our focus is on the
You are not just testing software, you environments where you can do proper testing based on the test type.
are testing the people being trained, the
processes that represent the operation,
the data that keeps the business alive, and
This environment can be a different variation of a development or test
finally the technology.
environment, but it is important that you never complete or plan for
regular testing in production environments.

Production environments are meant to run the real business opera-

tions. Doing tests, or making changes to it without being tested first,
is a high risk for the stability of the solution because of the unknown
variables it can bring. The only test type that can be executed in a
production environment is a mock cutover test, and this test happens
for a very limited period before the solution is live.

You can require testing be done in an environment different than

Plan for non-Dynamics 365 testing

Dynamics 365 environments, like in the case of integrations. Plan for the
environments where you have a availability of the testing version of third-party systems that will integrate
dependency in the solution that
requires validation. with Dynamics 365; integrations or other dependencies will need to be
validated. This is a commonly missed opportunity during implementation.

Plan for additional hardware required to emulate the real operation

like scanners, printers, handhelds, etc. These resource needs are dictated
Planning to document the results of
by the test case.
your testing cycles helps to highlight
the wins, but also the bugs and
patterns. This report determines Documentation
the next course of action and can
Documenting testing has two angles, the test cases and tracking the
impact future milestones. It allows
stakeholders to make decisions to outcome of the test.
correct the paths if necessary.

Use Azure DevOps to build Having a model to document the outcome of a test case and test cycle
dashboards that can help to report
progress and quality of the testing. allows the implementation team to validate the performance of the
test cycle. Preparing to document and report the outcome highlights

Implementation Guide: Success by Design: Guide 14, Testing strategy 371

other challenges as well, though not necessarily ones connected to the
test cases itself. These issues can be solution performance, connectivity
issues, usability, continuity of the transaction, gaps in the product, etc.

Other important benefits of documenting the outcome are that it

tracks the progress of the testing and keeps the big picture in sight. It
allows us to detect and build the backlog for new non-high-priority
opportunities that are discovered during the testing. It can also trigger
a change of the statement of work when the newly discovered require-
ments are critical. This is highlighted as project risk.

The different roles involved in testing

The people involved in testing depends on the type of testing being
done. For example, performance testing types require more technical
roles and consultants familiar with that type of testing, while user
acceptance testing types require your end users.

Testers can be developers, functional consultants, customer subject

matter experts, quality assurance testers, or end users.

It is important to plan for the resources required for building your

solution and for the different types of tests that require special skills.
Consider the volume of test cases involved and the business areas
impacted so you have good representation of the business.

Roles for testing are determined by who to involve in tests depending

on the following variables:
▪ Test preparation (system, data, test cases, training, etc.)
▪ Test execution
▪ Test administrator or manager
▪ Test results triage
▪ Test reporting
▪ Test defect fixes

The business team involvement during testing is critical, they own the
solution and plan for enough time to allow for proper testing by the
key business users. One common pattern of failure is poor involvement
from this group of testers.

372
 Test types
In the previous section, we covered the different roles and environment
types that may be needed for testing, but we mentioned the need to
consider the test type. Let’s now combine the concepts in this section
so you can see how to use different test types to implement Dynamics
365 under the Success by Design framework.

In Figure 14-6, we show when we recommend executing the most

common test types when implementing Dynamics 365 apps across the
different Success by Design implementation phases. Note that some
test types are more relevant to specific phase than others.

Fig.
14-6 Testing throughout the solution lifecycle

Initiate Implement Prepare Operate

Get ready to start Design Build Exchange-to- User acceptance Deploy Business as usual and
exchange testing continuous update

Unit test

Functional test
* Go live

Testing with customer data

Process test

End-to-end test

Performance test

Security test

Regression test

Interface test

Data validation test

User acceptance test

Mock go-live

The minimal recommended test types are:

▪ Unit tests ▪ Performance tests
▪ Functional tests ▪ User acceptance tests
▪ Process tests ▪ Regression tests
▪ End-to-end tests ▪ Mock cutover

Implementation Guide: Success by Design: Guide 14, Testing strategy 373

 Each test represents an incremental testing approach. Some are
required for any implementation and others depend on the specific
implementation circumstances based on the complexity of the solu-
Test types need to be planned with
tion, performance, and security requirements. Some test types are
consideration to the complexity of the
solution, performance, and security especially relevant for specific stages in the implementation and others
requirements, and should be executed
during specific times within the
can run for the life of the solution. For all cases you need to test with
implementation. real customer data and migrated data as soon as possible.

Unit test
This test type focuses on individual function, code, and configuration
testing. It is done by developers and is the lowest level component of
the testing. In this test type, the developer verifies the requirements,
validates, improves the code design, and finds and fixes defects.

See it in action
This testing happens in a development environment and is expected to
Let us assume we are building a solution be done during the implementation mainly, or during any bug fixing.
for a customer that ships goods and
uses a specialized transportation system
It is a fast, iterative process. This test type is required if you are custom-
that is part of the final solution. For izing or extending your solution and is one of the first opportunities to
this example, we use unit testing to
test specific validations during a sales introduce a quality check in the solution. At this point you should val-
order confirmation and interface with
idate performance behavior of the individual components in addition
the transportation system using Power
Automate. The unit tests are written and to keeping security in scope. Being diligent from the start with this test
executed for each validation requirement
and interface in this scenario.
type can save you important time during implementation by avoiding
rework or bug fixes for issues that should be detected during this test.

Functional test
Functional tests can be either a manual test or an automated one. They
are done by the functional consultants, customer SMEs, or testers. The
purpose of functional tests is to verify the configurations of the solu-
tion or any custom code being released by the developers. The primary
objective is to validate the design as per the requirements. This is done
See it in action
in a test or developer environment during the Implement phase of the
Following the previous example Success by Design framework. At this point, testing automation can
regarding the sales order validation, the
functional testing of the interface with
also be introduced.
the transportation system focuses on the
configuration and setup, like customer
data, products, pricing, warehouse setup, This test type is the first test to be done by consultants and customer
and the dependencies with payment
SMEs. Consultants still need to do it, as it is important that the first line
processing and other module-related
configuration and parameters. of testing is done by the consultant’s previous customer testing at least
to verify stability of the feature.

Implementation Guide: Success by Design: Guide 14, Testing strategy 374

 At this point the consultants need to map the test to the requirements
and processes and the test case is detailed. The link with process is
agreed upon with the business so the test case is not focused only on
the gap, but also to fit in the whole picture of the solution.

The data used to test beyond this point requires it to keep evolving
continuously using customer data and should reflect the reality of
the operation.

Process tests
The solution continues to be built, with unit testing being done by
developers and functional testing by consultant and customer SMEs.
The work from the development team is validated, and bugs and
corrections are completed while the test is mapped to the process.

This is the point at which running connected test cases is ideal. The
point where we raise the bar in our testing approach by connecting a
collection of test cases that are all connected to our process.

Process tests can be manual or automated at this point. The business

has visibility and is actively participating in this testing. It is our first
opportunity to go beyond testing that focuses on just functionality or
modules, and it gives different people the opportunity to start playing
an important role in the test by handing over the outcome from one
test case to another. This should be your first comprehensive testing
See it in action cycle, starting during the Implement phase of Success by Design.

At this point we know how our sales

The objective is to verify the solution allows us to operate business
order function works, but now we want
to test how the sales order works in scenarios, and the testing needs to be done in a test environment.
connection with test cases preceding
and following the process. Let us call
this process prospect to cash. We want Tracking the outcome becomes more critical at this point, so bugs fixes
to test the collection of money after the
shipment of goods happen. We involve and other dependent improvements must be tracked. During this time,
other teams in the implementation team tracking tools like Azure DevOps become crucial as part of your appli-
that handles that part of the process,
and we validate that we can generate the cation lifecycle management.
invoice and collect money while closing
the cycle. From the business point of view,
the customer can confirm they can sell, For this test type, role security becomes crucial. We want to make sure we
ship, invoice, and collect the money so the
process prospect to cash works.
have the right segregation of duties and that the different personas can
operate as they should when they are running the solution in production.

Implementation Guide: Success by Design: Guide 14, Testing strategy 375

 This testing should look for unexpected outcomes of the process,
commonly known as negative testing, and not just the happy path.

Performance of the solution starts to become more evident with this

test type since more testers are involved, but also because it touches
more areas of the solution so the solution can start to enter into stress
mode when executing parallel functions. Nonfunctional requirements
need to be observed; for example, proper batch processing jobs are
properly configured.

End-to-end tests
After validating all the individual processes, it is time to connect all of
them and increase their complexity with new process variations. This is
the first test cycle that looks like a complete operation.

This test is manually executed, but automation starts becoming more

relevant in preparation of future iterations, regression testing, and
system maintenance because of continuous solution updates even
after go live.

The test is done by functional consultants who are preparing the cycle
and guiding the team, but the overall cycle is mainly executed by
customer SMEs and testers.

The main objective of this test type is to validate all full business
processes in scope and it needs to be done in an integrated test envi-
ronment since now it is connecting to other systems that interact with
the solution. It is iterative and the prerequisite to being able to execute
See it in action your user acceptance test (UAT).

On previous test cycles, we were able to

collect the cash of our sale. Now we want This test type starts in the Implement phase and goes through the
to connect other processes dependent on
this one; for example, accounting process
Prepare phase as per the Success by Design framework.
to report taxes, update and interact with
other systems like the transportation
system, optimize inventory, and It is important to execute as many of test cycles with end-to-end
introduce new products. This allows us to
testing, doing only one and at the end of the build of the solution is
combine different Dynamics 365 apps to
work together. not recommended since it add risks to confirming readiness by having
less time to react to final fixes.

Implementation Guide: Success by Design: Guide 14, Testing strategy 376

 Another important aspect here is that you run this test not just with
customer real data but with migrated data coming from the legacy
solutions as soon as possible.

Plan for this test by making sure This test is key to validating the entire solution works in conjunction with
you include real customer data and
other systems that are part of the business, and testing is done by having
migrated data.
role-based access control enabled so it validates a real end-to-end test.

Performance tests
Successful testing is not complete until we not only make sure we
can run the business on top of the solution, but also that we can do it
at scale. We are implementing Dynamics both for today and for the
future, and we need a solution that lasts and performs.

Performance testing is required, especially if there is concern over

operational volumes, peaks, variety of integrated systems, usability
challenges on the UX design, etc. In this section we focus on some of
the considerations you need to keep in mind for performance tests,
but we have dedicated an entire chapter just for this in Chapter 17, “A
performing solution, beyond infrastructure.”

We put special emphasis on this test type since there are many mis-
conceptions defining whether this test needs to be executed. Our
experience has shown that performance is one of the most common
reasons for escalation since teams often miss this test when it is needed.

In the end, a performing solution is a combination of data, code, con-

figuration, and infrastructure. Microsoft, partners, and customers play
important roles for this test type, but we look to the implementation
teams to play their important role on validating performance beyond
infrastructure and at the right time of the implementation, so plan for it.

This test is executed by developers, functional consultants, customer

SMEs and testers mainly.

The objective of this test is to ensure the solution performs while focus-
ing on critical processes that require scaling with load and growth. Not
all the processes are involved in this test.

Implementation Guide: Success by Design: Guide 14, Testing strategy 377

 This test is completed in a dedicated performance testing environment.
The basic version of performance testing starts during unit testing
so developers can influence proactive improvement of performance.
Regular dedicated test environments can also be used depending on
the load to be tested.

Performance testing happens as soon as the critical processes are

available to be tested throughout the implementation.

Performance testing needs to be visible at the start of the imple-

mentation and be part of the test plan. It requires agreement on the
performance test objectives, and it can require proper environment
planning to provide the necessary resources.

Do not delay performance testing till the end of the implementation

or avoid the review of the need for performance testing. This test can
bring to surface important challenges that can require fundamental
architectural fixes, so doing it late in the implementation puts the
entire solution at risk of wasting important resources.

It is important to mention that the infrastructure required to execute this

test type, especially when the plan requires a higher spec environment,
does not need to be available during all the implementation. You bring
See it in action
them a needed environment or repurpose ones for other test types
In previous example we involved during the lifetime of the project. This does not mean you should not log
invoicing and the integration with the performance bugs observed during other test types at any time.
transportation system. Now we want
to test if we can process the operation
peaks considering the seasonality of the
Remember, you are not just testing the standard product in this
business. Invoice posting operation is
crucial since we need to have almost real- test type; you are testing the solution that you are implementing in
time interaction with the transportation
system as per the business needs, so we
combination with your business operation needs. Some very basic im-
need to test a day in the life with relative plementations can justify not executing a performance test, but before
volumes across key business processes,
not just prospect to cash. deciding you can learn more about what you need to have a performing
solution in Chapter 17, “A performing solution, beyond infrastructure.”

User acceptance tests

We are now getting ready for prime time. Our solution works in unison,
it performs, and it is ready for final validation by the actual people that
execute the business, our end users. User acceptance tests are the final

Implementation Guide: Success by Design: Guide 14, Testing strategy 378

 calls for the business to declare readiness for go live.

This test is manually executed, never automated. It is executed with

customer data, including migrated data, and with the latest solution
version. This test is the closest to being like running live operations, it is
an actual business operation simulation.

User acceptance testing (UAT) can be a great

The objective is obtaining business sign-off of the solution, collecting
opportunity to start building automated end user feedback, helping to manage organizational change, and it is
testing for recording the tests so you can
have repeatability on testing. This sets the done in a dedicated and integrated test environment.
path for regression testing, optimizing the
investments to build that automation.
UAT is done during the Prepare phase of the Success by Design framework
UAT is the last comprehensive test before
going live, and every test case emulating
and it happens prior to preparing for go live, which is a prerequisite to
real operation is executed so end users can reach this milestone. For more details, see Chapter 18, “Prepare for go live.”
record their test as a proof of the test case
validation, therefore making this the first
step toward automation. This test type needs to be executed by users from the business, the
implementation team is just a facilitator. The business team is the main
owner of this test, failing to fully participate or test thoroughly is a risk.

See it in action
The business users must be trained prior to UAT, not just on the solu-
The business team now brings in a select
tion but also on how the pending process works once the solution
group of people who run the operation. is deployed to production; otherwise, this group causes false error
This group consists of order processors,
the accounting team, the accounts reports because the lack of training.
receivable team, shippers, and others. The
selected group of people run a real-life
operation simulation. All the processes are At the end of the successful test, the end user connects the new solution
executed in parallel and in-sync between
teams. For the prospect to cash process,
to its reality of their daily tasks and confirms the readiness of the solu-
they involve a selection of different end tion, but also their own readiness at being familiar with the new system
users connected to this process to run the
test. The team tests all the variations of the after being trained. The new solution fulfills the business operation need.
processes in scope.

The final iteration of this test type requires the business sign off on
acceptance. We describe the importance of this step in later sections in
this chapter.

This is one of the most important milestones in terms of testing types.

You are required to plan for the UAT
milestone and to have business stakeholders
This test happens at the end of the implementation and prior to the go
sign off on the acceptance of the complete, live. It is the entry criteria to prepare for operations. UAT is an end-to-
satisfactory results of this test prior to final
preparations to operate (go live). end testing approach where all the processes in scope are tested and it
requires sign off on the acceptance by the business stakeholders.

Implementation Guide: Success by Design: Guide 14, Testing strategy 379

Regression tests
Now we need to look at how to scale our future testing cycles, but also
to make sure prior positive test results continue as you keep evolving
the build for the solution.

Regression testing is testing that requires you to repeat previously

executed tests, due to changes or updates, to the solution to ensure it
stays healthy. Keep in mind that the solution is dynamic, new features
are introduced or new processes are required. Running regression tests
manually works of course, but it is also important that you consider the
need to prepare to automate this type of testing as soon as possible. It
is an investment that pays off with time.

The objective is to ensure the solution still performs as expected after

a change and that it does so in a pre-production environment, test
environment, or development environment.

A regression test happens when you have change in the code, or any
configuration or new solution pattern that can impact different pro-
cesses. This test type is done by testers, developers, and end users.

It is important to perform this test type before change is introduced to

the production environment.

As we realize that this necessary quality gate is needed, there are tools
available to help you to automate. You can find links to more information
about these tools at the end of the chapter in the “References” section.

There are different techniques you can follow to run your regression test.

You can opt to test almost 100 percent of the processes, which can
provide you comprehensive coverage, but it is expensive to run and
maintain especially if there is no automation.

You can prioritize the test based on business impact. This technique
ensures you have coverage of the processes that are mission critical for
the operation, it is also a more affordable approach and the only risk is
that you cannot guarantee a solution 100 percent free of regression.

380
 Another technique is to focus on the areas that you expect to be
impacted based on the change. This technique is targeted where the
change is happening and requires less effort, but it has the limitation
of not being able to confirm that you are free of regressions and the
impact cannot be visible on the direct change itself, but instead down-
stream on other processes.

You can also combine all the previous techniques, making sure you
test critical business processes, and you can target more testing on the
features being changed.

Always keep in mind that bugs discovered late in the implementation

process have a lifecycle to get fixed but also require testing. Testing
allows you to detect those bugs early, and if the bugs are detected late
in the process, let’s say in the last test cycle, to think twice on the value
of the fix versus the risk it introduces with not being able to do proper
regression testing. The fix can be more costly at that point, being so
close to go live, than waiting and doing a proper regression testing or
delaying the go live.

Every time you test successfully, you put money in a trust. Every time
you have a failed test or fail to properly test, you deduct money from
that trust leaving a debt in the solution completeness. At the end, you
need to reach the end goal with the trust full of money.

Again, automation is key, and you should plan for testing automation.
See it in action Your solution is alive and ever evolving so change is a constant, not just
coming from the application, but from the customer business needs.
Microsoft has released new functionality
that enriches the order processing feature
in Dynamics 365 Finance and Operations Start building your automated testing progressively, focusing first on
apps. When the new update becomes
key business processes then expanding further over time. Do it as early
available, teams can execute regression
testing to key business processes as possible during the implementation process and always test it after
connected to this new feature. Testing the
order and warehouse processes following
a change and prior to production deployment.
a new configuration change is done
because it can impact the picking process.
Once the update is done in the test Regression is important but it can be costly, especially if you do not
environment, the team runs an automated automate in the long term. Automation brings the benefit of testing
test to confirm the solution produces the
expected outcomes. faster, having better testing coverage, and providing repeatability of
the test. Automation takes as much planning as was needed for the

Implementation Guide: Success by Design: Guide 14, Testing strategy 381

 manual testing planning, but it is a long-term investment that pays
itself off with time.

Finally, keep in mind that finding bugs during regression testing could
be due the change of the solution design, a standard product update,
or just the resolution of previous bugs redefining how the solution
works, which can require recreating the test case.

Mock cutover
This is a special test since it is executed in a special environment, the
production environment.

This is a temporary use of the production environment to do testing

activities and it happens during the Cutover phase. To learn more
about the cutover activities refer to Chapter 18, “Prepare for go live.”

This test brings important value because it helps to validate aspects like
connectivity, stability of the access by users and data, integration end
points configuration, device connectivity, security, network, and many
of the test cases in your processes may be environment dependent.
See it in action
During this test, you are able to validate confirmation of the estimated
Continuing with the previous examples
times between environment preparation for production for all the
in this chapter, the team has tested
the entire solution and they are ready planned tasks during go live. It is a confirmation that all the planned
to move forward. The team wants to
confirm the cutover plan in terms of
activities run smoothly once you do the actual final execution.
sequence of activities, timing, and
production environment stability. They
execute the cutover plan by rehearsing Once the testing is completed you rollback your data up to the point
and documenting so they can adjust that there are not transactions being created because of the testing,
it. After executing the mock cutover,
they found the time to load data and allowing you to finish the environment preparation and start running
sequence it required adjustment due to
your system for real. If the mock cutover fails, there is a risk of delaying
other conflicting cutover tasks. The team
adjusted that sequence and confirmed the the go live.
readiness to execute the updated Cutover
plan knowing it works without surprises.
We recommend you always plan for a mock cutover test type.

Other types of testing

We have described the incremental approach for testing through
the different test types, scale needs, and continuous validation of the
solution. But there are other types of tests that can be unique for every

Implementation Guide: Success by Design: Guide 14, Testing strategy 382

 implementation based on the design of the solution. These test types
are below.
▪ Smoke tests  Where you test if basic functions work as expected
so you can confirm the component can continue to test further
with other test types.
▪ Data acceptance tests  You validate if the migrated data goes
into the target environment correctly and if it is usable for actual
operation.
▪ Interface tests  Here you confirm if your interfaces can be utilized
as intended.
▪ Environment testing  In this test type, you validate if the new
environment is fit for the purpose.
▪ Reporting testing and business intelligence testing  Validates
that reporting and Business Intelligence can be executed and
operate properly with the expected outcome and performance.
▪ Device testing  Focused on validating devices are operational,
connect, and perform as expected, for example, RFID readers,
scales, and warehouse devices, etc.
▪ Network/infrastructure testing  Validation of relevant underly-
ing networking, firewalls configuration, Wi-Fi, printers, etc.

While some projects are too small to justify separate planning to

execute some of these types of testing, the concept and meaning of
the tests should be folded into other types of testing.

Test types and outcome ownership

We described the importance of defining the owner of the outcome at
testing, now we need to connect this ownership role to the test type.

In other words, who takes accountability to drive the resolution of the

outcome derived from testing based on the test type. It is important
we have clear ownership to avoid defects bouncing around with
unclear roles driving the actions, even though the fix can come from
When you define your test plan,
different roles. As an example, unit testing outcomes tend to be owned
define ownership for the type of fixes by partner technical architects or dev leads. User acceptance testing
but also who is accountable to drive
the fix overall by the test type. outcomes tend to be owned by customer steering groups devolved to
lead SME or PM.

Implementation Guide: Success by Design: Guide 14, Testing strategy 383

 Test types and test plans
Test plans are living documents that keep evolving since project needs
can change in terms of adding new test cycles or test types. This is due
to business needs changing or the need to increase the quality due to
unknown factors at the start of the project. For the most important test
types, like user acceptance testing or performance testing, you need to
create specific test plans or keep enhancing your master test plan while
always bringing enough details to provide clarity for each of them.

The bottom line on

defining a test strategy
Defining a test strategy for your implementation is a combination of
considering the project scope, testing planning with a test plan and test
cycle schedule in combination with the phases and solution versions
availability, and selecting all the test types and how you execute them.

Testing is iterative and incremental; it grows as you progress in the

implementation and it is a permanent activity once you are live based
on the regression testing technique you select.

Always test under the umbrella of the processes, in the end the pro-
cesses are the language of the business and the testing is proof that
the solution “can speak” that language.

Executing testing
In a previous section, we focused on the importance of defining a test-
ing strategy. Now we are going to explore the minimum components
you need to keep in mind to execute the testing.

The prerequisites required to move to execution are having a test plan,

test cases, and clear testing cycles that are defined according to the
progressive readiness of the solution.

Implementation Guide: Success by Design: Guide 14, Testing strategy 384

 Now we focus on the tactical components for prepping the execution,
what to keep in mind to communicate the scope, and how to track the
progress and outcome.

Communicating the plan for the test cycle

Aligning the teams prior to testing execution is important so that
everybody stays focused on the expected outcome of the test cycle
being executed. This is especially important when you test with several
people involved, like process testing or functional testing.

During the communication effort you share the test plan for the test cycle.
You need to communicate the following with the team during testing.
▪ Scope  Before you start testing, you describe the scope and pur-
pose of the testing cycle, and what processes, requirements, and
tests cases are included in that test cycle. Every new testing cycle
requires detail on how the iteration has been designed in terms of
incremental testing and what is expected to be tested. The scope
of the test cycle is aligned to the solution version being used.
▪ Schedule  The time expected to be used to run the test cycle.
▪ Roles  Who is testing and how are the test cases distributed?
How do they report test case execution so dependent teams can
continue the testing? Who is the orchestrator of the test? Who
resolves questions to test cases per area?
▪ Resolution process  One of the objectives to test is to identify
any defect in the solution. The communication plan needs to
specify how those bugs are reported but also how to document
the feedback from the tester.
▪ Progress  How is the progress recorded and communicated so
everybody can see the current state of the testing cycle?
▪ Resources  The communication needs to specify where the
testing happens and how to access the apps. It determines any
additional equipment needed for testing, for example printers, bar
scanners, network requirement, etc.
Part of a successful test cycle is to set the
▪ Test sequence  Especially on process test, end-to-end test, and
expectations with the participants, so
everyone keeps the focus on the objective. user acceptance test types. You need define and align how the
Consider the test cycle like a mini go live
for the scope in place for the cycle.
different teams interacts.
▪ Test objectives  Here you explain the purpose of the testing cycle.

Implementation Guide: Success by Design: Guide 14, Testing strategy 385

 Tracking during testing
Detecting bugs and properly documenting them is key. It facilitates the
resolution. But before reporting bugs, the tester needs to understand
the scope and objective of what the test is. Building the solution is pro-
gressive and it is easy to report issues where the scope of the test does
not include a solution for that and is part of future iterations. Keep the
focus with the team testing.

Tracking captures the outcome and performance of the test cycle; it is

how the development team and consultants understand the quality of
the test cycle. Azure DevOps is a recommended tool for this tracking
since it integrates into the entire application lifecycle management
and any bug reported is visible to the team that needs to fix it, but also
shows what is or is not working in general.

Let us explore some tactical components of the tracking of issues

discovered during testing.

Keep track of changes in configuration

It is likely some tests result in tuning configurations as you test.
Adjustments to the solution in configurations can change the behavior
of the app, and those changes can happen during tests. This brings
flexibility into the resolution, but also raises the risk of the team not
documenting those changes. Keep in mind that for the next test cycle
Making configuration changes can invalidate
previous test cases connected to those the data can be reset and if you do not document and make sure those
changes. Those changes need to be handled
changes are not reflected in your golden configurations, you lose the
much like it is a production change to make
sure those changes do not negatively impact important opportunity of what you fixed during testing. In the end,
previous passes test. Track and execute
regression testing.
you do not want to repeat the same research over and over, so docu-
ment the configuration update.

Keep record of the progress

At every test cycle you come with a scope. Depending on the type of
test you have some dependencies. Teams need to know when they
should start testing each type of testing. A testing heatmap can help
you to show the progress of testing by process, test cases, and its
statuses. For example, in the scenario used to explain the test types in
previous sections we used the analogy of testing a prospect to cash.

Implementation Guide: Success by Design: Guide 14, Testing strategy 386

 A meaningful process test type makes sure the process being tested
has continuity, from the order to the invoice and shipment and later to
the collection. If you are testing a sequence of processes, you do not
go ahead of the game and straight to the collections part. If you do,
you lose the opportunity to properly test the larger process sequence.

Keep record of the bugs

You know how to report the bugs and that the reports need to be ac-
curate. Make sure you can report the issue, but also that the case being
tested is in scope for that test cycle. Not doing that creates rework in
the investigation for resolution. Also, be factual. Is the test passing or
not based on the test case documentation and the expected outcome
when following the exact steps?

Bugs versus gaps

One common issue during bug reporting is that we can mix a bug with
a required feature or gap. Stick to the test case, if you identify a gap,
great! Just make sure the feedback is properly classified. We do not
want to lose the opportunity to capture that great idea or finding, we
just need to funnel it properly to the right tracking system.

Dealing with the

outcomes
Finishing the test cycle requires reporting the results and maintain-
ing the bugs discovered during the testing, but also funneling the
feedback received. In the end, all these outcomes add onto the top of
the next wave of processes or features to be introduced in future test
cycles, as shown in Figure 14-7.
Fig.
14-7

Reporting the outcome measures

% Failed the performance of the test cycle
Performance of Scope
and helps to identify areas of op-
the test cycle portunity for the implementation
% Passed
team since it triggers clear actions
to improve the quality, so the next

Implementation Guide: Success by Design: Guide 14, Testing strategy 387

 test cycle provides better performance.

For the next test cycle, we start over, as even passing test cases during
Sign off by the business team on the
this cycle need to be re-tested in the next one.
overall pass of the final test cycle is a
required step prior to executing cutover
preparation for go live. This creates

Solution acceptance
accountability on the acceptance of the
solution by business. Any non-blocker bug
needs to be document and they should be

and operation
low risk for the operation at not bringing a
solution at going live without them.

Finally, the last test cycle takes the main objective of why you started
implementing the solution to confirm the readiness to run the business
using the new solution.

We described the importance of reporting the outcome of the test

cycles for user acceptance test types, this report helps to determine
the readiness to prepare for go live and is the foundation of the go or
no-go decision by the stakeholders. The performance should be high
at this point, and if not? Then the solution is not ready.

This confirms the readiness of the solution but also the readiness of the
production environment if you run a mock cutover test. It is important
that for any final test the business team signs off and confirms they can
operate the business with the new solution.

Accepting the solution does not mean it is 100 percent free of bugs.
You need to assess the value to bring a fix, at this point if the issue is
very low risk for the operation it is often better to go live with known
nonblocking bugs and fix them later than to introduce risk by making
an unnecessary rushed fix and not having time to re-test properly.

From here you move to the maintenance mode of the solution if you
are implementing only one phase, or you continue adding new work-
loads or expanding the solution. Keep the discipline for testing and
scale using the tools for automation. Testing is a continuous practice,
the difference is the frequency, scope, and tools used to test.

Implementation Guide: Success by Design: Guide 14, Testing strategy 388

References Conclusion
Finance and Testing is about quality, it is used to confirm the effectiveness of the path
taken by the implementation team to solve the challenge of how to
Operations apps operate the business with the new solution, it confirms the initial vision
Test Strategy TechTalk by putting it into action. It is a required step in the process to implement
Dynamics 365, regardless of how simple or complex your solution is,
Performance benchmark
or if you are implementing standard functionality or extending or
Performance patterns and antipatterns customizing it. If you run into a large number of quality related issues
after you are live, it is generally attributed to ineffective testing.
Performance troubleshooting tools

Performance testing approach Testing during implementation is the step that builds the trust for the
business to run their operation. Always test and do so as early as possi-
Regression suite automation tool
ble. It is never too early to start testing.

Customer
Engagement apps
FastTrack automated testing in a day
workshop offering

EasyRepro – Getting started

EasyRepro in 60 Seconds

Automated and manual testing with

Azure Test plan in Azure DevOps

Build end-to-end UI tests for your

canvas app using Test Studio

389
 Checklist

Planning Carry out performance testing throughout the

Implement and Prepare phases on critical processes to
Plan the test case creation, forecast the resource re-
ensure the solution performs while focusing on scaling
quirements (people, hardware, environments, devices),
with realistic load and growth.
and implement the test cycles in accordance with the
processes in scope. Throughout the Implement and Prepare phases,
perform iterative and cumulative end-to-end testing
Have a tracking mechanism that monitors the overall
of all full business processes in scope and connected
progress, pass/fail rates, ownership, and issues tied
integrations and solutions in a test environment with
to specific test cases, and have a process defined for
increasingly “real” migrated data.
remediation, retesting, and deployment when errors are
found and corrected. Perform UAT in a test environment during the Prepare
phase to test the preparedness of users, processes, and

Implementation the solution to operate the business.

Ensure the Software Development Lifecycle (SDLC) Consider automation for functional, process, end-to-
includes unit testing by the developer in a development end, performance, and regression testing.
environment, with focus on function, code, and configu-
Establish a test communication plan with clear scope
ration of all extensions and integrations.
and objectives, specifying the schedule, sequence, roles
Ensure the functional testing is performed by functional involved, and issue resolution process.
consultants or SMEs in a test or developer environment
Perform regression testing in a test environment, or
with the primary objective to validate the design against
development environment for updates, throughout the
the requirements.
Implement and Prepare phases as well as post-go live to
Ensure process testing is performed by SMEs in a test ensure the solution performs as expected after a change.
environment, running multiple processes consecutively,
focusing on testing whether the solution allows business
Go live
operations and monitoring for unintended outcomes.
Plan a mock cutover in the new production environment
to validate confirmation of the estimated times between
environment preparation for production for all the
required tasks during go live.

Implementation Guide: Success by Design: Guide 14, Testing strategy 390

 Case study
The right environment for the
right test: A missed opportunity
A customer in the manufacturing industry that manufactures rec-
reational vehicles is implementing Dynamics 365 Supply Chain
Management and has created a testing strategy that included four
testing cycles, including user acceptance testing, as part of their plan.

The team included the most common test types except for perfor-
mance testing under the assumption that the first rollout would not
require it since the extension of the system was low, and the customer
will be implementing just manufacturing business unit. Next rollouts
will include the servicing business, including the warranty manage-
ment of recreational vehicles plus other operations. The servicing
operations will continue to be executed as a third-party solution for
now and their dealership network will be using this system to order
parts and honor warranties.

Very soon, during the first testing cycle, the team started to test across
the first wave of the solution where there were no integrations, so
completing testing was satisfactory in terms of the cycle performance.

As the team got ready for their second wave of testing, they started
to introduce some of the integrations as part of the scope for the test
cycle, but those integrations were emulated with dummy data and the
team introduced some migrated data. Integrations worked for their
purpose and the volume of migrated data was small.

Implementation Guide: Success by Design: Guide 14, Testing strategy 391

 For their third testing cycle, things looked great and they were ready
to do a more comprehensive testing by using end-to-end testing, this
time with real integrations and real volumes of data migration.

At combining the testing for having a larger volume of users for the
test and integrations running with similar volume of operations to
what they expected in production, the team hit the first major blocker.
The solution was not keeping up with the needs of the test cycle, and
they were not even in production. The first reaction from the team was
an underperforming infrastructure, which raised the concerns of the
business stakeholders upon learning the test cycle outcome.

The team was ready to prepare for UAT and decided to continue ex-
pecting that this would not be an issue in production due to it having
higher specs. They assumed the production environment would be
able to solve this performance challenge, so they decided to continue,
complete UAT, and move to production. The customer signed off and
preparation moved to the next stage to get ready for go live.

The big day came and production and all the departments switched
to new system. The first day was normal and everything was working
great. The team decided to turn on integrations for the servicing
solution on the second day. When the second day came, they were
ready to go interconnect with the service department, and integrations
started to flow into Dynamics 365. This is when they had their first
business stopper: they had a sudden decrease in performance, users
were not able to transact in Dynamics 365, service departments from
the dealerships were not able to connect effectively, shipments started
to slowdown, and the shop floor was having a hard time to trying to
keep inventory moving to production.

The implementation team was providing hyper-focused care and

immediately started to check for what could be blocking the operation
and they found the problem. The integration was creating a bottleneck
by making important OData calls to check inventory, creating sales
orders, and invoicing while connected to the dealership’s partners using
the servicing solution. They stopped the integration and things went
back to normal, but it was clear that the integration was the cause.

Implementation Guide: Success by Design: Guide 14, Testing strategy 392

 The team found a solution after troubleshooting, but it will require a
change to the integration design pattern and fast movement to solve
it and keep operating. Part of the revenue from the recreational vehicle
business comes from servicing and providing parts to the dealerships.
It took days of manual work to keep things working during this painful
first week of go live, which impacted the perception of the organiza-
tion in the solution.

In retrospect, validating designs for integrations with a volume of data

was key to preventing this scenario and the issue they had during real
operation. The team’s morale was affected after working hard during the
implementation season, all because they missed one important test;
the performance test. The team thought emulating early and validat-
ing in production would be OK, but the cost of that decision was high.

This team learned that including performance testing as early as possi-

ble, with actual volumes in the right environments that perform at the
same volumes expected in production, would help to detect a design
problem and correct it in time. The team also learned that “testing” in
production for the first time, and with the real volumes, was not the
best idea since the time available to make any fixes was reduced due
to the live operations and correcting as they go introduced bigger
risks. The assumption of the team that they were implementing almost
standard functionality led them to think that the solution would per-
form fine, but the actual customer challenge made the solution unique
which requires extensive testing.

For the next phases, this customer included performance testing. They
dedicated an environment to stress test the solution with their own busi-
ness needs and they executed these tests earlier with additional scenarios
and UAT that included parallel processing of people testing and inte-
grations running. They were able to have second go live to include their
accessories manufacturing business and it was a breeze in comparison.

It is never too early to start testing and do so according to the project

needs. The later the testing, the bigger the risk.

Implementation Guide: Success by Design: Guide 14, Testing strategy 393

15 Guide
Extend your
solution
Implementation Guide: Success by Design: Guide 15, Extend your solution 394
The power of making it your own.

Introduction
Business solutions offer a rich set of capabilities
to help drive business value.
Still, in some situations, you need to extend the solution and adjust off
the-shelf functionality to accommodate organization or industry spe-
cific business processes. These adjustments can change how a feature
works or bring additional capabilities to meet specific requirements.

While Dynamics 365 apps natively offer rich capabilities, they also
offer powerful options to customize and extend them. Extending the
solution can open even more opportunities to drive business value. It
is important to note, however, that extending should not compromise
the fundamental advantages of an evergreen cloud solution, such as
usability, accessibility, performance, security, and continuous updates.
Complex business requirements lead to
highly advanced solutions with customiza- These are key to success and adoption.
tions and extensions to applications.

Defining your
Advanced implementations bring an
increased risk that user experience suffers
because of the introduction of perfor-

extensibility strategy
mance, stability, maintainability, sup-
portability, and other issues.

In this chapter, we delve into key factors to

consider when determining the scope of
depth of solutions. When organizations shift toward software as a service (SaaS) cloud
solutions, one of the main reasons they do so is to enjoy the many

Implementation Guide: Success by Design: Guide 15, Extend your solution 395
 advantages, including out-of-the-box security, native integrations with
other cloud solutions, cost savings, and improved agility. The ever-
green cloud approach also enables the continued ability to evolve with
the latest modern features. In addition, it is often easier to use off-the-
shelf solutions with embedded modern processes and further extend
it. This tailors a solution to meet business needs, rather than having to
build a custom solution from scratch with all the associated hassles and
expense to meet the precise requirements of your organization.

What is extending?
Even though Dynamics 365 apps cover a wide variety of business
processes and industries, there is often some degree of specialization
or customization needed. , We refer to this as extending the solution.
Extending can vary from minor changes to the user interface of a
particular feature to more complex scenarios like adding heavy
calculations after certain events. The depth of these extensions has
important implications on how much the off-the-shelf product needs
to change to meet specific business requirements.

We can also categorize these extensions into distinct types:

▪ Extending by modifying off-the-shelf functionalities.
▪ Extending by introducing new functionalities on top of a native one.
▪ Creating custom solutions on the platform.

Determine the need

It is important for organizations to be aware of the impact of extend-
ing the solution. They want to strike the right balance between the
features required to empower a user or bring business value, as well as
efficiency and the value it brings compared to what it costs.

A key driver in the decision to extend is understanding the potential risks

of extension to key solution characteristics, like its robustness, reliability,
performance, usability, security, and ability to take updates. Consider
the example of extending an invoicing system that could pose risks to
the operations or could disrupt the business because of increased com-
plexity; such risks are best avoided. When opting to extend the solution,

Implementation Guide: Success by Design: Guide 15, Extend your solution 396
organizations may also depend on a third-party to maintain and evolve
it, and in some scenarios an unresponsive third-party can block the
organization from adopting new functionalities.

Extending to integrate with other solutions is a common scenario that

we cover in Chapter 16, “Integrate with other solutions.”

Do not replicate your legacy solution

When gathering solution requirements, organizations may want to
reproduce every piece of functionality from the legacy solution or even
try to mimic the user experience. This could come across as a strategy
to reduce user resistance and drive adoption, but it can also lead to a
highly customized solution that fails to leverage the strengths of the
new platform in favor of familiarity.

Legacy solutions may have taken years to develop and evolve and may
not use the latest and greatest functionality available on the market. As
an example, Dynamics 365 apps natively use artificial intelligence and
machine learning to provide insights that help users make the best and
most informed decisions.

Major deviations from the native features or repurposing existing

features can limit the value of an evergreen platform by compromising
flexibility and the ability to evolve with changing business needs. That
increases technical debt and maintenance costs.

Understand the power of the platform

Another challenge arises when an organization does not fully know the
power of the new solution and opt to extend before even understand-
ing the impact of their decisions. The platform may have numerous
ways to achieve the same functionality as the legacy solution already.
Reasonable customization of the off-the-shelf solution might be re-
quired, but the point is to find the right balance. Thus, each scenario
should be evaluated, explored, and in some cases piloted. And of
course, it’s important to involve key business users in these decisions.

397
 Leveraging ISV solutions
Leveraging independent software vendor (ISV) solutions from the
app marketplace instead of extending the solution to achieve the
same results may save development cost and time, as well as testing
and maintenance resources. ISVs typically support and maintain the
solution at scale for multiple organizations. Their experience can be an
advantage for organizations that require additional functionalities that
are already provided by ISVs.

When deciding to use ISV solutions, organizations should be aware of

the ISV support model and its alignment with continuous updates as
well as protection clauses.

Extensibility scenarios
Extending off-the-shelf solutions occurs when functionality is changed
to fulfill an organization’s requirements.

Customizations and extensions performed to out-of-the-box appli-

cations could stretch from simple setting and configuration updates
to customized platform as a service (PaaS) solutions that extend their
functionality. In this section, we explore levels in which customization is
possible and discuss their relevance.

Some organizations prefer to stay as standard as possible. That does

not mean that they cannot make simple changes with configurations
or low-code/no-code customizations using modern tools. Staying
standard allows for slight changes, as long as those changes do not
highly impact key characteristics of a solution.

App configurations
Configurations are the out-of-the-box controls that allow makers and
admins to tailor the app to the needs of a user. These setting changes
are low effort, requiring no support from professional developers. They
are a powerful way to make the application your own, for example

Implementation Guide: Success by Design: Guide 15, Extend your solution 398
changing a theme to reflect business branding.

In some cases, additional due diligence might be required for config-

urations, even though they are tested and shipped as part of the core
service. Examples include security model configuration, out-of-the-box
integration setup, and enabling embedded intelligence. Also consider
new capabilities delivered via the service updates that might be con-
trolled using the app settings.

App settings are the safest and the least impactful way of tailoring the
solution to your needs and should be the preferred approach before
exploring another extensibility technique.

Low-code/no-code customizations
A differentiator for Dynamics 365 apps and the latest generation SaaS
products is the powerful customization capabilities made available
through “what you see is what you get” (WYSIWYG) designers and de-
scriptive expression-based languages. This paradigm helps significantly
reduce the implementation effort and enables businesses to get more
involved with design and configuration.

This low-code/no-code approach also guides makers to take ad-

vantage of the optimized patterns and best practices, avoiding the
potential pitfalls and antipatterns that are observed in a complex code.
Because the platform takes care of lower-level details and automati-
cally optimizes the logic, testing is focused on functional components.
The impact on performance, usage pattern and service boundaries,
however, still needs to be taken into consideration.

Typically, customizations result in user interface modifications and

changes to the field settings, security model, business processes,
visualization components, etc. Changes implemented through custom-
izations are performed to ensure data quality, reduce repetitive tasks,
and adjust visualization components that determine how information
is organized and displayed to the users.

With customizations, applications can be tailored to more closely fit

399
 organization requirements that are specific to an industry or unique
business processes, including specific functionality focused on specific
roles or personas. This allows personalization that streamlines the user
experience so a user can focus on what is most important.

Low-code and no-code customizations are the preferred extensibility

approach when the requirements cannot be satisfied by app configu-
ration changes.

Extend using pro dev

Over the years the low-code/no-code base configuration capabilities
have advanced, reducing the dependency on professional develop-
ers for application changes. Edge cases or scenarios that cannot be
achieved using the configuration-based approach can still exist. In that
case, the powerful custom code execution capabilities of Dynamics 365
apps can be leveraged.

Dynamics 365 apps support extension programming models that

enable organizations to leverage existing expertise within the orga-
nization for advanced scenarios. These models, depending on the
specific app and functionality extended, generally support .NET based
languages, Software Development Kits, and open-source libraries
around JavaScript and HTML5.

The key aspect to keep in mind when using code-based customization

is to understand the extensibility framework of Dynamics 365 and
only use the documented extension patterns. Use of unsupported
techniques breaches the service level agreements and can have a
potentially severe impact on the live service.

It is important that your developer teams have the latest software

development toolkits available to them. They should also know about
community based third-party tools and samples. These can drive
productivity even though Microsoft may not provide direct support
for them.

Implementation Guide: Success by Design: Guide 15, Extend your solution 400
 Extending into PaaS
In some scenarios, organizations leverage PaaS components to extend
solutions, which adds powerful capabilities that help address complex
requirements. Dynamics 365 apps has a core design philosophy that
allows our SaaS applications to be extended by leveraging the underly-
ing Azure PaaS capabilities. This is referred to as the no-cliffs extension
approach. The approach enables businesses to start with SaaS, and
then for the most complex long-tail scenarios, extend into the Azure
PaaS. Doing so alleviates the fear of being limited by the platform.

This no-cliffs extension provides the best of both worlds. The SaaS appli-
cation provides the off-the-shelf functionalities as well as the options and
methods to extend them. The PaaS extensions further enrich the solution
architecture by providing rich and powerful mechanisms that scale and
allow heavy processing of operations outside of the business solution.

An example of how this approach is natively used is the Internet of

Things (IoT) Intelligence add-in for Finance and Operations apps. This
add-in integrates IoT signals with data in Finance and Operations apps
to produce actionable insights, as illustrated in Figure 15-1.

Fig.
15-1 Dynamics 365 Dynamics 365
Finance and
Operations apps
Azure IoT IoT Intelligence Add-in
Central
Workspaces

Machine Product Production

status quality delays Dashboards

Connected Data
technology ingestion
Alerts

Analytics Insights Actions

Automation

The same happens with Connected Field Service. Connected Field

Service for IoT Central provides a direct integration of Dynamics 365
Field Service with Microsoft Azure IoT Central, an add-on solution

Implementation Guide: Success by Design: Guide 15, Extend your solution 401
 Connected field service

Fig.
that brings in Azure IoT PaaS on top of Dynamics 365 Field Service, as
15-2
shown in Figure 15-2.

In both examples, organizations can leverage PaaS to further extend

Azure IoT
Central the solution. The same can be applied to specific organization require-
ments that can use PaaS features as part of the extension’s architecture.
This approach can be particularly valuable because it reduces the
maintenance requirements for parts of the solution compared to a fully
custom-built extension.

One example of this approach is when organizations leverage Azure Logic

Power Automate
Apps. Logic Apps provide a serverless engine to build automated work-
flows to integrate apps and data between cloud services and on-premises
systems. Logic Apps provide the ability to trigger workflows based on
events or timers and leverage connectors to integrate applications and
facilitate business-to-business (B2B) communication. Logic Apps are
integrated seamlessly with Azure Functions, as illustrated in Figure 15-3.

Dynamics 365 Azure Functions is an event-driven serverless compute platform that can
Field Service
also solve complex orchestration problems, by using it you can move
some of the heavy computing processes away from Dynamics 365 apps.
Fig.
15-3

With Azure Functions organizations can code their

Azure functions
Structured query
language (cloud)
own workflow definition and simplify complex,
stateful coordination requirements programmati-
Service bus Office 365
Logic applications cally in event-driven applications, as well as taking
advantage of numerous native connectors.
Cognitive Services Microsoft
Dynamics 365

Machine learning DocuSign

A multitude of examples demonstrate where PaaS
On-premises data gateway can be leveraged to further extend the capabilities
of the solution. Leveraging a platform helps elim-
inate the costs and complexity of configuring and
managing the required infrastructure, while empow-
SharePoint Structured query IBM DB2 BizTalk server
ering development teams to focus on building the
language
apps and services that drive value.

Implementation Guide: Success by Design: Guide 15, Extend your solution 402
 Considerations
Every piece of an extension should focus on bringing efficiency or
value to the organization. Inherent costs do exist when implementing
changes and can vary from the cost of building, testing, maintaining,
and support. These should be taken into consideration when planning
to extend a solution.

In this section, we delve into the key considerations and impacts exten-
sions can have on a solution.

Usability and accessibility

User experience is a key characteristic of a business solution. Providers
invest heavily in the user experience to ensure that the users have a
pleasant experience when interacting with the solution, and that the
solution operates seamlessly between devices, such as laptops, tablets,
phones, and platforms like Outlook and Microsoft Teams.

We earlier discussed that the purpose of extending off-the-shelf func-

tionalities should be to add further value to the business. One way to
achieve this is to improve user experience. This is achieved by consider-
ing and adapting user requirements and expectations.

While the main purpose of extending may not be to improve the user
experience, it should not negatively impact the user experience and
how the solution behaves across different devices and platforms. In
addition, extending should not negatively impact the responsiveness
and performance of the solution.

Security, privacy, and compliance

Security and compliance are important aspects of a solution. When
extending, it is crucial to honor the security and compliance require-
ments. Access to data, either implemented by security roles or any
other feature, should not be bypassed when customizing or extending
the native functionality. Breaking the security model can cause a large

Implementation Guide: Success by Design: Guide 15, Extend your solution 403
 negative impact, and security breaches are not to be taken lightly.

The same happens with compliance, for example the General Data
Protection Regulation (GDPR) policies. Compliance functionality im-
plemented natively also need to be inherited in any customizations of
extensions. Failing to do so may have consequences for organizations
that do not comply with regulations.

GDPR is just one set of regulations. Regulation of privacy and data use
exists in many different forms across several markets. While there is a
great deal of overlap in terminology, privacy and security are not identical.
Security is about preventing unauthorized access to any data, while priva-
cy is ensuring, by design, proper acquisition, use, storage, and deletion
of data defined as private under local regional and global regulations.

While security is natively built into Dynamics 365 apps and highlighted
as best practices, privacy requirements tend to have a higher probabili-
ty of being overlooked by developers when extending native apps.

Performance
Although cloud solutions provide a high degree of scalability and
performance, when extending a solution, it is important not to com-
promise its performance.

When extending the user interface and/or the business logic, addi-
tional efforts are added to create, retrieve, update, or even delete data.
Those additional efforts may have an impact on the user experience,
depending on the amount of extended functionality added.

Service protection limits also exist to ensure consistent availability and

performance. These should also be considered when extending the
solution to avoid random or unexpected surges in request volumes that
threaten the solution’s availability and performance characteristics.

Performance goals should be identified, agreed upon, signed-off on by

the business, and measured as part of the testing strategy to identify any
deviations from those goals. These goals are especially important when

Implementation Guide: Success by Design: Guide 15, Extend your solution 404
the solution is extended, as those extensions can highly impact them.

Depending on the environment strategy, organizations may have

globally distributed solutions where factors like latency, firewalls,
network traffic, organization proxies, and routing by Internet Service
Providers (ISPs) are important to ensure a good user experience. When
extending a solution, especially integrations with external systems, the
number of external calls, round trips, and volumes, on top of all previ-
ously mentioned factors, are crucial to ensure that the solution meets
the performance goals.

Scalability
The scalability of a business solution is also a key consideration to de-
termine how you extend it. While the cloud platform includes scalable
servers and micro services, other aspects of the platform need to be con-
sidered to determine the impact of your business solution architecture.

Parameters, such as the number of parallel connections, requests, and

concurrent users, can determine how well the business solution scales,
for example when new business units, countries, or regions are added.

It is better to assess the impact of the chosen architecture as use of

the solution grows, for example, when the scope is small or when the
solution is first released.

When extending a feature, it is also important to understand how the

functionality applies to different business units, countries, or regions,
and what is the level of adjustments needed for each of them. If each
business unit, country, or region requires its own level of customiza-
tion, the solution at the extension level is not scalable by itself and
needs further adjustments.

Service protection and limits ensure consistent solution availability and

performance, as well as a level of protection from random and unex-
pected surges in request volumes that threaten the availability and
performance characteristics of the platform or solution.

405
 When extending the solution, it is important to understand how use
grows over time and the impact of the design on parameters like:
▪ How much storage is required by the extensions and how does it
grow over time?
▪ How many additional application programming interface (API)
calls do the features require?
▪ What are the limits on workflows, code execution, or API calls?

Impact on application lifecycle management

When extending the business solution, all code and customization
must be packaged and shipped from their various environments. This
means extensions need to be included in application lifecycle manage-
ment (ALM). Typically, custom code, customizations, and development
created from the user interface or pro dev tools need to be developed,
tested, approved, and released into production. In addition, for easier
maintenance, they also should be added to a repository where they
can be backed-up, merged, and properly handled.

This means that all extended functionalities are added on top of ALM
practices, which also increases the complexity. As an example, if the
extended features required different configurations, or are only ap-
plicable to specific business units, countries or regions, there needs to
be a separation at the ALM process, which may just be that they are
shipped in different packages (or solutions).

Aspects of the packages should be considered, depending on the num-

ber of extended functionalities added to the native solution. Examples
include the order these functionalities are introduced, how they are
split, and how the size affects the ALM operations.

Maintainability, supportability,
and future-proofing
In a cloud world where updates are continuous, and new features are
released with frequency, it is important to think about the impact and
costs of maintaining extended functionalities.

Implementation Guide: Success by Design: Guide 15, Extend your solution 406
 When extending a solution, additional maintenance requirements are
added to the business solution, so it is important to understand and be
aware of deprecations and roadmap. This helps avoid building something
that might soon be offered natively by the solution or being forced to re-
place parts of the extended solution because of deprecated functionality.

Supportability
Extending a business solution can also complicate the support re-
quirements of the solution. Normally, the first line of support is at the
organization itself or a vendor. Support resources must have expertise
on the business solution and extensions built on top of off-the-shelf
capabilities. This requires a specialization of resources to support the
extended business solution.

The solution provider is typically the last line of support, so it is import-

ant to use proper techniques when extending a solution. Not complying
with these rules can void the warranty or support terms of your contract.

Because some apps provide the ability to customize them to meet

user needs, these customizations can also impact the time required to
maintain them. For example, when a new user is onboarded, the time
it takes to onboard and whether existing security roles and apps can
simply be applied or if additional requirements must be added that
require more maintenance time.

As the solution is extended, the importance of having a good testing

strategy grows because it is necessary to validate how the solution
behaves with each update.

Update and future-proofing

In a cloud solution, organizations must embrace the change and inno-
vation that comes with frequent new features and capabilities. These
capabilities, in turn, add value to the business.

When it is necessary to extend functionality, it should always be done

in a way that makes it easy and safe for other developers to extend
them down the line.

Implementation Guide: Success by Design: Guide 15, Extend your solution 407
 It is good practice to take the following perspectives into account. Firstly,
having sound developer craftmanship. This applies to basically any lan-
guage or technology and dictates that when APIs and extension points
are created, they are well thought through, robust, and well defined.
Also, that the extension is made in a way that allows for other extensions
to use the same extension point or API side-by-side with ours.

Secondly, you want to be a “good citizen” in the new solution. That

is to say that by your example, you encourage other developers who
may expand your code later to follow the same principles. This includes
writing code others can easily read, APIs and extension points others
can extend without surprises, unexpected results, or side effects, and
design structures and frameworks that enable others to utilize your
effort in a clean and concise way.

Thirdly, it is important to establish a process that anticipates the future

and works to minimize potential negative effects extending the solu-
tion can bring. This is because, while it is common for organizations
to build custom extensions, deviations from native functionalities can
have an impact on the ability to take advantage of future capabilities.

Product-specific guidance
In the following sections, we will look at Dynamics 365 Customer
Engagement apps and Finance and Operations apps individually.

Finance and Operations apps

In Finance and Finance and Operations apps, the extensibility model
provides the ability to extend the out-of-box functionality to fit addi-
tional requirements.

The extension model itself, the toolset, Application Lifecycle

Management (ALM) in Azure DevOps, and the SDK is well described in
Microsoft Docs, Microsoft Learn, and in multiple community sources.
There are two great entry points for learning about extensions and
development in Finance and Operations apps in the links below and a

Implementation Guide: Success by Design: Guide 15, Extend your solution 408
 list of additional relevant links at the end of the chapter.

Visit the Extensibility home page Introduction

for reference information or follow In this section, we first look at what we can do to enhance functionality
the learning path at Introduction to
developing with Finance and Operations and UI without extending the application. Then we give a very brief
apps - Learn | Microsoft Docs.
overview of the extension model, and finally highlight some of the
tools and practices that are available for professional developers when
they extend the application.

Let’s first look at what we can extend without changing the application
components in Figure 15-4.
Fig.
15-4
Extension example #1 Extension example #2 Extension example #3

Requirement: Create a custom Requirement: Add a table to hold a list of Requirement: Add code to automatically change
workspace with tiles and lists from food allergens. Add the allergen field to the the status of the new production order to “started”
multiple forms and queries across the sales line record for the user to indicate that a when the user firms a planned production order.
manufacturing module. product contains the specific allergen.

Tools and components: Use the Tools and components: Use Visual Studio in a Tools and components: Use Visual Studio in a
personalization feature and “add to developer environment to add the extended developer environment to extend the X++ business
workspace” functionality in the UI to data types, the tables and fields and the logic, add appropriate event handlers, class, methods,
put together the desired workspace extensions to the sales line table and the Sales and X++ code to catch the event that the planned
and components. No change to code Order form. This change requires new and order is firmed, and execute the correct code pattern
or application components is needed. changed application components and must to bring the new production order to started status.
follow software development lifecycle (SDLC), This change requires new and changed application
build, and deployment guidelines. components and X++ code, and must follow SDLC,
build, and deployment guidelines. Considerations
about scalability and performance when large
numbers of planned orders are firmed using
appropriate developer tools should be considered.

This requires user level experience with Requires entry level professional developer Requires medium- to expert-level professional
navigation of the relevant module, experience in Dynamics 365 Finance and developer experience in Dynamics 365 Finance and
navigation, and personalization. Operations apps and familiarity with Visual Operations and familiarity with Visual Studio, build
Studio, building procedures, and best practices. procedure, best practices, and, ideally, frameworks
like SysOperations Framework, Multithreading, and
performance-checking tools and patterns.

Personalization and no-code options

As we mentioned above, even though Finance and Operations apps of-
fer a rich extension model with minimal impact on future updates, it is
always preferred not to extend the product in the first place if possible.
This lowers risk, maintenance, and project complexity.

Before deciding that a certain requirement must be solved through

Implementation Guide: Success by Design: Guide 15, Extend your solution 409
 an extension to the standard product, it is a best practice to consider if
one of the options for noninvasive personalization or configuration can
meet the required functionality. Finance and Operations apps offer wide
options for personalizing the UI, creating user specific experiences, and
automating processes without having to do any programming at all.

Figure 15-5 is a table of these options.

As the table below shows, many options are available to find alter-
native approaches to extending the application. The below is not an
exhaustive list. Additional options for low code and no code options
are mentioned in Chapter 16, “Integrate with other solutions.” The
decision of whether to extend comes down to user efficiency and value
Fig.
15-5 for your customers.

Tools What is it? Description Characteristics

The application remembers the While this is hardly considered a way of extending, it Personalizations can be shared
Restricted last settings for sorting, column does give the user the opportunity to store selections in by users or managed centrally
width, criterial values in queries dialogs, expansion of relevant fast tables, and aligning by an admin from the
personalization
and dialogs. the column width so that more columns are visible on a personalization page.
given form.

Personalization bar in forms and Personalization allows users or admins to add or hide Personalizations can be shared
Personalization workspaces. fields or sections, change labels, change the order by users or managed centrally
of forms/UI of columns, and edit the tab order by skipping fields by an admin from the
when pressing tab. personalization page.

Saved views is a combination of Saved views is a powerful tool that allows the user Saved views can be shared by
personalization of the UI for a the ability to quickly switch between tailored views of users or managed centrally by an
form and filtering and sorting of columns, filtering and sorting on the same screen admin from the
the form data source. depending on the specific task at hand. For example, Personalization page.
Saved views
a buyer in a pharmaceutical company may need a
simple view of the purchase order screen for May require the Saved views
nonregulated materials purchasing regulated materials feature to be turned on.
used for manufacturing

Users can use personalization This functionality allows users to tailor their experience Custom and personalized
and the “add to workspace” to their needs. Workspaces provide glanceable workspaces can be shared by
Custom button to add tiles, views, or information about the most important measures, users or managed centrally by
Workspaces links to an existing or custom actionable items, and relevant links to other pages. an admin from the
workspace. personalization page.

Users with certain security roles Finance and Operations apps have a rich set of Custom fields and the
can add up to 20 custom fields features that apply across a wide set of industries. personalized forms to show the
to tables Some organizations require additional fields on fields can be shared by users or
certain tables; for example, the item or customer managed centrally by an admin
Custom fields
master or the sales order header. This feature allows from the personalization page.
the user to create these fields. Note that these fields Deleting a custom field is
are specific to the environment that they are created irreversible and results in the loss
in and cannot be referenced by developer tools. of the data in the custom column.

Implementation Guide: Success by Design: Guide 15, Extend your solution 410
 Fig.
(continued)
15-5

Tools What is it? Description Characteristics

The grid on forms in the system The grid offers the following capabilities. The admin can enable the
have some extended features New Grid Control feature from
▪ Calculating totals for columns in the grid footer
that may eliminate the need for feature management. Note that
an extension. ▪ Pasting from Excel there are certain limitations, see
▪ Calculating math expressions. For example, the reference at the bottom of
if the user enters 3*8 in a numeric field and the section.
Grid capabilities presses tab, the system calculates and enters the
result of 24

▪ Grouping tabular data in one or more levels in

expandable groups

▪ Freeze panes so important information does not

scroll out of view when scrolling horizontally

The user can add a canvas app to The ability to embed a canvas power app enables Please see more about UI
a form or workspace as embedded citizen developers to use low-code/no-code options integration for Finance and
into the UI or as a menu item for interacting with data in Dataverse directly from Operations apps in Chapter 16,
Embedded that can pull up the app from the the UI in the Finance and Operations apps. “Integrate with other solutions.”
power apps menu.
canvas apps
It is important to note that if the app must interact
with Finance and Operations apps data, that
integration to Dataverse must be in place and the app
must of course support the required actions.

Users can view, edit, and act on IT admins can build and publish mobile workspaces that Simple actions can be done from
business data, even if they have have been tailored to their organization. The app uses the mobile workspaces. Most
intermittent network connectivity existing code assets. IT admins can easily design mobile more advanced actions require
Mobile on an app for iPhone and workspaces by using the point-and-click workspace extension.
workspaces Android. designer that is included with the web client.

The application has 13 predefined workspaces out-of-

the-box that cover Finance, AR, AP, and HR functions.

Users can extract or edit data The Excel integration allows for a wide variety of With great power comes great
on most forms in the system by scenarios for entering, presenting, and interacting with responsibility. While it is easy to
clicking on the office icon. data in way that is not possible from the UI. In addition change a whole column of data
Excel to the export-to and open-in Excel option, the user can and publish that data into the
integration create workbooks and templates for specific purposes. system, it is equally easy to make
Excel has many features for presentation and offers a mistake.
data manipulation capabilities for larger datasets that
users cannot do in the system UI.

Extending the application

In this section, we provide a high-level introduction to the extension
model and discuss required skill sets and practices used when extending.
We also highlight some of the tools and techniques that we recommend
you use to ensure a clean, high-performing, and future-proof solution.

Implementation Guide: Success by Design: Guide 15, Extend your solution 411
 Figure 15-6 is an overview of the components, the consideration for
Fig. extensions, and the related characteristics.
15-6

Component Editor Considerations Do Not

Graphical editor in Visual Studio Forms must adhere to patterns. The editor in Visual Deviate from the predefined
with preview. Studio has a rich modeler that will help the developer patterns and create monster all-
applying the right structure to the screens. This in-one screen style forms. They
User interface
will ensure performance when loading the form, are often a result of the designer
adaptability across different resolutions and screen trying to replicate a legacy
sizes and consistency across the UI. system experience.

Metadata editor in Visual Studio Follow best practices and frameworks. Apply indexes Create redundancy or replicate
Data model for Tables, Fields, Extended Data and define delete actions. Normalize. Use effective poor modeling from legacy apps.
and queries Types, Enums Queries and Views date framework when applicable. Keep performance
in mind. Use field lists when performance is a risk.

X++ Editor in Visual Studio. Adjust compiler setting to alert about best practices, Ignore best practices, write long
code with the goal of zero deviations. Use code patterns methods or over comment
and frameworks. Practice good developer the code.
Business logic
citizenship. Write clean easily readable code.
Run CAR report and use compatibility checker.
Unit test the code.

SSRS report editor in Visual SSRS reports are good for certain precision Reach for the SSRS report option
Studio designs and tabular lists. See Chapter 13, if there is a better way.
Reporting
“Business intelligence, reporting, and analytics,”
for more information.

Metadata editor in Visual Studio The out-of-the-box data entities are general purpose Create data source proliferation.
entities are built to support a wide variety of features See Chapter 16, “Integrate
surrounding business entity. In scenarios where a with other solutions,” for more
high volume, low latency interface is required it is information.
Data entities recommended to build custom data entities with
targeted and specific features needed to support high
volume interfaces in the implementation.

Development architecture
The architecture of the development environment, as shown in Figure
15-7, includes the software development kit (SDK), which consists of
Visual Studio development tools and other components. Source con-
trol through Azure DevOps allows multi-developer scenarios, where
each developer uses a separate development environment. Deployable
packages are compiled and packaged in a build environment or a build
service and uploaded to Dynamics Lifecycle Services (LCS) for deploy-
ment to nonproduction environments. Deployment to the production
environment happens after proper testing has been done in the UAT

Implementation Guide: Success by Design: Guide 15, Extend your solution 412
 environment and users/stakeholders have signed off as described in
the SDLC.

See Chapter 11, “Application lifecycle management,” for more informa-

tion about the SDLC.

Fig.
15-7

Visual Studio Debug/Run (F5)

Local Runtime

Batch
Storage
Project system Cloud manager
service
X++ code editor instance service

Designers
Application explorer
Design-time meta-model Deploy
Application object server Service
Best practice integration hosted by Internet information service

Package
(metadata, binaries, and data)

Business
database

Build
Model binaries Runtime
(file system)

Model store
Metadata API
(file system)

Development environment

Cross app integration

Cross app integration options such as dual-write, Virtual Tables, and
configurable app-to-app integrations allow developers extending
Finance and Operations apps to utilize all the capabilities of Dataverse
and Microsoft Power Platform to make extensions. An example
could be a requirement for a lightweight product catalog with crisp,
high-resolution images in Finance and Operations apps. This extension
could be made by integrating to Dataverse and embedding a canvas
app into the Finance and Operations apps UI to show the catalog.

Implementation Guide: Success by Design: Guide 15, Extend your solution 413
 By bringing data from Finance and Operations apps into Dataverse, it
is possible to leverage the Power Platform and use low-code/no-code
developer models on Finance and Operations apps data. Either as embed-
ded power apps or PowerBI reports or dashboards, in standalone apps,
or integrated with other Dynamics 365 apps, without making changes
through extensions to the Finance and Operations app and data layer.

Please see more details about integration to Dataverse in Chapter 16

“Integrating to other solutions”.

Advanced practices and tools

As we mentioned earlier, it is always sound advice to be good citizen
developers. In Finance and Operations apps specifically, you can
ensure that you follow that principle by understanding, learning, and
using the following tools and best practices as shown in Figure 15-8.

Fig.
15-8

Tools What is it? Description Characteristics

Best practice checks for X++ and Developers should strive for zero deviations. The You can stop developers from
Best practice application components are built best practices are made to ensure and updatable, checking in code with best
check into Visual Studio. They can be performing and user friendly solution. practice deviations.
errors, warnings, or informational.

The compatibility checker tool The compatibility checker tool is available as one Not all breaking changes can
Compatibility can detect metadata breaking of the dev tools. You can use it to ensure that your be detected by the tool. See the
report changes against a specified solutions are backward-compatible with earlier Compatibility checker docs page
baseline release or update. releases before you install or push updates. for specifics.

The users can take a trace of You can use the trace parser to consume traces and The trace parser tool can be
runtime execution directly from analyze performance in your deployment. The trace found in the PerfSDK folder in
Traces and
the UI. The trace parser can be parser can find and diagnose various types of errors. your development environments.
trace parser used to read the trace. You can also use the tool to visualize execution of X++
methods, as well as the execution call tree.

Performance timer is a tool in To open the Performance timer, open your web page The tool itself has a performance
the web client that can help you with the added parameter debug=develop. You can impact.
to determine why your system's see counters for client time and server time, and the
Performance performance might act slow. total time. Additionally, you can see a set of performance
timer counters, a list of expensive server calls, how many
SQL queries were triggered by this individual call and
which SQL query was the most expensive.

Implementation Guide: Success by Design: Guide 15, Extend your solution 414
 Fig.
(continued)
15-8

Tools What is it? Description Characteristics

Under Environment Monitoring The logs provide for example: The tools under Environment
in LCS there a comprehensive Monitoring are very effective at
▪ Activity Monitoring: A visualization of the activity
collection of tools and information diagnosing potential or growing
that has happened in the environment for given
LCS Logs that you can use to analyze and
timeline in terms of user load, interaction and
performance issues. Keeping
diagnose your cloud environment. an eye on these metrics can
activity.
help pinpoint problems with
▪ Information about slow queries, deadlocks, etc. extensions.

The CAR report is an advanced The CAR report can be run by command line in a A clean CAR report is a
Customization best practice check tool. development environment. The output is an Excel requirement for the go-live
Analysis Report workbook with recommendations issues readiness review prior to
(CAR Report) and warnings. enabling production.

A breaking change is a change Breaking changes are, for example, changes to data That although the application is
that will break all or some of the model and extended data types, changes to access massive, we tend to only extend
consumers of your code that modifiers on classes and methods and many others. the same relatively small subset
already exist. of elements. It is not as unlikely
Understand and
This is especially important in heavily extended solutions, that you may think that other
avoid breaking
if you are making a basis for a global rollout, if you developers use your components
changes
are making an ISV, or if you have multiple developers or code.
sharing common custom APIs or constructs but should
always be considered.

If you find a need for an Extensibility requests are logged to a backlog. Extensibility requests are
extension point that is currently Microsoft engineers prioritize all requests, and then following the same cadence as
not available, log the request work on them in priority order. Please note that the platform updates.
Log extensibility
early. This is done via an Microsoft is not guaranteeing that all requests will be
requests early extensibility request in LCS. fulfilled. Requests that are intrusive by nature will not
be supported, as they will prevent a seamless upgrade.

Sometimes developers put a Developers are the first line of defense against bugs, It is a lot easier and more cost
lot of effort into building the performance issues and semantic issues that may effective for the developer to
extension, but little effort into exist in the specification. By simply going through the find and fix a problem before it is
unit testing it before determining intended functionality from perspectives such as: checked in, built and deployed.
whether it is ready to deliver. ▪ Will this code scale with high volumes?
Proper unit
▪ Does it do what I expect?
testing ▪ Can I do things I am not supposed to?
▪ Could the user accidently do something unwanted?
▪ Does the requirement make sense or force me to
break best practices or patterns?

For more on these topics, see the “Reference links” section later in
this chapter.

Additional components
It is important to note that the extended product architecture contains
several additional components. Along with that are multiple avenues

Implementation Guide: Success by Design: Guide 15, Extend your solution 415
 and tiers for approaching a requirement for extension. It depends on
the specific area and nature of the requirement.

In addition to the components in the Finance and Operations applica-

tion stack, there are components for Commerce and WMS that require
different technologies as well as developer toolsets and skillsets. These
include the Point of Sale (POS), The Commerce Scale Unit (CSU), the
eCommerce components, the Warehousing Mobile App, and the
Finance and Operations Mobile App.

That means separate requirements for extensions may have to be

applied to different components. Slightly different approaches may
also be needed, and the complexity and skill sets required may vary.
For all of these, it is critical that developers and administrators familiar-
ize themselves with each of the specific SDKs, guidelines, application
lifecycles, builds, and deployment methods and the skill sets required
before deciding to extend the component.

Reference links
▪ Extensibility home page - Finance & Operations
▪ Application stack and server architecture - Finance & Operations
▪ Microsoft Power Platform integration with Finance and Operations
▪ Microsoft AppSource – destination for business apps
▪ Commerce for IT pros and developers - Commerce
▪ Write extensible code - Finance & Operations
▪ Breaking changes - Finance & Operations
▪ Extensibility requests - Finance & Operations
▪ Grid capabilities - Finance & Operations
▪ Mobile app home page - Finance & Operations
▪ Take traces by using Trace parser - Finance & Operations
▪ Testing and validations - Finance & Operations

Implementation Guide: Success by Design: Guide 15, Extend your solution 416
 Customer
Engagement apps
The Power Platform, shown in Figure 15-9, provides the ability to use
configurations, low-code/no-code customizations, and still allows de-
velopers to programmatically extend first-party customer engagement
apps like Dynamics 365 Sales. Custom business applications can also be
created. You can even mix the two approaches to provide applications
adjusted to specific organizational needs.

Fig.
15-9
Microsoft Power Platform
The low-code platform that spans Office 365, Azure, Dynamics 365, and standalone applications

Power BI Power Apps Power Automate Power Virtual Agent

Business analytics Application development Process automation Intelligent virtual agents

Data connectors AI builder Dataverse

Until recently, creating custom apps would require professional de-

velopers. But by using Power Apps, organizations can quickly build
custom business applications that provide rich business logic and
workflow capabilities to transform manual business processes into
digital, automated processes.

Implementation Guide: Success by Design: Guide 15, Extend your solution 417
 Power Apps democratizes custom business app building experience
by enabling non-professional developers to build feature rich, custom
business apps without writing code.

When offering this ability to users, it is important to have a central

governance entity to determine guidelines for certain aspects of the
solution, like security, data model, etc. This prevents redundancy, gaps
in the security model, and bad practices/architecture that impact how
apps are maintained, secured, scaled, and shipped.

The Microsoft Power Platform Center of Excellence (CoE) starter kit

provides a collection of components and tools that are designed to
help organizations develop a strategy for adopting and supporting
Microsoft Power Platform. This helps organizations implement a CoE
to drive innovations and improvement while also providing standards,
consistency, and governance.

Power Apps also provide an extensible platform that lets professional

developers programmatically interact with data and metadata, apply
business logic, create custom connectors, and integrate with external data.

Model the user experience

Power Apps is a high-productivity development platform for business
apps and has four major components:

Canvas apps are intended to build a business app from a canvas in

Microsoft Power Apps without requiring professional developers to
create code. The apps can be built by using drag and drop elements
onto a canvas. It also provides the ability to use connectors to integrate
data from, and to, a variety of sources.

Model-driven apps are a component-focused approach to app devel-

opment, which also does not require code. These apps can be simple
or complex. With Model-driven apps much of the layout is determined
and designated by the components that are added to the app. The
Model-driven apps can be completely custom or from first-party apps
such as Dynamics 365 Sales.

Implementation Guide: Success by Design: Guide 15, Extend your solution 418
 Both types of apps are built to be responsive by adjusting and being
accessible from different devices.

Cross app integration open opportunities

Power Apps makers can also create external facing Portals that allow
to extend solutions on Dataverse with data
from Finance and Operations Apps. users outside their organization to create and view data in Microsoft

Virtual entities offer CRUD integration

Dataverse. The Portals experience reuses components in other apps to
to data in Finance and Operations from determine the experience. It is also possible to customize the Portals
Customer engagement apps without the
need for extensive integration work. experience, similar to how other apps are customizable.

Dual-write is available for scenarios where

the business process requires the data to Microsoft Dataverse securely stores and manages data that is used by
be physically copied between Finance and
business applications.
Operations and Dataverse.

You can read more about Cross app

integration in Chapter 16 “Integrating to
other solutions”. Business logic and validation
within the platform
Business logic and validation to ensure data quality and reduce re-
petitive tasks can be applied within the Power Platform by using the
following approaches:
▪ Business rules validate data insertion and provide validation mes-
sages within the user interface. As an example, with business rules
it is possible to display a message when a certain field does not
have data or does not have data in the expected format.
▪ Business process flows guide users to ensure they enter data and
use the solution in a consistent way. An example would be dis-
playing the information that is required to complete a business
process, like lead to invoice, this provides a consistent approach
for all users while it also improves usability of the solution by
making the process more efficient and improving data quality.
▪ Workflow allows business processes to be automated within the
process designer to apply a condition or perform a new action.
For example, when closing a support case to send a survey to the
customer.
▪ Business logic with code supports advances developer scenarios
to extend the application with code, for example using java script
code for form events or plug-ins that apply business logic to data
transaction.

Dynamics 365 Customer Engagement apps, such as Dynamics 365 Sales

Implementation Guide: Success by Design: Guide 15, Extend your solution 419
or Dynamics 365 Customer Service, use Dataverse to store and secure
the data they use. This enables organizations to build or extend apps by
using Power Apps and Dataverse directly against the business data.

Dataverse data is structured according to the Common Data Model,

which is a shared data language for business and analytical applications
to use. By using Dataverse, makers can simply extend the data model
of first-party apps such as Sales, Customer Service, Field Service or
others, and jump-start app development by using the Common Data
Model with business logic, security, and integration already built in.

Customer Engagement apps and the Power Platform provide a powerful

customization and app making capabilities by using several components
that provide the ability to build the appearance and functionality of an
app. These components are distributed in four categories.

Data
The entity designer and option set designer determine what data the
app is based on and allow changes to the data model by adding ad-
ditional tables and fields as well as relationships and components that
use predetermined options for users to select.

User interface (UI)

User interface components, like the app designer, site map designer,
form designer, and view designer, determine how users interacts with
the app and allow changes to be made to the components that display
in the app UI.

Logic
The business process flow designer, workflow designer, process designer,
and business rule designer, determine the business processes, rules, and
automation of the app.

420
 Visualizations
These determine what type of data visualization and reporting the
app includes, like charts, dashboards, and reports based on SQL Server
Reporting Services.

You can create a range of apps with Power Apps, either by using canvas
or model drive apps to solve business problems and infuse digital
transformation into manual and outdated processes.

Solution analyzers
Solution checker
The solution checker can perform a rich static analysis of the solutions
against a set of best practice rules to quickly identify patterns that may
cause problems. After the check is completed, a detailed report is gen-
erated that lists the issues identified, the components and code affected,
and links to documentation that describe how to resolve each issue.

It is important to include solution checker with the solution release

cycle, and after releasing an updated version in a lower environment
(sandbox), to make sure that any deprecations or issues related to
extensions are identified.

Portal checker
The portal checker is a self-service diagnostic tool that can identify
common issues with a portal by looking at various configuration
parameters and providing suggestions on how to fix them.

Power Apps checker web API

The Power Apps checker web API provides a mechanism to run static
analysis checks against customizations and extensions to the Microsoft
Dataverse platform. It is available for makers and developers to per-
form rich static analysis checks on their solutions against a set of best
practice rules to quickly identify problematic patterns.

Implementation Guide: Success by Design: Guide 15, Extend your solution 421
 Conclusion
Cloud-based solutions offer ready-to-use applications solutions that
can be easily delivered, reducing the time required for an organization
to start taking advantage of them. SaaS solutions typically reside in
cloud environments that are scalable and offer native integrations with
other SaaS offerings. These solutions benefit from continuous updates
that add innovation and new capabilities multiple times per year. This
reduces costs and effort and eliminates the downtime associated with
upgrades in a traditional on-premises model.

This means that organizations can start using the solution as-is. Still,
in some scenarios, additional requirements are needed to add value
to the business or to empower users and drive adoption of the new
solution. Thus, modern SaaS solutions also provide rich capabilities to
further extend it. These can range from simple configurations using a
low-code/no-code approach or an extension by using custom code by
professional developers.

It is important to consider that the level and depth of these extensions

can impact key characteristics, like performance, or increase the cost
of supporting and maintaining the solution, but they should hinder
the ability to take advantage of innovative functionalities that are
delivered as part of the continuous update strategy. Normally, when
deciding to extend a solution, a trade-off is necessary between new
functionalities and the impact of adding them. For example, if adding
too much information to a form it impacts the time required to load
the information included in it.

In this chapter we have discussed how these key characteristics can be

impacted and the consequences of overextending the solution.

With Success by Design, during the Solution Blueprint Review work-

shop, such topics are discussed to establish a clear understanding
of the impact that the extensions have on the solution, as well as to
identify risks and issues that come with these decisions. One of the
most important factors to consider when deciding to extend a solution

Implementation Guide: Success by Design: Guide 15, Extend your solution 422
References is how performance is affected. During the Solution Performance im-
plementation workshop, the FastTrack solution architect works along
Business Apps | Microsoft Power Apps with the implementation partner and the customer to review how the
extensions can impact the overall performance of the solution. Typical
Build Apps – Canvas Apps or Model-driven
Apps | Microsoft Power Apps risks and issues are identified related to overextending forms, synchro-
nous events that impact the user experience, impact on capacity such
Microsoft AppSource – destination for
business apps as service protection limits, resources, and allocation limits.

423
 Checklist

Define your Considerations

extensibility strategy Validate that the extensions don’t negatively impact
the user experience, responsiveness, performance, and
Check that all requirements to extend the solution are
how the solution behaves across different devices and
driven by the need to empower users or bring additional
platforms as part of the organization test strategy.
value to the business.
Ensure the extensions honor the security mechanism,
Ensure the solution doesn’t mimic the ways of achieving
privacy, and compliance requirements.
the same results as the legacy solution or the system
being replaced. Ensure extensions are scalable, tested for high volume,
and capable of handling peaks like holiday seasons.
Understand the platform capabilities and use its
strength to simplify and optimize the overall process to Align extensions with ALM automated processes to build
get most of the out-of-the-box experiences. and deploy them in an efficient and fast-paced manner.

Review if any potential ISVs were considered before Ensure code and customizations follow only the
deciding to extend the solution. The AppSource market- documented supported techniques, and don’t use
place contains ISV-managed solutions that may replace deprecated features and techniques.
the need to create a custom solution.

Implementation Guide: Success by Design: Guide 15, Extend your solution 424
 Case study
The power of making it your own
A wealth management services company that delivers personalized
services had the need to standardize, simplify processes, drive efficien-
cies, expose more data, and increase visibility and interoperability to
their users.

After evaluating solutions, the company decided to build a modern,

customized, cloud-based Customer Engagement apps platform on
Microsoft Dynamics 365.

The off-the-shelf functionalities gave the ability for the users to achieve
a single view of the customer lifecycle across all products. Through
customizations and specific extensions of the solution, the company
added special wealth management services and customized the user
experience to better serve their customers.

This helped the company streamline processes, increase revenue,

reduce costs, enhance transparency, and manage regulatory com-
pliance, and allowed the users to better understand each customer’s
story, goals, and capacity for risk. The company was able to deliver
appropriate, innovative, high-value services and products to maintain
successful end-to-end customer journeys across every account.

The power of using off-the-shelf functionalities and being able to

further customize and extend the experience for their users set the
foundation to improve their internal processes and empowered their
users to deliver better services in a more efficient matter.

Implementation Guide: Success by Design: Guide 15, Extend your solution 425
16 Guide
Integrate
with other
solutions
 Introduction
Together, Dynamics 365 apps provide a rich and
mature feature set across a wide variety of industries.
However, there are situations in which you might want to go beyond the
boundaries of the application suite and extend processes or incorporate
data assets from other sources. In this chapter, we examine how to
integrate Dynamics 365 apps with other solutions, or with each other.

We discuss the importance of keeping business goals at the forefront

while aligning integration projects with your organization’s overall
solution strategy.

In line with the design pillars, we look at how the choice of integration
This chapter covers the platform should fit into the architectural landscape.
following topics:
Defining business goals
We also look at how to choose an integration design that offers users
Choosing a platform the capabilities they desire in the short and long term.
Choosing a design

We discuss integration patterns available with Dynamics 365 apps and

Choosing a pattern
consider important characteristics and factors.
Challenges in integration

Product-specific guidance And finally, before diving into the product specifics, we walk through
some of the common challenges people face when integrating systems.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 427
 Defining
business goals Defining business goals
Choosing a platform
To ensure that your integration work aligns with the overall direction of
Choosing a design the business, it’s important to match each requirement of cross-system
Choosing a pattern processes against the overall goals of the project and the business. To
accomplish this, begin your integration work by defining goals that
Challenges in integration
map to the business perspective.
Product-specific guidance

Reasons for integrating

You might have many reasons why you want to integrate systems in
your solution. Given the array of tools available and the vast ecosystem
provided by Microsoft Power Platform and Azure, integration across
processes and systems is often readily available and a scalable and
reliable choice. Here are some examples of when you might want to
unify systems by using integration.

Integration to legacy system

You might have a highly specialized system, perhaps a custom-built
or heavily customized system that fits a specific set of requirements in
your business, that you want to keep in place or phase out over time.
With this approach, you would integrate Dynamics 365 apps into the
legacy system to get the benefits of both and allow for business pro-
cesses to span both systems.

Multi-phased implementation
You might be implementing Dynamics 365 apps in a planned multi-
phased approach, starting with a geographic location, a single division,
or a single Dynamics 365 app; Finance, for example. In this scenario,
some level of integration with the parts of the legacy solution that will
be implemented in future phases could be necessary.

Regulatory requirements
Regulatory, government, and industry data exchange or reporting
are often standard practice or required by law in some industries and
regions. It’s often reported electronically, but more complex exchanges
of data could require a more traditional integration approach.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 428
 Financial consolidation
Perhaps your organization is a subsidiary of a larger operation that
requires data to be consolidated and reported in a corporate parent
entity. This often requires extracts of data to be transformed and
loaded into the corporate consolidation system. In some cases, it’s the
other way around: your organization might expect integration of con-
solidation and other data from your subsidiaries into your new system.

Multiple system architecture

Sometimes the Dynamics 365 apps are links in a bigger chain of spe-
cialized line-of-business systems. For example, the Prospect-to-cash
process might start in one external system where the opportunity
is discovered, then go through quote and order management in
Dynamics 365 Sales app, and then integrate to other systems for war-
ranty tracking, billing, and financial reporting.

Extract to reporting solutions

Perhaps you might have invested in building a specialized analytics
solution that includes a data warehouse with many years of historical
data and tailored reporting on top of it. In that scenario, you might
want to extract data from Dynamics 365 apps for transformation and
load it into the existing data warehouse.

Many more scenarios are not described here. Some might even be
combinations of several of the examples.

When multiple systems “talk” to each other, systems that perform

distinct functions in the business, it’s not only about making it work.
Integration can enable business processes, reporting, and insights that
would not have been possible otherwise.

Goals and benefits

Organizations that integrate systems when building their business
solution can achieve several key goals and benefits, including:
▪ Process automation  Integration enables the automation of
repetitive tasks in the workflow.
▪ Data availability at the right time  Communication between

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 429
 different departments is key. If the sales department depends on
information from accounting, integration can help the data to flow
at the time that the sales department needs it. Whether the data
arrives in real time or in batches, you can program it to arrive at a
certain time. The right pattern of communication will help satisfy
the different business units.
▪ Reduced human errors  An error in the data like a typo, an
incorrect number, or a button pressed at the wrong time could
significantly affect your process, or even worse, degrade the cus-
tomer experience.
▪ Increased productivity  The users become more focused on the
processes that add value, since they won’t need to reach for or
search other systems to retrieve the information they need.
▪ Process optimization  Integration simplifies your processes so
that you spend less time executing tasks, thereby adding more
value to your business.
▪ Increased security  By automating processes and data move-
ment through integrations, organizations implement better
controls for data access and reduce the need for users to directly
work with sensitive data.
▪ Regulatory compliance  Automated integrations could help
meet some of the controls needed to meet regulatory requirements
like HIPAA (Health Insurance Portability and Accountability Act) or
FDA Regulations in the US.
▪ Reduced cost of operations  Integrations might help reduce
repetitive manual activities, human errors, and training activities,
which could lead to an overall decrease in cost of operations.
▪ Avoiding data silos  Integrations break down data silos and
For supplemental information, read the
Integration design for Dynamics 365
improve the value of the organization’s data. With AI, machine
solutions on Microsoft Learn. learning, and IoT usage on the rise, this data might help drive bet-
ter insights, data science, and predictive analytics in your company.

Integration planning
To properly incorporate business goals into the entire project lifecycle,
we recommend you plan the integration architecture in the initial
stages of the implementation project.

When planning integrations, it’s easy to miss a critical point or

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 430
 underestimate volume or complexity. To prevent this, we highly
recommend you create a blueprint of the design before starting any
specification work and do some calculations of loads and transaction flow.
Success by Design highly recommends
that you approach integration work the
same way as an extension project, by Conceptualizing
following a defined software development
lifecycle (SDLC) that incorporates business
Creating a blueprint and thoroughly planning the implementation
stakeholders and collects their buy-in. activities will also help you formulate the testing and performance test-
The SDLC should include requirements,
design, development, and testing, as well ing of the solution later in the implementation. That’s why integration
as performance testing, deployment, and architecture is a key part of the Success by Design solution blueprint.
application lifecycle management (ALM).
The work-to-define requirements should
be business driven and process focused.
To create a blueprint for integration, you can leverage several types of
For more information, refer to Chapter 7, diagrams, which we describe here.
“Process-focused solution,” and Chapter 11,
“Application lifecycle management.”
▪ Application integration diagram  The application integration
diagram is often a high-level representation of solution architec-
ture. An example of one is shown in Figure 16-1. Many styles exist,
but in its basic form it should provide an overview of which sys-
tems in the solution need to be integrated and, ideally, what data
is exchanged and the direction of the data flow. Once the overview
is established, details can be added about the specific interface
touchpoints, frequency, and volume information, and perhaps a
link to ALM information, such as a functional design document (or
Fig.
16-1
FDD) number or link.

Customer Finance and

Engagement apps Operations apps

System 1 Data App 1 Data App 3

App 2 App 4

▪ Interaction diagram  The interaction diagram defines system

and user trigger points, data and process flows, and the sequence
of events. It can also uncover unforeseen complexity. By following
the flow, even nontechnical team members can understand and
validate the design before any work has started. The diagram in
Figure 16-2 illustrates a simple three-system interaction diagram
for a simulated process flow.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 431
 Fig.
16-2

System 1 System 2 System 3

Process start

User trigger event System event

Message Request

Message Response
User action User trigger event

Fig.
16-3
▪ Process documentation, mapping, and volume calcu-
lations  Once the high-level overviews are defined, we
System 1 System 2
Sales Operations recommend you document the cross-systems processes in more
detail, for example by defining user stories, before starting the
AccountNumber Customer Account
system-to-system mapping. Mapping can be done as part of a
Name OrganizationName functional or technical design document. (An example of map-
Description SalesMemo ping is shown in Figure 16-3.) The design document, along with
the process definition, should be detailed enough to ensure that
Telephone1 PrimaryContactPhone

everyone agrees on the scope, process, desired outcome, test

EmailAddress1 CustomerGroupID
criteria, and benefits of the integration. Often such a document
CreditLimit PrimaryContactEmail
also contains information about the expected effort and duration.
CreditLimit Calculating the volume of transactions that flow through the
interface is important because it helps you decide what patterns to
use, and ultimately the size of the different platform components
and services needed for the integration.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 432
 Defining business goals
Choosing a platform
Choosing
a platform In this section we discuss what you should consider when choosing a
Choosing a design platform and, if needed, a middleware solution to achieve the previ-
ously mentioned business goals.
Choosing a pattern

Challenges in integration Data storage needs have increased exponentially in recent years, and
it will continue to intensify in this direction. The amount of data that’s
Product-specific guidance
generated and transferred today is practically unimaginable. In addi-
tion to the new data that’s being captured or generated, historical data
from legacy systems might need to be preserved. The platform you
choose will need to incorporate all of these realities, and must be able
to reliably handle the storage and transfer of vast amounts of data.

Cloud, on-premises, and hybrid platforms

Your organization’s size and maturity level are factors when consid-
ering various platform scenarios. Most organizations find themselves
using some variation of one of the following types:
▪ Simple cloud  In the scenario depicted on the left side of Figure
16-4, the organization might have acquired the first cloud soft-
ware as a service (SaaS) application. Overall solution complexity is
low. As part of the implementation project, they want to integrate
with another cloud service.
▪ Simple hybrid  The scenario on the right side of Figure 16-4 is
similar to the previous one in that the organization might have
acquired the first cloud SaaS application. Solution complexity is
also low. But in its implementation project, the organization wants
to integrate to a system on-premises, adding the need for a strate-
gy for crossing the cloud to on-premises boundary.

When implementing a simple cloud or simple hybrid system, be sure to

use the capabilities that Microsoft Azure and Power Platform provide.
With Dynamics 365 apps, out-of-the-box low-code/no-code options
are available for using Pull designs and RESTful service endpoints. Such
an approach will build a scalable foundation, allowing expansion for
years to come.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 433
 Fig.
Simple cloud Simple hybrid
16-4

System 1
System 1 System 2

Cloud
Cloud
On-premises

On-premises
System 2

Whether in cloud or hybrid, be sure to not fall for the temptation of

building tightly coupled point-to-point, custom synchronous service
integrations. They will likely not scale well, and future expansion will
likely make the system more fragile. It’s good practice to lay a solid
foundation early. Organizations that choose not to do that often find
themselves having to make tough and expensive adjustments to make
the architecture scalable as the business grows.
▪ Primarily on-premises  In the scenario depicted on the left side
of Figure 16-5, the organization has in place a rather complex
on-premises solution with multiple line-of-business systems.
The organization is using a middleware platform that’s based
on-premises to connect systems and might be integrating its first
SaaS cloud application.
▪ Hybrid  The hybrid example on the right side of Figure 16-5
represents large, diversified enterprises or organizations that are
halfway through their cloud journey. The organization is using
connected cloud and on-premises middleware platforms that can
integrate in hybrid scenarios, and it might add an additional cloud
application to the solution architecture.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 434
 Fig.
16-5 Primarily on-premises Hybrid

System 1 System 2 System 3 Storage/DW

Cloud

Cloud
System 5

Middle-
ware

Middle- Middle-
ware ware
On-premises

On-premises
System 1 System 2 System 3 System 4 System 4 System 5 System 6 Storage/DW

When planning an incremental migration to the cloud, consider the

best options for where and how to achieve a reliable hybrid middle-
ware-to-cloud platform. Consider whether the current middleware
platforms and on-premises components are scalable and provide
opportunities for leveraging modern technologies, such as low-code/
no-code development options, or whether it would make sense to
leverage the possibilities within Azure and Power Platform to migrate
the current line-of-business systems to the cloud to achieve scalability,
elasticity, and extendibility.

Be sure to not silo data assets in on-premises line-of-business systems

and disjointed data stores that are difficult for citizen developers,
makers, and modern frameworks to access.
▪ Hybrid without middleware  In the scenario shown on the
left side of Figure 16-6, the organization has likely evolved the
integration architecture over time. Multiple systems are integrated

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 435
 using different technologies, and it seems System 4 is acting as
a hub. No middleware platform is applied, and strategy is likely
somewhat loosely defined. The perceived cost of applying an
integration platform and middleware is making it necessary to use
point-to-point integration when new systems are introduced.
When implementing a system that’s
▪ Primarily cloud  In the scenario shown on the right side of
primarily cloud, compare the cost of Figure 16-6, the organization has a higher level of cloud maturity,
implementing a clear strategy and
using a platform and middleware such platform, and middleware are platform-as-a-service (PaaS) cloud
as Azure and Power Platform with the
services. The organization has integrated a few on-premises
cost of operating and maintaining the
existing integrations. components that are necessary, and additional line-of-business
systems are all in the cloud.

Primarily cloud
Fig.
16-6 Hybrid, no middleware

System 1 System 2 System 3 Storage/DW

System 1 System 2 System 3 Storage/DW
Cloud

Cloud
Middle-
ware
On-premises

On-premises
System 4
System 4 System 5 System 6 Storage/DW

It’s important to not continue using resources on maintaining an

integration architecture that spans multiple technologies and doesn’t
have a centralized mechanism for error handling, notification, high
availability (HA), and disaster recovery (DR).

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 436
 Be sure to continue using the cloud-based strategy and the toolset
provided by Azure and Power Platform to integrate to on-premises
components when necessary.

Middleware
Integration middleware is software or services that enable communi-
Middleware provides specialized cation and data management for distributed applications. Middleware
capabilities to enable communication,
transformation, connectivity,
often provides messaging services on technologies such as SOAP, REST,
orchestration, and other messaging- and JSON. Some middleware offers queues, transaction management,
related functionality. Dynamics 365
Finance and Operations are business monitoring, and error logging and handling. Different middleware
apps that provide integration capabilities
platforms can support on-premises, cloud-based, or hybrid scenarios.
to support interfaces, but they are not
designed to replace middleware. The following are characteristics you should consider when choosing a
middleware platform.

Key characteristics
When deciding whether to integrate to an existing system, you should
consider several important characteristics of the solution, the system,
and its context.
▪ Scalability and performance  The planned platform, middle-
ware, and supporting architecture must be able to handle your
organization’s expected persistent and peak transaction volumes
in the present, the short term, and scale well in the long term.
▪ Security  Authentication defines how each system confirms a
user’s identity, and you should consider how that will work across
systems and with middleware. Authorization specifies how each
system grants or denies access to endpoints, business logic, and
data. It’s important to ensure that an integration platform and
middleware are compatible with system security and fit into the
overall security architecture.
▪ Reliable messaging  Middleware typically provides messaging
services. It’s important that the messaging platform supports the
architecture of the integrated systems and provides a reliable
mechanism or technology to ensure that messaging across sys-
tems is accurately sent, acknowledged, received, and confirmed.
This is especially important in situations in which a system or
part of the supporting architecture is unavailable. Error-handling

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 437
 concepts related to messaging, such as idempotency and transient
versus persistent errors, are discussed in the “Mind the errors”
section later in the chapter.
▪ HA and DR  The middleware platform supports the expected
level of stability and availability across the interface in line with the
cross-system requirements, for example, HA for mission-critical
scenarios. Requirements for DR is another consideration for the
platform and middleware. For example, if the integrated systems
or the supporting infrastructure experiences an outage, it’s im-
portant to consider what might happen to shared or synchronized
data when the outage is resolved or if either system has to roll
back to the latest backup.
▪ Monitoring and alerting  The platform, middleware, and
supporting infrastructure must support the requirements for
monitoring activity and alerting users and administrators if there’s
a problem. In addition to ensuring that the platform can manage
adequate levels and venues of alerts and notifications, it’s equally
important to consider who in your organization is responsible
for responding to those alerts and notifications and whether the
necessary workload is feasible.
▪ Diagnostics and audit logs  It’s important to also consider the
requirements for monitoring cross-system activity and for audit
trails and logging. You should consider whether the planned
platform and middleware support that.
▪ Extensibility and maintenance  It’s important to consider the
impact of additional business requirements in the long term.
Considerations include the following: the level of effort required to
extend the integration beyond the first implementation; whether
the integration requires expertise in multiple technologies and
changes to multiple systems or whether there is a single low-code/
no-code point of extensibility; how and how often updates are ap-
plied; and what level of technical and functional support is available.
Dynamics 365 apps and Power Platform
▪ ALM  To keep costs as low as possible, consider how you can
allow users to design, configure, and
customize an application to meet a ensure effective lifecycle management, version control, and doc-
customer’s business needs. In doing so, it’s
important to consider performance and
umentation of the integrated solution. You should verify that all
scalability. Ensuring a performant system is components and parts of the solution can use ALM tools such as
a shared responsibility among customers,
partners, ISV providers, and Microsoft. Azure DevOps. For more information about ALM, refer to Chapter
11, “Application lifecycle management.”

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 438
 Fig.
Data consistency across multiple systems is important, and for that to
16-7
occur you need to ensure data quality. Keep in mind the saying that
“if garbage comes in, garbage comes out,” it’s important to verify
Microsoft Power that your data flows correctly and in a timely manner across all the
Platform integrated systems.

The low-code platform that spans The following section discusses the preferred platform to use to inte-
Office 365, Azure, Dynamics 365,
grate with Dynamics 365 apps on Azure.
and standalone applications

Power Platform and Dataverse

Power Platform enables integration between your organization’s
systems. It also provides opportunities for your organization to use
Power BI low-code/no-code development models as well as more complex
Business analytics developer platforms to integrate solutions. In Dynamics 365 apps, this
avenue for integration and much more becomes available when you
use Microsoft Dataverse.

Dataverse provides a data foundation for applications in Customer

Power Apps Engagement apps and Power Platform. When your solution com-
Application development
ponents live in Dataverse, you can natively connect and extend the
solution using Power BI, Power Apps, Power Automate, and Power
Virtual Agents, as shown in Figure 16-7.

Power Automate The primary benefits of using an integration pattern based on

Process automation Dataverse are as follows.

Works with any type of app

▪ Dataverse works with a broad set of data sources through connec-
tors or Microsoft Power Query.
Power Virtual Agent ▪ Dataverse has built-in connectors through Power Automate that
Intelligent virtual agents
offer configurable, deep integration to not only Microsoft cloud
services such as SaaS, Dynamics 365 apps, and Azure, but also to
popular applications such as Adobe, Mailchimp, and Twitter.
▪ Dataverse provides a modern REST-based API, a vast developer
toolkit, and a growing list of samples that link two or more con-
Dataverse nectors, which means that you often don’t need to start your
integration project from a blank slate.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 439
 Works with any kind of data
▪ Dataverse is integrated into data-centric tools and services such
as Microsoft Excel, Outlook, Power BI, Power Query, and AI Builder
A data warehouse provides capabilities to
that are traditionally used by knowledge workers and integration
store, process, aggregate, analyze, data
mine, and report on both current and engineers.
historical data sets typically with data
aggregated from a varied set of cloud
▪ The Dataverse has built-in analytics, reporting, and AI that you
and line-of-business (LOB) applications. A can use to provide insights into your organization and support
data warehouse has specific architecture
optimized for analytical functions. An decision making. You can obtain analytics and reporting data
Enterprise Resource Planning (ERP) using low-code/no-code options or by using the full capability set
solution such as Finance and Operations
apps maintains a highly normalized data of Azure Data Factory, Power BI, and Azure Databricks. For more
set optimized for transactional processing
information about analytics and reporting, refer to Chapter 13,
and business functionality. It’s important
to select the appropriate data warehouse “Business intelligence, reporting, and analytics.”
platform to meet your organization’s data
warehouse and analytical needs.
Let’s take a look now at how you can use Power Platform and
For more information on reporting and
analytics solutions, refer to Chapter 13, Dataverse with Dynamics 365 apps.
“Business intelligence, reporting, and
analytics.”

Dynamics 365 Customer Engagement apps

Customer Engagement apps data is securely stored and managed in
Dataverse. This means that the Customer Engagement apps can na-
tively use Dataverse capabilities to integrate to other applications.

Dynamics 365 apps and Dataverse

Finance and Operations apps are built on a specialized cloud SQL sche-
ma, optimized for online transactional processing (OLTP). However,
you can integrate into Dataverse through several design patterns,
some of which are seamless and synchronous or near real time. When
you connect Dataverse and Finance and Operations apps, you’re
using a rich set of predefined denormalized endpoints in Finance
and Operations called data entities. For more information about how
to connect Finance and Operations apps to Dataverse, refer to the
“Dynamics 365 cross app integration” section later in this chapter.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 440
 Defining business goals
Choosing a design
Choosing a platform
Many factors influence the choice of patterns for your integration.
Choosing
Integration scenarios can be grouped roughly into three kinds: UI,
a design
data, and process integration. While there is some degree of gray area
Choosing a pattern
and overlap between them, there are also some distinct characteristics
Challenges in integration and expected behaviors.

Product-specific guidance
UI integration
In UI integration, the primary point of integration is centered around
an action that’s performed on the UI. The integration might or might
not trigger business logic or cause anything to be written to the sys-
tem. UI integration creates a seamless user experience even though
the data and process might exist in separate systems, as shown in the
example in Figure 16-8.

In this figure, an order taker at a company works in the Sales system

Fig.
16-8
to take an order from a customer, and in that process the order taker
looks at real-time information from an ERP system.
System 1 System 2
Sales Enterprise The UI integration allows the order taker to answer
Resource Planning
the customer’s inquiries accurately and immediately
without switching systems. The information might or
Create quote
might not be used for the order-taking process.
Enter item

The simplest form of UI integration is when a widget,

Look up price
Get price canvas app, or view from one system is embedded
Return price into the UI of another system without ever touch-
ing the business logic or data of that system. This
enables users to focus on their tasks, for example,
Look up availability
Get date scheduling interviews with job applicants or pro-
Finalize quote
Return price viding services to customers, without the need to
switch between several LOB systems. Sometimes
Send quote values can be passed from the hosting system to the
hosted system to provide session context from one
system to the other.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 441
 The following are additional examples of UI integration:
▪ Commodity price ticker; an email view
▪ On-hand inventory
A key characteristic of UI integration
▪ Embedded Power BI dashboard showing the current day’s produc-
design is that it’s embedded.
tion quality issues in a manufacturing plant
The benefit of UI integration is that a
user can retrieve and provide real-time
information from multiple sources
without switching between systems,
thereby saving on training, user licenses,
Data integration
and more importantly, time when Data integration, shown in Figure 16-9, is integration between systems
interacting with customers.
that takes place at the data layer, and data is exchanged or shared
between systems. Data integration is different from process integration
in that both systems work with a representation of the same data,
whereas in process integration the process starts in one system and
continues in the other system.

We use data integration patterns in the following scenarios.

▪ Master or transactional data that isn’t the central part of a single,
continuous business process is synchronized between a process in
one system to a process in another system.
▪ Data is shared or exchanged between systems when needed for
calculations, classifications, or references.
▪ Data is shared or exchanged between systems so that actions that
happen in one system are reflected in the other system.
▪ Aggregate data from a system with a detailed level of data is
Fig.
16-9
exchanged to a system with a higher level representation of data.

System 1 System 2 The following are examples of data integration:

▪ Aggregate marketing data assets
▪ Bank account reconciliation
UI Layer UI Layer
▪ Order status updates

Process / Process /
▪ Exchange rate updates
Business logic Business logic ▪ Postal code and city list synchronization
▪ Accounts and customer data synchronization
Data integration

Keep in mind that often data doesn’t originate from a

system within your own organization; it can come from an external source
to upload. Conversely, data might be extracted from your system to be
sent to an auditor, a regulatory body, or an industry data exchange.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 442
 Fig.
In the example in Figure 16-10, orders are synchro-
16-10
nized between a Sales and an ERP system. When the
order is created in Sales, the ERP system sends regu-
System 1 System 2
Sales Enterprise lar updates of the order data to Sales, which enables
Resource Planning
users to find accurate order status information.
Create order
Create order
Data integration is also a useful integration type
Ship order
because multiple features within Dynamics 365 apps
Update status enable administrators or users to configure many of
these integration points out of the box, for example,
Look up status Invoice order Microsoft Teams, journal uploads for expenses or bank
Update status account statements, exchange rate synchronization,
and features for regulatory reporting and extracts.
Order paid

Update status When designing data integration, you should

consider which system will be the system of record,
Look up status or owner, of the information. There are scenarios
in which this is clear cut, for example, uploading
worker data from Dynamics 365 Human Resources
into Finance and Operations apps (the former is the
Fig.
16-11 system of record). But there are also scenarios in
which each system owns a separate part of the over-
System 1 System 2 all entity, for example, the integration of Accounts to
Sales Enterprise Customers between Dynamics 365 Sales and Finance
Resource Planning and Operations apps. In this scenario, Sales is the
owner of the basic information, but that information
is just a small subset of the fields in the customer
Create order
entity in Finance and Operations apps.

Ship order
Another type of data integration, shown in Figure
16-11, is when two systems share the data layer, so
Look up status updates to one system are instantly reflected in the
other system. This is possible with cross app integra-
Invoice order tion with Dynamics 365 apps and Dataverse.

First-party model-driven applications such as

Look up status
Dynamics 365 Sales, Field Service, and custom
Power Apps all share the same data layer, even if

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 443
 applications are built and deployed separately.

With Dataverse it’s also possible to integrate Customer Engagement

apps and Finance and Operations apps so that a similar seamless
experience can exist between the Sales system and the ERP system.

Systems may have separate processes and even different business

logic, but if the underlying data layer is the same, the need for transfers
of data, synchronization, and programming to transform the data is
completely eliminated. This kind of integration is now possible because
the development of shared data stores such as Dataverse set standards
for how a certain kind of data is defined in the data layer.

Process integration
Process integration refers to when a business process is designed to
span multiple systems. There are many varieties of process integrations,
such as from a plan-to-produce workflow in which production fore-
casting or block scheduling occurs in a Production Scheduling System
and the rest of the manufacturing and supply chain management
process (production, order management, and fulfillment) and the
Fig.
16-12 billing process occur in an ERP system. Figure 16-12
shows this type of integration.
System 1 System 2
Production Enterprise
scheduling Resource Planning In this scenario, the business process spans two
systems. The user in the Production Scheduling
Plan production orders
System creates and maintains the production sched-
Firm production order Create order ule, thereby initiating the business logic to procure,
Create
production produce, and reserve the order in the ERP system for
order Procure raw materials
later picking and shipping. User interaction might
Update Consume raw materials be required in the ERP system, but the point is that
production
order
the event of creating the order in the Production
Update production order Update order
Scheduling System triggers certain business logic in
Report finished the ERP system, and the rest of the business process
steps are handled either automatically or by the
Pick, pack, and ship order
appropriate users in the ERP system.

Process integration can span multiple systems and

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 444
 often requires batched, scheduled, and sometimes near real-time
integration.

A key characteristic of process integration In the example shown in Figure 16-12, without the integration part,
design is that it’s event driven.
orders would have to be manually created twice, increasing time and
The benefits of process integration are
risking errors such as typos and missed updates.
accuracy, efficient timing of information
and activities in the organization, and
reduction of manual errors.

Design options on Power Platform

Let’s now discuss some of the common services available across
Power Platform and Azure that can help in designing a dynamic,
scalable solution.

With Dataverse, you can leverage Power Platform to integrate apps.

You should consider several components and features of Azure and
Power Platform when designing an overall solution as well as
individual integration.

Power Automate
Power Automate provides low-code/no-code solutions to help you au-
tomate workflows and build integration between various applications.
Power Automate automates repetitive and manual tasks and seamless-
ly integrates Dynamics 365 apps inside and outside Power Platform.

In fact, Power Automate has connectors to more than 300 specific

applications and technologies and a vast selection of integration
samples to start from, all with the goal of helping users to be more
productive. Power Automate is available to Office 365 users who have
the appropriate license and access. For more information, read about
Power Automate and What is Microsoft Dataverse?

Azure integration frameworks

Azure provides designers with a rich, scalable feature set and
user-friendly design elements as well as the option to build more com-
plex integration solutions. Developers can use tools including Visual
Studio, Visual Studio Code, and Azure DevOps.

Power Platform and Azure together can handle integration

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 445
 requirements from simple scenarios to enterprise architecture scenarios.

For more information, visit Azure Logic Apps and Azure

Integration Services.

Defining business goals

Choosing a pattern
Choosing a platform
Choosing the right pattern is a critical part of successfully implementing
Choosing a design
integration between systems. When choosing an integration pattern, you
Choosing a should consider factors such as what its main functionality is and how it’s
pattern built, including platform, language, user interface, and the connectivity

Challenges in integration type it handles. We recommend that you also consider what type of
actions you need the integration to perform, such as the following:
Product-specific guidance ▪ Data types and formats  What types of data are you sending?
Transactional, text, HTML?
▪ Data availability  When do you want the data to be ready, from
source to target? Is it needed in real time, or do you just need to
collect all the data at the end of the day and send it in a scheduled
batch to its target?
▪ Service protection and throttling  When you use certain inte-
gration patterns, service protection might be built in, for example,
by limiting the number of records, or maximum system resource
utilization used for interface activity.. Sometimes the provider
places this limitation on you, but in other cases, if you’re offering
integration points to other systems, you want to impose a limita-
tion or throttling to ensure that the general performance of your
system is not impacted by traffic from external requests.
▪ Transformation of data  Do you want to convert or aggre-
gate transactional data into analytical data? Are you changing
.txt to HTML?
▪ Triggers  What action will trigger sending data from the source
to the target?
▪ Trigger actions  Do you want specific actions to be automated
after the data arrives at the target?
▪ Process error handling  What type of monitoring will you put in
place to detect any issues with the interfaces? What type of alerts
do you want to handle the patterns that you’ll be using?

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 446
 ▪ Flow direction  Is the flow bidirectional or unidirectional?
▪ UI  The UI determines how the data is going to be displayed once
it arrives at the target system. The UI determines the user experi-
ence; how the user will interact with and handle the information.
▪ Scalability  How will the interface pattern handle the expected
transaction volumes in the present, short term, and long term?
What would happen if you exceeded those expectations?

Pattern directions
Let’s take a closer look at some common patterns for individual inte-
grations and the pros and cons for each. This table is generalized; for
more information, refer to the “Product-specific guidance” section in
this chapter.

Pattern Mechanism Trigger Considerations Use when

Real time or Data is exchanged User action or Pros: Fast request and response Generally, not a best
synchronous synchronously, invoking system event. round trip. Real-time values and practice to use because of
actions via services. information. the risk of processes getting
stuck and solutions that
Cons: Small payloads. Action-
are fragile.
reaction can cause tight coupling of
systems. Risk of ripple effect from Use when real-time
transient errors. Sensitive to latency. information is critical.

Asynchronous Data is exchanged Scheduled and user Pros: Loose coupling of systems Most recommended
unattended on a initiated. Can wait for off makes the solution robust. Load integration patterns and
periodic schedule or as a hours or idle time. balancing over time and resources. technologies supported are
trickle feed using Can be very close to real time. asynchronous, although
messaging patterns. Timely error handling. they can be near real time.

Cons: Delay in response and

visibility to changes across systems.

Push One system puts Originating system user or Pros: If technical expertise lies For reactive scenarios.
(pushes) data into system event. within the pushing system, the Receiving system provides
another. Information custom effort lies here. Good for a turnkey API and
flows from the originator reactive scenarios. organization’s developer
to the receiver. Cons: Pushing system might not skillset is with the
have visibility into availability and originating system.
load and idle times in the
receiving system.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 447
 Pattern Mechanism Trigger Considerations Use when
Pull Receiving system requests Receiving system request Pros: If technical expertise lies For proactive scenarios.
data from the originator, based on schedule. within the pulling system, the We might not have the
a subtle but significant custom effort lies here. Good for option to add triggers or
difference from the proactive scenarios. Clear visibility events in the originating
Push pattern. into availability and load and idle system. Originating system
times in the receiving system. provides a turnkey API and
Cons: Originating system might not organization’s developer
have the APIs needed to pull from. skillset is with the
receiving system.

One-way sync Data from one system is Data state, system, or Pros: Establishes a clear system of One system is owner or
synchronized to another user event. record. Simple conflict resolution. system of record and
by one or more trigger other systems consume
Cons: Sometimes the receiving
events. that data.
system or non system of record
doesn’t have built-in capability to Consider scenarios in
make the data read only, which which the originating
can confuse users. table is a subset of a
similar table in the target.

Bidirectional sync Data from two or more Data state, system, Pros: Data is kept in sync across Scenarios in which there
systems are synchronized. or user event. applications. Acquired divisions on isn’t a clear system of
multiple platforms can continue to record.
use their existing systems. Users can Data from best-of-
use their system to make changes. breed systems should be
Cons: Complex conflict available in Dataverse for
resolution. Redundant data is Power Apps and other
replicated for each system. tools and services.
Synchronized data might be a
subset of data in systems. The rest
must be automatically given values
or manually updated later.

Aggregation Data from a specialized Any. Pros: Detailed data is kept in the Aggregates are needed for
system is integrated to system where it’s used. Aggregation calculating or processing,
another system on an can derive a small dataset from a for example, on-hand
aggregated level for large one, thus limiting traffic inventory by warehouse,
processing or reporting. across platforms. revenue by invoice header,
or revenue by customer by
Cons: Users often expect to be able
day, or operational data
to drill down to the detailed level.
for posting in a finance
While this could be done with
system.
embedding, it does require
additional integration complexity or
users operating in two systems.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 448
 Pattern Mechanism Trigger Considerations Use when
Embedding Information from one User event. Pros: Simple because the data A mix of information from
system is seamlessly remains in the originating system. first-party applications (for
integrated into the UI of example, Bing, Power BI,
Cons: Difficult to use the data for
another system. and Exchange), third-party
calculations for processing.
components, canvas apps,
or other information
embedded in the UI of
an application.

This is common practice

in Customer Engagement
apps.

Batching Batching is the practice Any. Pros: Great for use with Whenever it isn’t
of gathering and messaging services and other necessary to transmit
transporting a set of asynchronous integration individual records.
messages or records in a patterns. Fewer individual
batch to limit chatter and packages and less message traffic.
overhead. Cons: Data freshness is lower.
Load in the receiving system can
be affected if business logic is
executed on message arrival.

Additional pattern considerations

Let’s take a moment to discuss When choosing a specific pattern, it’s important to take a broader look
synchronous integration patterns and near
real-time patterns because these are often at the integration context.
mischaracterized and mistakenly used
interchangeably.

An example of a synchronous integration Messaging patterns, middleware,

pattern is a process in which a user
performs a web service-based lookup of and service bus
current on-hand values from one system
to another. In this example, the user A service bus refers to asynchronous messaging on a well-equipped
waits for a response before continuing messaging platform that includes a service-based channel and
the process, thus making the process
both latency and availability dependent. endpoint connectors. This is a common, modern way of managing
This is known as coupling, and it should
complex integration architecture across distributed systems. The
generally be avoided.
messaging platform is often referred to as the middleware, and it often
A near real-time pattern for the same
design scenario provides a user’s system
offers message routing and message orchestration, transformation,
with fresh on-hand inventory information and even in some cases, load balancing message error handling, moni-
every one to three minutes, thereby
enabling the user to continue the process toring, logging, and notification services.
even if there’s a transient outage in the
providing system or the supporting
architecture. This pattern combined with Messages are pieces of information or datasets that can be transferred
using a messaging-based integration
architecture results in a robust solution.
between systems. Each message can contain instructions such as an
event fired or a delete request, or it can contain a payload like a record

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 449
 or dataset. Some messaging platforms allow for prioritizing messages
based on type, origin, destination, or payload.

Modern messaging platforms are based on service endpoints, REST,

and JSON, but the platforms can offer many additional features.
Connections between endpoints are called channels and the message
paths are called routing.

The benefit of messaging patterns is that the messaging platform

handles the flow of information between systems and detects individu-
al systems or SaaS service endpoint transient outages and handles the
queueing of messages to the affected systems until they’re back online.
Figure 16-13 shows the typical components of a messaging platform.

Fig.
16-13
Monitoring Orchestration
and alerts and routing

Endpoint Channel Routing Channel Endpoint

Message

Data logging

The asynchronous nature of the messaging patterns allows for a robust,

decoupled integration architecture with less manual intervention and
lower maintenance costs.

When implementing a messaging pattern, consider the following:

▪ Frequency and volume  It’s critical to have a clear picture of the
frequency and transactional volume of data that will go through
the interface, how that load is distributed over time, and what the
requirements for these are in the longer term.
▪ Idempotency and out-of-order messages Mechanisms
are available to ensure that the messages are idempotent, for
example, using a message ID to check whether the message is

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 450
 a duplicate or an out-of-order message. Services such as Azure
Service Bus support this capability and allow you to implement it
without the need for additional code.
▪ Service-level agreement (SLA) or end-to-end cycle latency 
Consider whether there are minimum requirements for how fresh
the data in the receiving system must be. For example, orders
taken on an e-commerce site should reach the warehouse in less
than five minutes.
▪ Triggers and actions end-to-end process  It’s also important to
look at the cross-system business process in the bigger picture:
▫ What happens before the integration steps?
▫ Is there potential risk of introducing semantic errors earlier in
the process?
▫ What event triggers the integration?
▫ Does the integration trigger action in the receiving system?
▫ What business process steps happen after the integration steps
are done?
▫ If an error is identified, logged, and notified, how is the problem
corrected? How often might errors occur, how long might it
take to correct them, and who is responsible for making the
corrections?
▪ Batching  Batching messages or datasets enables less frequent
communication, or “chatter.” But it also typically makes messages
and payloads bigger.
▫ Do the requirements support a batched approach in which
datasets are consolidated into each message, or does the
integration require individual records or rows?
▫ If a batched approach is suitable, how many rows and records
are in each message?
▫ Are there service protection volume limits to consider?
▪ Topology and architecture  Topology and architecture are
important considerations for the short and long term as your
organization grows internationally or as you migrate systems to
the cloud and need to address new challenges and considerations.
▫ Are there integration components that will span the cloud and
on-premises boundary? Does the messaging service handle that?
▫ Where in the world is the data located, and does the inherent
latency between geographic regions support the requirements?

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 451
 ▫ Does any data exist in countries or regions of the world that
have local privacy policies and regulations?
▫ Are there service protection limitations in technology?Protocol
or platform dependencies in any of the integrated systems?
▫ Does the data exchanged require transformation, reformatting,
or remapping as part of the integration?

For more information, read about Messaging services on Azure.

We highly recommend that when you design integration, you don’t

assume that everything will always work as intended. In fact, when
working with integration, there are many external factors that we often
cannot completely control. Therefore, we need to make sure that our
integrations are robust by design, which we discuss in the next section.

Mind the errors

Designing for failure is a crucial component of architecting any solu-
tion, including integration components. As part of the integration
architecture, it’s important to incorporate the following:
▪ Error logging  Select the platform (such as a log file or database)
to log the errors.
▪ Error monitoring and notifications
▫ Define the process and technical approach to monitor errors.
▫ Select an error notification approach, which is the process to
notify administrators and other stakeholders if there’s an error.
Business-critical errors might need a different notification
approach than non-critical errors.
▪ Business continuity  It’s important to plan for business continu-
ity with minimal disruption to business in the event of errors. For
more information, see Chapter 20, “Service the solution.”

The more integration touchpoints there are, the greater the potential
for errors. Therefore, error handling also needs to be planned in ac-
cordance with the integration scenario. For example, in the scenario
of a synchronous integration pattern, an error might require a com-
plete rollback of the entire process, whereas in an asynchronous data
integration scenario, it might be acceptable to fix a data issue just by

452
 Fig.
notifying the administrator. (See Figure 16-14.)
16-14

System 1 System 2 Let’s now discuss error management for two key pat-
Quote Enterprise terns of synchronous and asynchronous integration.
Resource Planning

As a reminder, in a synchronous integration pattern,

Create quote
each step is dependent on the completion of a
Enter item preceding step. In the event of an error, consider a
retrial or rollback, which option to use is dependent
Look up price
Get price on whether the error is transient or persistent. In the
scenario of a transient error, the normal behavior of

?
Return price
the application will resume after a few retries. Note
that retrial limits should be predefined to avoid a
situation in which all resources are blocked. Once
the retrial limit has been crossed, the entire process will need to be
rolled back, and appropriate error messages should be logged.

For an asynchronous integration, the error management strategy

might differ depending on your business requirements. Consider
creating a queue that can continue to validate the message being
exchanged and retry until the retrial limit is reached. The message that
fails from being exchanged can also be stored in another queue for
later intervention.

Transient errors such as network timeouts get fixed after a few retries.
However, persistent errors require intervention to be fixed.

An important consideration as you build your retrial mechanisms is to

ensure that your requests are idempotent. This will prevent data dupli-
cation issues due to retries.

The following are some of the most common error scenarios in any
integration between Dynamics 365 apps and other applications:
▪ System becomes unavailable
▪ Authorization and authentication errors
▪ Errors caused by platform limits
▪ Errors caused by service protection limits applied to ensure
service levels

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 453
 ▫ API limits (limits allowed based on the user license type)
▫ Throttling
▫ Process errors
▪ Runtime exceptions

It’s critical to plan error logging, monitoring, and communication as part

of an overall error-handling approach. Logging errors with the exact
timestamp and message details helps the implementation team and
application owners troubleshoot the errors. Error tracking also enables
the application owners to identify potential areas of improvement in
their process by reviewing frequent errors, if any. Figure 16-15 shows a
persistent blocking error. The platform supports logging and notifica-
tions so that an administrator can be notified and resolve the issue.
Fig.
16-15

System 1 System 2 It’s also important to identify owners who

Quote Enterprise
Resource Planning should receive error details through email
as part of regular monitoring. The monitor-
Create quote
ing and notification approach might differ
Enter item
for business-critical errors versus non-crit-
Look up price
Get price
ical errors. In the case of a mission-critical
application integration, some organizations
Return price
Notification might choose to send text notifications to
the administrators to speed up the process.
Log

Error messages displayed to end users

through the UI should be clear, precise, and short. They should also
include information users can act on, such as point of contact informa-
tion or a link to an FAQ. This makes the user experience a more positive
one and should be considered a best practice as part of error handling.

Consider using tools such as Azure Monitor that help maximize your
applications’ availability to collect, analyze, and act on telemetry.

For more information, visit Azure Monitor overview.

To summarize, error logging and monitoring have a significant impact

on the following key factors and therefore are considered essential:
▪ Application health including performance and stability

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 454
 ▪ Ease of troubleshooting and resolving errors
▪ Minimize business impacts when errors occur
▪ Increase user satisfaction

For more information about integration patterns and technologies,

refer to the “Product-specific guidance” section later in this chapter.

Defining business goals

Choosing a platform
Challenges in integration
The diverse applications used in organizations make up the IT back-
Choosing a design
bone of those companies. Organizations use applications that might
Choosing a pattern be a combination of on-premises, cloud, and third parties. These
applications need to communicate with each other for different busi-
Challenges in
ness needs. Integration challenges during a project implementation
integration
can cause delays or cost increases. A key focus of any IT department is
Product-specific guidance
to ensure that these integrations enable business productivity and not
become a blocker to business growth.

Integration challenges can be categorized into the following areas.

Business
Each integration scenario has a direct impact on the application you’re
integrating with and on its users. Any downstream applications might also
be indirectly impacted. For integration to be implemented in a successful
manner, you should address the following during the Initiate stage:
▪ Identification of application owners and stakeholders 
Application owners need to identify downstream applications that
might be impacted in an integration. However, a common scenar-
io is to bring in these owners after the planning is complete. This
often results in misaligned timelines and scope and in turn creates
project delays and poor quality. Integration design needs to take
every impacted application into consideration.
▪ Alignment between business owners  Business stakeholders
have different needs for their organization’s various applications.
Unless there is a collective understanding of integration scenarios
and approaches, requirements might be mismatched among the

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 455
 various owners. This in turn often results in delayed timelines, cost
overruns, and a lack of accountability. System integrators should
consider the following:
▫ Identify the key owners and bring them together to walk
through the scenarios.
▫ Differentiate between process, data, and UI integration to
simplify and streamline the integration scope.
▫ Outline the impact on business groups affected by the
integration.
▫ Highlight issues and risks in the absence of following a consis-
tent approach.

A transparent conversation enables business stakeholders to understand

the underlying risks and benefits and thus build a common perspective.
▪ Ambiguous and unrealistic expectations  Integration require-
ments can sometimes be ambiguous or incorrectly perceived
as mandatory. Common examples of the latter are unnecessary
synchronous integrations and replicating high transactional data
volumes into your business application to make it a reporting
application. Such architecture decisions can result in solutions that
are neither scalable nor performant. Encouraging and facilitating
conversations to help business stakeholders define their require-
ments clearly and understand the necessary tradeoffs as part of
the architecture decisions is a critical step in ensuring a smooth
implementation.
▪ Lack of business continuity planning  This is an aspect of
implementing any IT project that most often is ignored. However,
no matter the type of integration, cloud to cloud, on-premises to
cloud, every business group must define operating procedures
that will be enforced as a part of any availability and disaster
recovery situation. As an example, consider a scenario in which
the on-premises application of a Dynamics 365 app is temporarily
unavailable. This can impact integration components, resulting in
data or process issues.
▪ IT-driven requirements and design  Implementation teams
sometimes consider integration between systems as a primarily
technical exercise and overlook the critical role played by business
teams. However, successful integrations between Dynamics 365

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 456
 apps often depend completely on a well-defined cross-system
business process, and if the specific process details and broader
requirements are not properly defined, the integration project
might take much longer or go through several iterations of trial
and error. For more information, refer to Chapter 7, “Process-
focused solution.”

Technology
Most enterprises have legacy applications with traditional, on-premises
architecture. The move to cloud applications requires consideration of
the patterns they support and the best practices when planning for inte-
gration. A low-code/no-code pattern should now be at the forefront of
any integration architecture. Engaging in conversations about not just the
current setup but also about future planning for performance, extensi-
bility, and maintenance plays a key role in choosing the right technology.
When choosing the appropriate technology, consider the following.
▪ Does one size truly fit all?  Many enterprises have an enterprise
architecture approach that might or might not be aligned with
the modern approaches for cloud applications. Prior to investing
in a specific approach, evaluate whether the existing architecture
aligns with cloud patterns. Sometimes, a generic approach is
taken; this can result in inefficiencies in integration, unscalable
architecture, and poor user experience and adoption. Therefore,
it’s crucial to consider design paradigms such as the following:
▫ Definition of the integration approach based on multiple
parameters
▫ Benefit of a proof of concept to determine the pros and cons of
one approach over another
▫ Synchronous versus asynchronous integration
▫ Process, UI, and data integration
▫ Single record or batch
▫ Frequency and direction of the integration
▫ Message reliability and speed
▫ Data volume
▫ Time expectations (some scenarios require a batch integration
to be completed during a specific time window)
▫ Error management and retries

457
 ▪ Will sensitive data be exposed?  System integrators must un-
derstand IT and customer concerns around security, especially in
the context of integrating on-premises applications with Dynamics
365 apps. Categorizing security concerns as follows aids in identi-
fying who and what is required to help address them:
▫ Access control
▫ Data protection
▫ Compliance and regulatory requirements
▫ Transparency
For more information, refer to Chapter 12, “Security.”
▪ Storage costs and platform limits  To ensure service quality
and availability, Dynamics 365 apps and Power Platform enforces
entitlement limits. These limits help protect service quality and
performance from interference by noisy behavior that can create
disruptions. System integrators must incorporate these limits as
part of the overall architecture. If these aren’t planned for, the
service will be throttled, resulting in failure and errors within the
integration layer. Storage costs are also often ignored. Although
this might not have an impact initially, over time, it can result
in increased storage costs and therefore should be planned for
appropriately.
▪ Connectivity  Network latency can become a constraint, espe-
cially in heavy data-load scenarios. System integrators must ensure
that they design payloads accordingly for the efficient use of
network resources without compromising performance.
▪ Anti-patterns  Implementation architects should follow the
best practices for Dynamics 365 apps. Sometimes these architects
don’t take cloud patterns sufficiently into account in integration
scenarios with on-premises applications, resulting in poor perfor-
mance. The behaviors leading to such situations are referred to as
anti-patterns. Consider the following common anti-patterns:
▫ Are there repeated connections between on-premises compo-
nents and Dynamics 365 apps that impact performance? If so,
consider sending data in batches.
▫ Is there latency between a customer’s on-premises applications
and a Dynamics 365 datacenter? If so, consider using a cloud
service such as Power Automate, Azure Functions, or Azure SQL
to reduce the latency impact.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 458
 ▫ Is a lot of data being synchronized with Dynamics 365 apps
for reporting purposes? Keep in mind that the database under
Dynamics 365 apps isn’t intended to be a data warehouse for
all of the organization’s data assets. Consider using a dedicated
datastore for reporting purposes.
▪ Proprietary technology  Customers might be using third-party
technology within their IT landscape that doesn’t provide interface
details or adequate support to enable integration easily. Often
such issues are identified either toward the end of design or during
the development stage. This causes delays in the project timeline,
burdening the customer with time constraints to mitigate such risks.
System integrators must highlight such dependencies in the plan-
ning stage to ensure adequate support or an alternate approach.
▪ Readiness  With the increasing pace of transformations in the
technology world, architects sometime choose an approach due
more to its familiarity than its applicability. Customers and system
integrators must evaluate whether to request additional resources
specialized in the specific technology who will be a better fit for
their current and future needs.

Project governance
The initial stage of a project should include a defined project gover-
nance model. Integration between on-premises and Dynamics 365
apps can range from simple to complex, and the lack of well-defined
project governance areas results in gaps and issues in the smooth im-
plementation of a project. Following are common project governance
concerns specifically for the integration components:

▪ Has the impact of the integrations been identified for the end user,
process, and reporting? This might require planning for change
management activities, including communication and training.
▪ Making a solution performant should be at the forefront of any
design decisions made by the implementation team. This applies
equally to the integration layer and the application layer. Is perfor-
mance testing planned and does it cover integration components?
Performance testing is another activity that tends to be considered
optional. However, architects and project managers must consider

459
 embedding this in their Dynamics 365 apps implementations. This
will help identify any performance bottlenecks prior to deploy-
ment for end users.
▪ Are development and test environments available for all appli-
cations for thorough system integration testing? Is a plan for
stub-based testing during the unit testing phase required?

Asking these questions during the initial stages of the project enables
both the implementation partner and customer to proactively identify
and plan for any dependencies and risks.

For more information, refer to Chapter 8, “Project governance.”

Defining business goals

Product-specific
guidance
Choosing a platform

Choosing a design

Choosing a pattern The following sections discuss specific integration-related consid-

erations for Finance and Operations apps; specific technologies and
Challenges in integration
platforms for integration from Customer Engagement apps; and sce-
Product-specific narios for cross app integration for Dynamics 365 apps.
guidance

Finance and Operations apps

Many ways of integrating into and from Finance and Operations apps
use an abstraction layer with integration and data migration capabili-
ties that shield users, administrators, and developers from the heavily
normalized database. This abstraction layer consists of data entities.

Data entities
In Finance and Operations apps , a data entity encapsulates a business
concept, for example, a customer or sales order line, in a format that
makes development and integration easier. It’s a denormalized view in
which each row contains all the data from a main table and its related
tables instead of the complex view of the normalized data model
behind it.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 460
 Said another way, data entities provide conceptual abstraction and
encapsulation, a denormalized view of what’s in the underlying table
schema to represent key data concepts and functionality. There are
The out-of-the-box data entities are more than 2,000 out-of-the-box data entities.
general purpose entities built to support
a wide variety of features surrounding
a business entity. For implementations Data entities help encompass all the table-related information and make
in which a high-volume, low-latency
the integration, import, and export of data possible without the need to
interface is required, it’s recommended
to build custom data entities with the pay attention to the complex normalization or business logic behind the
necessary targeted and specific features.
scenes. For more information, see the “Data entities overview.”
You can replicate an out-of-the-box data
entity and remove any fields and logic
around features not used by your Finance and Operations apps integration patterns
organization to create a highly performant
The following table is a list of integration patterns with pattern descrip-
interface.
tions, pros and cons, and use cases.

Pattern Mechanism Trigger Considerations Use when

OData Data is exchanged User actions and Pros: It allows developers and Use with messaging services
synchronously, invoking system events. service platforms to interact with and point-to-point integration.
actions using RESTful web data by using RESTful web services. Inbound (Push) or outbound
services and the HTTP It provides a simple and uniform (Pull) of datasets from entities
protocol stack. way to share data in a discoverable when the volume and payload
All data entities are manner. It enables broad integration are relatively low.
marked public, and CRUD across products.
operations are mapped. Cons: The pattern is synchronous
and subject to Service Protection
limitations and throttling. Cold start
might be slow.

Custom A developer can create User actions and Pros: Easy for developers to add and Used when invoking an
Services—SOAP, external web services by system events. expose service endpoints to use with action or update, for example,
REST, and JSON extending the application integration platforms. invoicing a sales order or
with X++. Endpoints are returning a specific value. We
Cons: Requires ALM and SDLC
deployed for SOAP, REST, recommend using REST Custom
for coding and deployment of
and JSON. Services in general because it’s
extensions into Finance and
optimized for the web.
Operations apps. The payload is low
compared to other patterns. REST is preferred for high-
volume integrations because
there’s reduced overhead
compared to other stacks such
as OData.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 461
 Pattern Mechanism Trigger Considerations Use when
Consuming A developer can consume Scheduled and user Pros: Easy for developers to add and Use when consuming services
web services external web services initiated. Can wait expose service endpoints to use with from other SaaS platforms
by adding a reference in for off hours or idle integration platforms. or products, for example,
X++. time. commodity tickers and
Cons: Requires ALM and SDLC
lookups of real-time values.
for coding and deployment of
references into Finance and The recommended pattern
Operations apps. The payload is is to use Power Automate
low. Risk of hardcoding connection instead when possible.
strings and credentials in related
code. Maintenance workloads and
risks when the service is updated.

Data The REST API helps Originating system Pros: On-premises and large- Large-volume integrations.
Management integrate by using data users or system volume support. The only interface Scheduling and
Framework REST packages. events. that supports change tracking. transformations happen
Package API outside Finance and
Cons: Supports only data packages.
(asynchronous, Operations apps.
batched, cloud, or
on-premises)

Recurring Data With the Data Receiving system Pros: On-premises and large-volume Large-volume integrations
Management Management REST API, requests based on support. Supports files and packages. of files and packages.
integration REST you can schedule the the schedule. Supports recurrence scheduled in Scheduling and
API (asynchronous, recurring integration Finance and transformations (XSLT) if transformations happen
cloud, or of files and packages. the file is in XML. inside Finance and
on-premises) Supports SOAP and REST. Operations apps.
Cons: None.

Electronic A tool that configures Data state, system, Pros: Data extracts and imports Electronic reporting to
Reporting formats for incoming or user events. are configurable in Finance and regulatory authorities and
and outgoing Operations apps. It supports several similar entities.
electronic documents in local government formats out of
accordance with the legal the box. It can be scheduled for
requirements of countries recurrence.
or regions. Cons: Comprehensive configuration
when used for messaging purposes.

Excel and Microsoft Office Data state, system, Pros: Out-of-the-box integration (ex- Extracts for ad hoc reporting
Office integration integration capabilities or user events. port and edit) on almost every screen or calculations.
enable user productivity. in the product. Fast editing of column values
Cons: Performance decreases with and entry of data from manu-
the size of the dataset. al sources.

Import of journal records such

as general journals, expenses,
and bank statements and
similar transactional data.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 462
 Pattern Mechanism Trigger Considerations Use when
Business events Business events provide a User or system Pros: Provides events that can be To integrate with Azure
mechanism that lets events. captured by Power Automate, Logic Event Grid, Power Automate,
external systems, Power Apps, and Azure Event Grid. or Logic Apps. To notify of
Automate, and Azure events inherently driven
Cons: Extensions are needed to add
messaging services receive by a single data entity, for
custom events.
notifications from Finance example, an update of a
and Operations apps. document or a pass or fail of a
quality order.

Embedded Power Finance and Operations Users. Pros: Seamlessly integrates Whenever the required data
Apps (UI) apps support integration information from Dataverse without exists in Dataverse and is
with Power Apps. Canvas integrating the backend. Opens loosely coupled with Finance
apps can be embedded opportunities for low-code/no-code and Operations apps.
into the Finance and options directly into the Finance and Using an app embedded in
Operations apps UI to Operations apps UI without the need the UI provides a seamless
augment the product’s for updates and compatibility. experience for users.
functionality seamlessly Cons: Power Apps and related
with Dataverse. artifacts aren’t deployed with the
build and must be configured in the
environment directly. Separate ALM
stories for Power Apps.

Embedded Power Seamlessly integrates Users. Pros: By using graphics and visuals Whenever reports,
BI (UI) Power BI reports, supported by Power BI to present dashboards, or visuals exist in
dashboards, and visuals data from any source, workspaces Power BI.
with information from can provide highly visual and
Dataverse or any interactive experiences for users
other source, without without leaving the Finance and
integrating the backend. Operations apps UI.

Cons: Power BI artifacts are not

deployed with the build and must be
configured within the environment
directly. Separate ALM stories for
Power BI components and Finance
and Operations apps.

IFrame (UI) The Website Host control Users. Pros: Seamlessly integrates UI from When information from
enables developers to other systems or apps without loosely coupled systems
embed third-party apps integrating the backend. can be displayed within the
directly into Finance and Information goes directly into the Finance and Operations
Operations apps inside an Finance and Operations apps UI apps UI. The experience is
IFrame. without the need for updates or enhanced if the external
compatibility. system supports single sign-
on (SSO) and deep linking.
Cons: External systems might have
separate lifecycles, and updates to
those might affect user experience
with little notice.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 463
 Priority-based throttling
Service protection is important for ensuring system responsiveness,
availability, and performance. In Finance and Operations apps, service
protection is enforced by throttling. Throttling affects OData and
custom service pattern integrations only. Administrators can configure
priorities for external services (registered applications) directly in the
application so that lower priority integrations are throttled before
high-priority integrations.

For more information, read about Priority-based throttling.

Requests that are throttled receive a response containing a retry-after

value, indicating when the integrating application can attempt a retry.

Microsoft Dynamics Lifecycle Services (LCS) monitors throttling activity.

Customer Engagement
In this section we talk about frameworks and platforms to use when
integrating from Customer Engagement.

IFrames
IFrame is a popular approach commonly used for hosting external
URL-based applications. Consider using the Restrict cross-frame script-
ing options to ensure security.

Power Apps component framework

Power Apps component framework is a modern approach for creating
rich visual components that allow enhancement of the user interface of
model-driven apps. It leverages client frameworks such as React, and
enhancements can be reused across Dynamics 365 apps. It can also pull
data from an external application and display the data in a visual form.

Canvas apps
Canvas apps is a cloud service that enables citizen developers to easily
build business apps without the need to write any code. These apps can use
connectors from other cloud services and be embedded in model-
driven apps to present data from other applications on the user interface.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 464
HTML web resources
A precursor to Power Apps component framework, HTML web re-
sources were commonly used to render data in a visual form, providing
designers with more flexibility. They can be used to pull data into
external applications using the available endpoints.

Dynamics 365 Channel Integration Framework

Dynamics 365 Channel Integration Framework hosts the widgets and
triggers the events for scenarios such as telephony and chat integra-
tion. There are multiple versions of Channel Integration Framework for
different scenarios, including multi-session and single-session needs.
Channel Integration Framework also integrates channels such as voice,
chat, and messaging.

Virtual tables
Virtual tables pull data on demand from external data sources. This ap-
proach is implemented as tables within the Dataverse layer but doesn’t
replicate the data because the data is pulled real time on demand. For
more information, read about the limitations of virtual tables.

Azure Data Factory

Azure Data Factory provides a code-free, serverless data integration
service to seamlessly orchestrate ETL (extract, transform, and load) and
ELT (extract, load, and transform) operations with disparate big data
sources. It can ingest data from Dynamics 365 apps using the web API
endpoints. We recommend that you consider platform and service
protection limits as well as Azure costs.

Webhooks
Commonly used for near real-time integration scenarios, webhooks
can be invoked to call an external application upon the trigger of a
server event. When choosing between the webhooks model and the
Azure Service Bus integration, consider the following:
▪ Azure Service Bus works for high-scale processing and provides a
full queueing mechanism if Dataverse pushes many events.
▪ Webhooks enable synchronous and asynchronous steps, whereas
Azure Service Bus enables only asynchronous steps.

465
 ▪ Both webhooks and Azure Service Bus can be invoked from Power
Automate or plug-ins.

For more information, visit Use Webhooks to create external handlers

for server events.

Azure Functions
Azure Functions uses serverless architecture and can be used to extend
business logic, including calling external applications. It runs in Azure
and operates at scale. Azure Functions can be called through Power
Automate and Azure Logic Apps. For more information, read about
Azure Functions and Using Azure Functions in Power Apps.

Dynamics 365 cross app integration

In this section we look at the integration between two or more
Dynamics 365 apps.

Some of the patterns mentioned earlier in this chapter are specific

to when we want to integrate one of the Dynamics 365 apps into a
Because Dataverse is cloud based, this third-party or custom-built system.
direct integration option isn’t available for
on-premises implementations of Finance
and Operations apps. You can use the Data
Four options are available for integrating the Customer Engagement apps
Management REST API instead.
suite and Finance and Operations apps. Note that we’re actually inte-
For more information, read the “Finance
and Operations virtual tables FAQ.”
grating Finance and Operations apps into Dataverse. These options have
distinct characteristics, different scenarios, and different pros and cons.

Virtual tables
The virtual table option enables you to connect Dataverse to Finance
and Operations apps entities as virtual tables that offer the same full
CRUD (create, retrieve [or read], update, delete) capabilities as the
entity endpoint in the app. A benefit is that we can access the data in
Finance and Operations apps in a secure and consistent way that looks
and behaves the same as any other table or construct in Dataverse. We
can also use Power Automate to connect to almost anything.

Another benefit is that the data doesn’t have to live in both the
transactional database underneath Finance and Operations apps and

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 466
 Fig.
Dataverse but is still seamlessly available as tables and rows in Dataverse.
16-16

Dual-write Dual-write
The second option is dual-write integration, shown in Figure 16-16.
Model-driven Dual-write also provides synchronous, real-time integration between
apps in Finance and Operations apps and applications in Dataverse.
Dynamics 365

A common scenario for the dual-write option is one in which both

Customer Engagement apps and Finance and Operations apps are
working on data that is fundamentally the same, for example, custom-
er and account, product and item, and projects.

The data is shared to provide seamless integrated experiences, for

Dataverse example, for product mastering, prospect to cash, and project to cash
across systems.

The benefit of using the dual-write approach is that the two applica-
Dynamics 365 tions share the same dataset, so changes in one system are seamlessly
Finance and and instantly reflected in the other one.
Operations Apps
Another benefit is that in some cases the functionality is out of the
box and configurable with minimal effort, but it’s also extendable by a
developer. Additionally, the full set of capabilities of Power Platform is
available for the shared datasets.

For more information about the mapping concepts and setup, read
about dual-write.

Dynamics 365 configurable integrations

The third option is configurable integrations. These provide a starting
point for integration between specific Customer Engagement apps
and Finance and Operations apps through integration artifacts that
enable integration between Customer Engagement Apps in Dataverse
and Finance and Operations apps. Configurable integrations are built
on data entity integration to Dataverse and enable an administrator to
relatively quickly set up the supported predefined integration scenarios.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 467
 The supported scenarios are as follows:
▪ Prospect to cash integration  The Prospect to cash integration
is a process and data integration between Dynamics 365 Sales
and Finance and Operations apps . The process integration en-
ables users to perform sales and marketing activities in Sales, and
other users handle order fulfillment and billing in Finance and
Operations apps. For more information, read Prospect to cash.
▪ Field Service integration  Field Service integration adds inte-
gration Implementation Guide: Success by Design: Integrate with
other solutions points between the Dynamics 365 Field Service
and Finance and Operations apps to enable process integration
for work orders and projects. For more information, read the
Integration with Microsoft Dynamics 365 Field Service overview.
▪ Project Service Automation  Similarly, the configurable integra-
tions provide integration between Finance and Operations apps
and Dynamics 365 Project Service Automation by synchronizing
project essentials such as contracts, tasks, milestones, time, fees,
expense forecasts, and actuals. For more information, see the
Project Service Automation overview. Please note that Dynamics
365 configurable integrations (a.k.a. Data Integrator) is a viable
and supported solution, however Microsoft’s current investments
are focusing on the Dual Write integration technology which
should be always preferred where applicable.

Other integration patterns

The fourth option for cross app integration scenarios is either
Application Connector or point-to-point OData integration patterns
both connecting through data entities. For more information, read
about the Application Connector.

Out of the different options for cross app integration, they are all via-
ble, but Microsoft investment is focused primarily around Dual Write
and Virtual Entities. These are the preferred options.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 468
References Conclusion
REST vs CRUD: Explaining REST & CRUD
Operations Implementing Dynamics 365 apps is often a building block into an
Messaging Integration Patterns organization’s larger solution landscape. In that case, the organiza-
Open Data Protocol (OData)
tion can benefit from automation to gain streamlined and effective
cross-system processes and avoid manual errors.
Consume external web services

Custom service development

Integration is also key in closing the digital feedback loop and making
Data management package REST API
all the organization’s data available, not just for reporting and visibility
Recurring integrations
but also to make it accessible to users in the organization. With Azure,
Electronic messaging
Power Platform, and Dataverse you’re well on your cross-system
Office integration overview journey toward reporting, analytics, and AI-assisted forecasts and
Business events overview predictive analysis so that the right information is readily available for
Embed canvas apps from Power Apps the right people at the right time.
Analytical Workspaces (using Power BI
Embedded) With Microsoft cloud technology and Power Platform, low-code/no-
What’s new and changed in Platform code options, and the citizen developer concept, once the systems are
update 31 for Finance and Operations apps
(January 2020) integrated and the data is available in Dataverse, that journey can start
without requiring massive investments upfront.

In this chapter, we highlighted the path from business goals, high-level

architecture, design, and pattern considerations, the work that should
be done when integrating systems and processes, toward a successful
solution. We introduced integration patterns and discussed the dos and
don’ts and common challenges that could unnecessarily prolong the
start of the journey if not addressed properly, specifically about business
ownership, technology, and implementation project governance.

Finally, we referenced some of the technologies that people are using

across the globe to successfully integrate their Dynamics 365 apps into
the solution landscape.

469
 Choose a design
Checklist Align the designs of each integration with the overall
integration architecture.

Avoid unnecessary replication and transmission of data

into the solution in integration designs.

Define business goals Clearly state the options and benefits of each of the
following: UI, data, process integration, and Dataverse.
Document and define goals and expected benefits of
integrations being implemented in a business-centric way.
Choose a pattern
Align the planned integration’s purpose with short- and
long-term organization goals. Design integrations to favor robust, asynchronous
messaging-based patterns.
Ensure the overview of the integration architecture, sys-
tems, and integration points is clear and understandable. Align patterns used for each integration with expectations
for volumes, frequency, and service protection limitations.
Ensure that stakeholders have a shared understanding of
the purpose and scope of the integrations that are being Set realistic estimates of the operating costs for services,
implemented. platforms, and storage involved and be aware of how
scaling affects them in the future.

Choose a platform
Project governance
Ensure the organization understands the concept of
cloud versus on-premises platforms and the boundary Plan each integration for user and performance testing
between them. under realistic loads, as well as the end-to-end process
leading up to the integration, across the system bound-
Plan to use either an integration middleware or
ary, and after the point of integration.
messaging service.
Plan for testing the end-to-end process patterns used
Ensure the integration architecture, platform, or middle-
for each integration in line with recommendations for
ware supports the expectations for monitoring, audit,
volumes, frequency, and service protection limitations.
notifications, and alerts.
Have change management activities related to integra-
Ensure the integration architecture supports the expect-
tions that reflect and support overall business goals.
ed level of security, availability, and disaster recovery.
Complete the impact analysis on upstream and down-
Ensure all components of the integration architecture
stream processes.
support ALM and version control.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 470
 Case study
Public sector infrastructure
organization learns how to
choose the right solution
for integration
A public sector infrastructure service organization was implementing
the Dynamics 365 Customer Service app and required integration with
a public-facing website, several on-premises apps, and Office 365 ser-
vices such as SharePoint Online. One of the implementation objectives
was to gather insights from the data that was scattered across different
business applications.

The organization was working with an IT vendor that had several years
of experience building complex integrations using technologies such
as IBM MQ and Microsoft BizTalk.

They decided to keep a uniform integration architecture and proceed-

ed with a microservices-based architecture to build small, reusable
components using service endpoints to gain the following benefits:
▪ Faster and independent development, deployment, and release cycles
▪ Simplicity of architecture to allow decoupled and reusable components
▪ Fault isolation

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 471
 The organization chose Azure API Management to abstract their APIs
and implement a secure integration layer.

As the team started building and testing the initial components, they
identified some challenges due to the architecture:
▪ They experienced slow performance with batch-integration sce-
narios because they called the services as they would have in a
point-to-point integration.
▪ They couldn’t use standard functionalities that would have been
available with out-of-the-box approaches such as SharePoint
Online integration with Power Platform.
▪ For an aggregated view, they decided to replicate all data into
the Dynamics 365 Customer Service app, which led to additional
storage costs.
▪ They encountered throttling and API limits issues, which prevented
successful completion of the integration.

At this stage, the organization decided to re-evaluate their standardized

architecture principle and redefine their integration strategy. They divided
their needs into two categories: batch and point-to-point synchroniza-
tions. They also adopted a “Configure-before-you-customize” approach.

With a reasonable mapping of different integration needs to the

patterns, they were able to redefine a suitable integration architecture
for these components.

For point-to-point integrations, the organization used the previously

defined microservices-based architecture and incorporated the neces-
sary changes for aligning with Power Platform API limits.

For batch synchronizations, the organization designed the architecture

to manage the overall ETL processes with the following approach:
▪ They extracted the data from their source application by creating
SQL Server Integration Services packages and storing that data in
a staging database to perform all transformations.
▪ They hosted their staging environment in Azure SQL, within the
same datacenter as the Customer Service app.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 472
 ▪ They sent that data as batch data to the Customer Service app to
reduce the number of individual connections.
▪ They considered API and throttling limits and built retries as part
of the design.

For an aggregated view of the data, they realigned their approach to

use Azure Data Lake, as Microsoft recommended. The out-of-the-box
Export to Azure Data Lake feature required simple configurations to
export data from the Customer Service app into Data Lake and didn’t
involve writing any additional code.

These changes in their overall approach provided significant benefits:

▪ Their batch synchronization performance improved significantly,
and they cut the duration of the daily sync by more than 70 percent.
▪ Their “Configure-before-you-customize” approach let them
choose out-of-the-box options to reduce development and
maintenance costs.
▪ The retry mechanism, combined with monitoring, ensured that
any failure scenarios were well handled and negative impacts on
end users were avoided.

Even though the architecture realignment reset the organization’s

timeline, the benefits outweighed the risk because they were able to
align their design with product features and the recommended prac-
tices. The organization already had a change-management plan in
place to take care of stakeholder communication and alignment, and
to plan trainings for all affected teams in a timely manner.

Implementation Guide: Success by Design: Guide 16, Integrate with other solutions 473
17
Guide
A performing
solution,
beyond
infrastructure
Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 474
Patience is such a waste of time.

Introduction
When a new Dynamics 365 implementation is
delivered, users typically expect an improvement in
system performance.
Good performance is often assumed as a given, a default experience. The
The following are performance factors to
consider when designing solutions: reality is that although Dynamics 365 products are scalable and powerful,
• Solution performance is critical for user various factors are involved in achieving a high-performance solution.
adoption, customer experience, and
project success and necessary to enable
These include defining and agreeing to performance metrics, testing
businesses to achieve their goals. throughout the phases of a project, and taking design and build consid-
• Clear goals and realistic expectations
are vital to developing a solution that erations into account, particularly for customizations and integrations.
performs well.
• Scalable solutions begin with the
correct use of the right software. In this chapter, we explore various aspects of performance and how
• Customizations increase the risk of
they should be addressed throughout the stages of implementation. We
performance issues, but these risks can
be mitigated with the right mindset discuss why performance is directly related to the success of a project and
and planning.
• Performance testing needs to be
why it must be prioritized early. We also discuss how to align expectations
realistic to be meaningful. with business users to enable meaningful performance discussions so that
• Performance issues are complex
to resolve and therefore are better challenges are identified and translated into design requirements.
avoided than fixed.
• Ensuring performance is an iterative
process that involves incremental Finally, we cover performance testing strategies to protect organiza-
improvements throughout the
solution lifecycle.
tions from common performance risks and anti-patterns as well as how
to approach resolving any issues that do occur.

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 475
 Why performance matters
Performance can pose challenges as projects approach go live. These
challenges are typically due to a failure to properly prioritize perfor-
mance in earlier phases. Let’s back up a bit and consider first why good
performance is important.

Let’s consider city planning, specifically, the importance of street de-

sign to ensure livability. Streets enable residents and visitors to get into
and out of a city. A city could plan a basic two-lane road to accomplish
this goal. However, what happens when 1,000 cars are on that two-lane
road at the same time? What happens if an accident occurs? What
happens if you add 10,000 cars? Yes, eventually drivers will get to their
destinations, but it likely won’t be in the desired amount of time. These
are the types of questions to ask when looking at performance.

People expect superior response times and a smooth experience from the
organizations they choose to provide products and services. If customers
don’t get this, they tend to look elsewhere. Let’s go back to our street
design. Because of poor street design, people might decide not to live in
or visit that city. Residents might even elect different leadership because
of their poor experience. The takeaway is that system performance affects
people in numerous ways and can have serious consequences.

Customer experience is key

A poorly performing application solution can definitely have a negative
effect on your organization. Microsoft Dynamics 365 products support a
wide range of business functions related to customer interaction and are
an opportunity to either delight or disappoint customers.

Customers might use integrated applications with a direct depen-

dency on Dynamics 365 apps, for example, through a sales website. In
such a scenario, poor performance can quickly turn into lost sales as
customers abandon the website to shop elsewhere. No one enjoys a
frustrating and slow sales experience, and customers can quickly switch
to a different business to purchase goods and services.

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 476
Often, though, customers interact with systems indirectly through an
agent within an organization, for example, when a customer calls a call
center to discuss a faulty product or delayed order shipment. These
types of interactions can be challenging for a business to resolve in
a positive way and are only made more challenging if an agent can-
not access information in a timely manner. This illustrates again that
system performance is key to ensure a positive customer experience,
regardless of whether the customer interacts with the system directly
or indirectly.

System success
As businesses evolve and optimize, the ambition remains to achieve
more with less. The need to sell more products or service more cus-
tomers is always accompanied by the need to reduce the expenditure
of money, effort, people, and time, at all levels of the organization.

This constant need to achieve more with less creates pressure on em-
ployees to work in the most efficient way possible. Time spent waiting
for a system to perform an action is wasted time, and employees who
rely on these systems to do their jobs quickly realize this.

Poor performance can have significant effects on a project. It’s usually

not well tolerated by business users and can be detrimental in terms of
user adoption and system reputation.

User adoption
User adoption is a critical factor in the success of any software project.
Any business case (and projected return on investment) depends on
the system being used as intended. Poor performance directly drives
user dissatisfaction and can make user adoption incredibly challenging.

Users are keen to adopt systems that increase their productivity, which
essentially means minimizing wasted time. Business users achieve their
goals when solution performance is optimal and as expected. A poorly
performing system wastes users’ time and therefore reduces productivity.

If the impact of poor performance on users’ time is significant enough,

477
 they’ll likely find more efficient ways of working on their own. These
workarounds ultimately serve the interests of specific users rather than
the business and often result in data stored off-system (for example,
in spreadsheets). This might eventually lead to a situation in which the
system no longer serves its purpose, doesn’t deliver a return on invest-
ment, and ultimately fails.

System reputation
Even before go live, performance can help or hinder user adoption.
During the development phase, the application typically is presented
to a set of key users in different areas of the business to collect feed-
back. These users then talk to colleagues about the implementation.
In this way, the reputation of the application spreads throughout the
business long before most users touch the system. Keep in mind that
performance impressions tend to spread quickly. For example, if a
demonstration goes well, a wave of excitement might flow throughout
the company. This positivity can help increase user adoption because
of the anticipated improvement in productivity.

Performance issues discovered in the buildup to go live or shortly af-

terward cause a great deal of stress for the project team, which usually
works long hours to ensure a timely delivery of a functional system.
The realization that the system doesn’t perform well can quickly force a
change in priorities and derail progress. These issues are often visible at
the most senior levels and tend to escalate quickly.

But it’s in the cloud

Having established that performance matters, surely you can just deploy
the fastest computers available to run the solution, right? Well, not quite.

Dynamics 365 is a cloud-based software as a service (SaaS) solution.

Cloud solutions entail different challenges from on-premises deploy-
ments. For example, several years ago, Dynamics 365 deployments
were commonly hosted on infrastructure physically located close to or
even in the same office or warehouse as users. Today, single deploy-
ments are used by users across the world on a range of devices with
different network infrastructures. Project teams therefore need to build

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 478
solutions that accommodate users who have a variety of hardware,
increased network latency, and a range of network quality.

The SaaS nature of an application means that the infrastructure on

which the solution runs is housed and maintained by Microsoft. Other
than a few predefined configuration options, the infrastructure is
beyond the scope of a customer’s control, which offers benefits in
setup cost and ongoing maintenance responsibilities but does limit the
options to change performance.

Having said that, the performance of an application extends beyond

the infrastructure. Although a correlation exists between the perfor-
mance of software and the quality of its underlying infrastructure,
the on-premises mindset of fixing software problems with additional
hardware is expensive and usually only a temporary fix.

Using our street design example again, adding a privately operated

toll road or highway doesn’t necessarily mean drivers will arrive at their
destinations any faster.

We look at more examples later in the chapter, but for now it’s import-
ant to be clear that most performance issues are best solved by correct
implementation decisions and not by adding hardware. Moreover, it’s
crucial to acknowledge that performance is not guaranteed simply
because the software is running in the cloud. It’s still the responsibility
of the project team to deliver a well-performing solution.

As there are a number of factors which contribute to the performance

of a system, it’s important to understand the responsibilities of the
various parties to ensure the solution as a whole performs in an ac-
ceptable manner (Figure 17-1). Performance is a shared responsibility,
typically between Microsoft and our customers and partners.

479
 Fig.
17-1

Customer Partner Microsoft

Owns the solution Builds the solution Manages the platform

▪ Describe clearly the business goals ▪ Understand and document the ▪ Ensures the core platform
and predict transactional volumes customer processes components have sufficient
▪ Complete the usage profile based ▪ Capture the performance non- resources to deliver acceptable
on the licenses purchased functional requirements performance, when used
▪ Whenever applicable, optimize the ▪ Build a solution that is optimized/ reasonably
business operations to maximize fit-for-purpose that doesn’t disrupt ▪ Deliver solutions to performance
the efficiency of the solution the scalability of the Dynamics 365 issues within Microsoft’s scope (e.g.
▪ Plan the performance testing and product index, product hotfix, infrastructure)
allocate the relevant resources ▪ Provide the expertise to educate ▪ Support investigations into
the customer on performance performance issues outside
testing of Microsoft’s scope (e.g.
▪ Support the customer for the customization fixes)
performance testing execution

Prioritize performance
Given that performance is so important to users, customers, and ulti-
mately the success of the overall system, let’s look at how performance
relates to project delivery.

Performance doesn’t happen unless

you make it happen
Performance issues in Dynamics 365 projects are commonly analyzed,
understood, and fixed at a solution level, often against tight dead-
lines and high visibility. It’s fair to say that the most common cause of
unexpected performance issues is a failure to properly prioritize per-
formance from the beginning of a project.

Performance in any software project is not guaranteed. It requires

careful planning, expectation management, open communication, and
dedicated time and resources. This is especially true for projects that
feature integrations, data migrations, and extensions on top of the
base product. These are all common in Dynamics 365 implementations.

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 480
Anything added to the core Dynamics 365 app has the potential to
negatively impact performance, so the impact needs to be understood
and managed to ensure performance remains acceptable.

Performance influences decisions in many areas of the project, including

the following:

Data strategy
Is the volume of data stored likely to cause performance issues for users?

Integration strategy
▪ Are real-time integrations feasible given the performance
expectations of the users?
▪ Can overnight batch integrations complete within a given timeframe?

Data modeling
Do we need to denormalize for performance reasons?

Security modeling
▪ Will our security model work at scale?
▪ Are there bottlenecks?

Environment strategy
▪ Is a performance test environment available?
▪ Is our performance test environment representative of
production?
▪ Have we budgeted for performance testing?

Design and implementation of customizations

▪ Are the developers building to meet specific performance goals?
▪ Do user expectations align with what’s technically feasible?
▪ Do we need to use patterns (such as asynchronous or synchro-
nous) or batch jobs (such as background processes)?

Testing design and approach

▪ What’s acceptable? What’s considered a pass or fail?
▪ Is our testing representative of user behavior?
▪ Are we testing based only on today's business requirements or

481
 keeping future expansions in mind?

User acceptance and adoption

▪ Is performance measurable?
▪ Are user expectations realistic?

These aren’t the types of questions the delivery team should be asking
when a performance issue surfaces, especially when approaching the
go-live deadline. A successful project will have answers to these ques-
tions early on. At the least, the delivery team should identify risks and
seek possible resolutions in the early stages of delivery. This might lead
to proof-of-concept work to test performance.

Let’s put this into the context of our street design scenario and consid-
er the questions that need to be asked and answered to maximize the
design. For instance, how many residents currently live in the city? How
much traffic can we expect to be on the roads? What’s the projected
population growth and how long will the design support it? Will traffic
lights be installed? If so, how many and how will that affect traffic?
What’s the speed limit and are there risks associated with that limit?
Each of these questions helps determine the best street design before
we even start the project.

Resources
Considering performance from the early stages also ensures that the
correct expectations are set in terms of time, money, effort, and peo-
ple. For example, for performance testing, a dedicated performance
test environment is needed as well as the people to do the testing.
Business stakeholders might need additional time with the delivery
team to understand, agree with, and document performance require-
ments. It might even be necessary to allocate more development time
and resources for code optimization.

The most successful projects from a performance standpoint dedicate

workers to performance-related activities. Performance testing is a
subject on its own, and expertise is required to design test cases and
execute test runs. Furthermore, a single owner within the project
should be identified to drive performance planning and remediations

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 482
 across the development teams during all phases of the project.

Prevention is better than cure

A common approach with Dynamics 365 projects is to implement the
system, get it to a point where it’s functionally correct, and only then
deal with the inevitable performance challenges (see Figure 17-2).
Although this approach might seem logical under the pressure of
delivery, the inevitable performance issues place additional stress on
the team during the most difficult parts of the project and add work
during the buildup to go live.
Fig.
17-2 Go live

Initiate Implement Prepare Operate

Identify and fix performance issues in these phases Do not wait to fix issues in these phases

Dealing with performance issues after functional testing is completed

can create additional project risk because the changes needed to
resolve them are typically significant and therefore come with a high
risk of introducing bugs. This essentially invalidates any prior functional
testing. A regression testing cycle is then required to ensure perfor-
mance fixes haven’t broken the previously tested functionality. This
wastes time at a critical stage in the project.

User confidence
Performance is always linked to perception. It’s important to be aware
of user feedback during the implementation of the project because
it can affect users’ engagement during testing and ultimately help or
hinder user adoption at launch. Projects that prioritize performance
early on tend to present better-performing solutions during imple-
mentation. This early planning helps reassure users that they’ll receive
a solution that enables them to become more productive, and this
leads to better engagement and adoption overall.

Why not prioritize performance?

With all this in mind, it’s perhaps surprising that many Dynamics 365

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 483
implementations don’t prioritize performance. There are several rea-
sons for this. In some cases, people lack experience with the specific
products. Performance issues can be complex, and it’s often difficult to
foresee them without extensive knowledge of the platforms and asso-
ciated patterns and anti-patterns gained over years of experience.

Sometimes, however, projects intentionally don’t include performance

testing in an attempt to save on budget. For some projects, performance
activities are deprioritized because of timing constraints. However,
these are shortsighted approaches because they’re an opportunity for
performance issues to surface and risk the success of the project. The
best way to avoid performance problems is to be proactive and seek out
and address risks early to minimize their impact. It’s a more productive
approach than reactively responding to problems as they occur.

Establish requirements
Acceptable performance is the goal of every project, but the definition
of “acceptable” is often vague, if defined at all. To successfully deliver
acceptable performance, we need to be clear on what that means and
then be able to track progress against our performance goals.

When we look at the requirements for our street design scenario, we

know that drivers want to go from point A to point B quickly and safely.
However, each driver might have a different idea of what quickly and
safely means. To get somewhere safely, a driver might need to drive
slower to handle unexpected situations. But then some drivers might
become upset because it takes too long to reach their destination.
If you create a goal for drivers to get from point A to point B within
X amount of time and that time increases, it might indicate that you
need to change the design.

In the next section, we discuss how to define performance goals that

are achievable and acceptable to users.

484
 Why do we need performance
requirements?
If you ask a user for their performance requirements, you’ll likely get a
response that the system needs to be fast and not much else. However,
“fast” is an ambiguous term; people have different expectations of
what that looks like, which makes understanding whether you’ve
achieved acceptable performance nearly impossible.

A better approach is to establish a common understanding across the

implementation team and stakeholders on performance, which en-
ables us to:
▪ Align expectations of the definition of “acceptable.”
▪ Understand the business impact of not achieving acceptable.
▪ Understand constraints during the design process.
▪ Provide clear instructions to developers during implementation.
▪ Test in a meaningful way with behavior models representative of
expected usage.
▪ Determine whether acceptable has been achieved.
▪ Assess gaps between what’s been achieved and what’s acceptable.

This approach is the same for other system requirements. It’s vital that
performance be considered like other requirements gathered in the
initial stages of implementation.

More specifically for performance, it’s also important to know when

to stop optimizing. For example, developers might apply perfor-
mance tweaks to a piece of code, optimizing with no goal in mind.
Any optimization of tested code comes with some risk of regression
issues; therefore, this practice should be performed as infrequently as
possible. However, without a clear understanding of when enough is
enough, it’s difficult to gauge what level of optimization is sufficient
and when further optimization is unnecessary.

Work with users

Any performance-related discussion should include representatives of
the system users. They often give meaningful insight into what is and

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 485
 Fig.
isn’t acceptable. Decisions made without input from users risk rejection
17-3
in the later stages of the project, which is what we want to avoid.

This doesn’t mean users are solely responsible for deciding perfor-
mance requirements. It’s important to be clear that performance
comes with an associated cost, and business stakeholders need to
assess requests coming from users within the context of the wider
“I need to load this record
quickly for a customer on project. Aggressive performance requirements might be achievable,
the phone; otherwise, I but they require additional development, testing effort, and people.
might lose the sale.”
With this in mind, it’s important to understand the underlying need
for each performance requirement and for business stakeholders to
be prepared to consider a compromise where it makes sense to do
so. Performance for the sake of performance is expensive and unnec-
essary. Communicate this to users and take a pragmatic approach to
focus on what specific performance requirements are important.
“I need to be able to load
the products into the Identify performance-critical tasks
delivery truck to meet my
shipment schedule.” Users will use the system under development to perform their own
set of tasks. Some of these tasks will be more performance critical
than others. For example, consider a call center agent attempting to
access customer data at the start of a telephone call with a customer.
It’s critical that the agent is able to access the data, and every second
counts. Time spent waiting for a form to load is a waste of time for the
agent and the customer.

“I have to be able to do In contrast, a user preparing month- or year-end financial reports

this within a few minutes;
we have a service-level might expect to wait a little while for actions to complete. Performance
agreement to meet.” is still important, but for activities like this that occur infrequently, the
business case for investing time, effort, and money to improve perfor-
mance is more difficult to justify.

For example, business users might offer the justifications shown in

Figure 17-3 when asking for a performance improvement.

Spend time with users to understand the activities for which per-
“This overnight process
needs to happen within the
formance plays a critical role. Agree on these areas with project
time window; otherwise, the stakeholders and then focus performance-related work on these
users won’t be able to work.” activities to maximize the value of the efforts. Consider performance

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 486
 testing for each system area, including the following:
▪ Functional processes
▪ Background operations (for example, batch and workflows)
▪ Integrations
▪ Data migration
▪ Reporting and analytics

Understand user expectations

For each critical task you identify, have discussions with users about
what they deem to be reasonable performance. It’s important to
discuss each task separately and focus on the specific task when dis-
cussing acceptability. This leads to different requirements for different
tasks, which is a more granular, actionable, and sensible outcome than
having broad, systemwide requirements.

Also consider that a user with exclusive access to a system is likely to

have a different experience than someone who uses the system only
during peak hours, along with thousands of other concurrent users,
or while an intensive background process runs. We recommend that
you map out predicted system usage throughout the day, week, and
month to understand the volumes of users, data, and actions in these
time periods to better locate any potential performance issues.

The usual measurement of performance is time taken, so requirements

should be shaped around what is an acceptable time to perform a spe-
cific action, in a way that can be measured. However, it’s often easier to
ask users, “What does unacceptable performance look like?” and work
backwards from there. This approach tends to lead to more realistic
requirements, although they might be unnecessarily aggressive. Take
the time to understand the business impact and risk of unacceptable
performance in these scenarios.

Anticipate growth
When discussing performance requirements, consider the roadmap of
the business as well as design requirements for an increase in demand
on the system. Data and user volumes play an important part in how a
system performs, so it’s important to anticipate any growth expected
in the near future and design for that rather than focus on the current

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 487
requirements. Along with future growth, also plan for seasonality in the
system load, for example, during the end of the year.

Document requirements
It’s crucial that performance requirements be documented (the same
as for other system requirements). Documenting requirements pro-
vides visibility to all parties about the expectations of the software and
provides clear goals for the implementation team. Additionally, any
performance risks identified during discussions with users should be
documented in the project risk register to ensure they’re tracked and
mitigated as much as possible.

Assess feasibility
The implementation team should review performance requirements
along with other system requirements as early as possible to assess the
impact to the organization. Some performance requirements might
have a significant impact on design decisions, so you want to be aware
of these as soon as possible.

Occasionally there might be doubt about whether a particular require-

ment is even achievable with the available software. In this situation,
it’s advisable to run a small proof-of-concept pilot project to under-
stand feasibility as early as possible, thereby minimizing the risk of
accepting unachievable requirements. Discuss requirements deemed
unachievable with the business, it’s generally a more positive discus-
sion during the early stages of the project than during user acceptance
testing (UAT) or immediately before going live.

With clear performance criteria established, accepted, and docu-

mented, the project team can move confidently on to implementing a
system that will perform within expectations.

Design for performance

A common cause of performance challenges in Dynamics 365 projects
involves incorrect use of the applications. In other words, uses that aren’t

488
 aligned with the intentions of the software or are due to poor solution
design. Poor decision making during the design phases of the project
can open the door to performance challenges further into the project.

Looking at this from the perspective of our street design discussion, there
are different ways to tackle a situation like traffic. We could design our
roads to be 10 lanes; that would handle a lot of traffic. But this creates
other complications. Do we have land that will support that infrastructure?
What are the relative costs associated with that design? How easy will it
be for residents to cross the street? How long will a light need to be red
for them to cross it? Will the street need additional lighting?

In the next section, we discuss the architectural and design consid-

erations for the project design phases that ensure success from a
performance standpoint.

Use the right tool for the job

Dynamics 365 products are flexible and can be changed to suit the
needs of many businesses. However, exercise restraint when consid-
ering incorporating this flexibility into your design. Just because we
can adapt the products to achieve something functionally, it doesn’t
guarantee they will do them well.

For example, the xRM concept, which became popular around the
release of Dynamics CRM 2011, spawned many systems to manage any
type of relationship in the CRM product. The ease of development for
basic data access, including a built-in user interface and security model,
combined with its rich extensibility features, made it a popular choice
to begin system development. Although this proved successful in
many situations, many projects ran into trouble because they used the
product in a way unsuited to its strengths. Dynamics 365 products are
designed for specific use within a business context. They’re designed
and optimized for users to access master and transactional business
data, not for keeping high-volume historical transactions.

Additionally, keep the Dynamics 365 roadmap in mind. It’s worthwhile

to explore newer features because they often bring performance

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 489
enhancements compared to older approaches.

Many areas of this book discuss what to consider when making system
design decisions. Often the consequence of making the wrong deci-
sions during these stages are performance issues. It’s important that
use of Dynamics 365 products is in line with their intended purpose.

Design for scalability

Whether you design a solution for the first time or redesign one due to
issues, keep performance in mind from the beginning. The following
are recommendations for building a highly scalable solution.
▪ Design for parallelism and multi-threading for high-volume scenarios.
▪ Use set-based operations, RetrieveMultiple,
insert_recordset, and update_recordset, as appropriate.
▪ Build multiple lightweight processes connected to each other.
▪ Design shorter transactions to reduce lock escalations and increase
user concurrency.

Design for responsiveness

Response time is important for the success of your solution. The
following are recommendations for building a highly scalable solution:
▪ Consider whether to design for synchronous, asynchronous, or
background (batch) processing.
▪ Consider data growth over time because rapid data growth
requires a lot of tuning to keep the same response rate.
▪ Design for aggressive caching in Finance and Operations apps as
appropriate.

It’s important to note that users typically assess whether performance

is in line with their expectations, not within an arbitrary period. With
this in mind, we can consider performance solutions that don’t actually
involve speeding anything up.

For example, choosing to execute a task asynchronously doesn’t re-

duce the time to execute the task, but it does ensure the user interface
doesn’t prevent the user from working. As long as the user doesn’t
need the result of the task in real time, this could be an acceptable
performance solution.

490
 Data migrations and integrations
Many performance issues related to data migration and integration
happen due to unreasonable expectations of the platform, specifically
around data volumes. A good data strategy often works around this by
ensuring that Dynamics 365 products contain only the data required
for users to complete their day-to-day operations.

When planning a data migration, it’s important to understand the

business requirements for the data being migrated. Implementation
teams sometimes decide to migrate far more data than required and
then run into performance issues due to the size of the data set. For
example, importing the last 10 years of unconverted leads into a busi-
ness’s sales system will likely take a substantial amount of time. The fix
here is unlikely to be to speed anything up; it’s more likely to involve
questioning the value of 10 years of data, much of which is probably
stale and unlikely to provide value.

In the case of high-volume integrations, consider when and how often

these will run. Heavy processing can place a strain on the application,
and if this happens at the same time that users are interacting with it,
issues are likely to arise. This is particularly important to consider for
global businesses, whose systems might be in almost continuous use.

It’s also important to be aware of platform limitations when designing

any high–data volume or processor-intensive operations, which are
common in data migration and integration scenarios. For example,
Customer Engagement apps has documented throttling limits that need
to be accounted for within designs. Working within these limits helps
ensure that the performance of the underlying platform remains ac-
ceptable. As another example, the Finance and Operations apps OData
endpoint has a page size limit of 10,000 records and isn’t designed for
high-volume integrations. Other approaches, such as Data Management
Framework (DMF), are better suited for these types of integrations.

Chapter 16, “Integrate with other solutions,” discusses how to design

integrations and integration patterns and anti-patterns in Dynamics 365.

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 491
 Prepare for implementation
In this section, we discuss the design decisions the team should make
before proceeding with the implementation.

Environment planning
Chapter 9, “Environment strategy,” discusses environment planning in
detail. From a performance perspective, consider the following as the
team moves towards implementation:
▪ Performance testing typically occupies an environment for a signifi-
cant amount of time, so a separate environment is usually advisable.
▪ Latency adds overhead to every operation. Minimize overhead as
much as possible by ensuring that applications are located as close
to each other as possible.
▪ Choose a performance test environment that’s representative of
the production environment whenever possible. For example, for
Finance and Operations apps, the implementation team should
recommend the appropriate environment tier based on expected
load and volume.

Keep in mind that if performance issues arise within scaled-down

Proof-of-concept
development development or test environments, it’s inaccurate to assume perfor-
Teams sometimes run an initial proof-of- mance will improve in an environment with more resources. Without
concept project before implementing
the main project to prove out ideas and
understanding the root cause of a performance issue, it’s impossible
gain understanding of product suitability. to determine whether additional infrastructure will help at all, so it’s
Custom functionality built during proof-
of-concept projects is often carried over crucial to understand and address performance issues early in the de-
into the main project.
velopment process rather than assume they’ll go away in production.
Although this approach can significantly
accelerate development during the
main project, it’s worth bearing in mind
User personas
that there’s often little in the way of The organization should have a sufficient number of user licenses to be
developmental governance for proof-
of-concept projects, which are usually able to realistically model the anticipated behavior during performance
undertaken without any performance testing. Teams also need an understanding of the user personas for
requirements or considerations in place.
These projects can therefore become testing expected usage. Keep the following in mind for user personas:
a source of performance problems
▪ User locations
unless the team takes the time to
review the outcomes along with any ▪ User security configurations
new requirements prior to further
implementation.
▪ Data typically associated with each type of user
▪ Expected number of concurrent users divided by persona

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 492
Customization and
performance
The extensibility of Dynamics 365 applications provides the powerful
ability to tailor software to meet individual business needs but also
introduces risk, performance issues commonly involve product cus-
tomizations and integrations, particularly those involving code. This
section discusses these issues and provides guidance on how to avoid
common problems.

Just write fast code?

Many performance issues are due to custom code, and the fix inevitably
involves reworking this code to remove bottlenecks. Project teams often
question why such code wasn’t written correctly to begin with, but the
truth is that software development is difficult. Developers regularly wres-
tle with complex and abstract processes based on vague and changing
requirements along with competing priorities against tight deadlines.

The most productive developers tend to focus on what’s necessary

and disregard anything superfluous to requirements. Keep in mind
that if a requirement is missing key information, the developer won’t
be aware of this and will proceed with implementation based on the
given information. It’s common for requirements to lack performance
criteria, which isn’t typically top of mind for the developer or part of
sign-off criteria.

It’s also uncommon for a developer to deeply understand the business

context of the system for which they’re developing code. A developer
likely might not have experienced the time-pressured environment
of a call center or order fulfilment center and therefore won’t be able
to judge the need for performance. Therefore, specific performance
requirements should be mapped through to functional requirements
where appropriate to ensure that developers have visibility to perfor-
mance constraints on the functionality for which they’re responsible.

493
 Bad design leads to bad code
Performance issues related to poorly performing code often aren’t
due to the code itself but rather to a flaw in the design decisions. This
can result in situations in which developers are unable to write code
to perform an operation efficiently, even when carefully following
best practices, because of the constraints placed on the solution by its
design.

Retrofitted performance
Sometimes developers focus on getting code working correctly before
working quickly. Although this can be a reasonable approach, the pres-
sure of deadlines often means that the optimization planned for a later
date doesn’t get done, leading to technical debt and a need for re-
work. A better approach is to be clear on any performance constraints
from the beginning and then implement the solution accordingly.

Connected components
Performance issues can occur even when developers write fast,
efficient code. This is because in larger projects involving several devel-
opers, the isolated nature of development might result in performance
issues not being discovered until components are put together as part
of a larger process. This risk should be identified during design activi-
ties, and it can be mitigated using a code review process during which
team members can consider the impact of the implemented pieces of
code running together.

Requirement evolution
A change in functional requirements can be another reason for cus-
tomization-related performance problems. A developer decides how
to implement code based on the given requirement at a point in time.
Some of the content in this section
A change in requirements might invalidate some or all of these deci-
discusses technical concepts. For those
less interested in the technical aspects, sions and cause the implementation to become unsuitable.
note that the details aren’t essential to
understand. Also note that free tools
are available for less technical people
who want to learn about these concepts,
including the Customer Engagement apps
Common mistakes
Solution Checker and the Finance and
Code can become suboptimal from a performance standpoint for
Operations apps Customization Analysis
Report (CAR). a number of reasons, and specific guidance is beyond the scope of
this chapter. However, the following factors are often involved in

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 494
 performance challenges, so we recommend that you understand and
avoid them during implementation.

Chatty code
One of the most common causes of code-related performance issues is
excessive round trips. Whether between the client and server or between
the application and database, an excessive number of requests for an
operation can really slow it down. Every request carries latency and pro-
cessing time overhead, and it’s important to keep these to a minimum.

Round trip issues are sometimes due to poor control of variables; a process
might be so complex that it’s easier for a developer to retrieve the same
data multiple times than structure the code to minimize the calls. We
recommend that developers avoid this practice and that the implemen-
tation team identifies such problems during the code review process.

Round trip issues are also sometimes due to queries being executed
within a loop, or worse, a nested loop. The code structure works for
the developer, but the parameters of the loop can add a significant
number of calls into the process, which in turn results in a significant

Fig.
performance issue.
17-4

Consider the pseudocode exam-

foreach (var a in collection1)
ple in Figure 17-4 as triggered
{
manually by a button click in the
foreach (var b in collection2)
user interface.
{
ExecuteRequest ();
}
If ExecuteRequest() sends a
} single request to the server, the
total number of requests will be
Size: Size: the product of the sizes of
Total requests Total execution time
collection1 collection2
collection1 and collection2.
2 2 4 0.2 seconds
Assume each request takes an
5 5 25 1.25 seconds average of 50 milliseconds.

10 10 100 5 seconds
It’s common for developers to
50 50 2,500 2 minutes, 5 seconds write logic that iterates across

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 495
collections of data that are dynamic in size, due to the data-driven na-
ture of code within Dynamics 365 implementations. It’s also common
for developers to work with low volumes of data within development
environments. However, that means that these types of issues often
aren’t identified until the latter project stages, which include meaning-
ful data volumes. Low data collection can be avoided by prioritizing
minimal data retrieval during code design, retrieving each piece of
data only once, and identifying any deviations from these steps as part
of a code review process.

Retrieving too much data

Another common performance-related issue is retrieving more data
than necessary, often in columns. For example, the practice of selecting
all the columns in Dynamics 365 apps to avoid specifying individual
columns can cause performance issues. The number of joins executed on
the database server to provide lookup names and option set values can
be a significant overhead, particularly when querying large volumes.

The same applies for rows. Be mindful of the amount of data being
retrieved because it takes time to query, download, and process the
volume of records even before the code is able to interact with them.

Broadly speaking, only the data required by a process should be retrieved;

anything more is simply a waste of resources.

Intercept synchronous events

Many events in Dynamics 365 products can be customized, but if
the customization executes synchronously, it adds overhead directly
to the user’s experience. For example, if a customization placed on
RetrieveMultiple in Customer Engagement apps executes, or an
event insert or pre-insert in Finance and Operations apps executes, the
user has to wait for the customization to execute while they watch a
spinning icon. This approach isn’t recommended, but when it’s neces-
sary, exercise extreme caution to ensure minimal performance impact.

Unintended execution
Performance issues sometimes happen because customizations are
executed accidentally or multiple times in error, for example, duplicate

496
 plug-in steps or duplicate method calls. Project teams should ensure
that developers are aware of exactly how their customizations are trig-
gered and mindful of circumstances that might inadvertently trigger
their code. Background processes or batch jobs recurrence should be
set according to business needs.

Performance testing
approach
The project team and users need to be confident that requirements
identified earlier in the project are achieved when the system is
implemented. A performance testing strategy ensures that system
performance is measurable and provides a clear indication of whether
performance is acceptable.

What makes a performance test good?

A common approach to performance testing is to try to push the
system to the breaking point to learn whether the system can handle
the demand. However, this approach typically provides little value,
particularly in Dynamics 365 apps. The best approach is to model or
simulate expected system usage including peak load and day-in-the-
life business scenarios to make sure the solution is capable of achieving
defined performance goals.

Realistic
Being realistic means understanding the quantity and personas of
users using the system at a given time and defining day-in-the-life
activity profiles for the personas to understand the actions they’ll
perform. If the performance test incorrectly assumes that all the users
will run the most complex processes in the system concurrently, the
projected demand placed on the system will be far higher than reality.
Strive to model the users in an accurate way to get the most meaning-
ful results from the test.

It’s important to understand the difference between testing the

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 497
functionality and testing as a user would use the system. Functional
testing is often achieved by invoking specific logic and measuring the
outcome, for example, an API call, whereas user testing is testing that
models the actual behavior of a user. Testers often create a suite of
functional tests pointed at a web service, execute them with a large
number of concurrent users, and call that a performance test; but this
provides misleading results. Typically, this approach pushes certain
functionality much more than the usage expected once the features
are live (creating false performance issues), and other functionality can
be bypassed entirely (for example, the user interface). The result is that
genuine performance issues can slip through testing unidentified.

Keep in mind that user interaction with the application is fairly slow.
For a given process, users don’t typically click buttons as quickly as they
can; they take time to think between actions, and this can be incorpo-
rated into a performance test as a think-time variable. This can vary
from user to user, but an average figure is sufficient to model behavior.
The key point here is to develop a performance test that represents a
number of users working concurrently and place a realistic amount of
demand on the system.

Developing a reliable performance test also includes realistically

modeling client infrastructure. For example, there’s little value in
performance testing your Dynamics 365 Field Service or Warehousing
mobile application from the office when users will likely use the appli-
cation in areas of low mobile reception.

The configuration of the system should also be realistic and based on

the expected infrastructure and configuration of the production envi-
ronment, use real security configurations for the test users, and work
with volumes of data anticipated in production.

Isolated
A dedicated performance testing environment is generally recom-
mended for two main reasons.

Performance test results are meaningful only if testers are aware of the
activities occurring in the system during the test. A performance test is

498
 worthless if an unaccounted-for process places additional demand on
the system during test execution.

Additionally, it’s common for performance testing to occur around the

same time as UAT. Performance testing can take a substantial amount
of time to execute and push the system to its limits, thereby disrupting
any manual user testing activities. A dedicated performance testing
environment ensures that performance testing can happen in parallel
with other testing.

Business data
Key to good performance testing is using business data such as setups,
configurations, masters, and transactions. It’s recommended to use
the configurations and data to be migrated that will ultimately go live
in production. Additionally, all the data preparation activities must be
ready in advance, for example, data from 100,000 customers or sales
orders should be available via an import file.

Repeatable with low effort

Ideally, we execute a performance test once and use the results to
ensure that the criteria specified in the requirements are achieved.
However, performance testing is rarely a one-time activity, often,
it’s an iterative process. This happens for various reasons; for exam-
ple, functional changes might occur during UAT that risk a change
in performance. Similarly, if performance testing demonstrates an
unacceptable level, a retest of optimizations will be needed once the
solution is implemented.

If a low-effort performance testing approach isn’t available, it can be

tempting to skip performance test iterations, which makes it difficult to
assess the impact of any changes made between runs. Conversely, a fully
automated process, potentially including environment provisioning and
data setup, can be executed frequently with minimal effort, providing a
clear picture of variations in performance in response to specific changes.

Functionally correct
A performance test result is meaningful only if the system functions
correctly. It’s tempting to focus on the performance metrics of

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 499
 successful tests. However, if errors occur, they should be corrected and
the test should be executed again before any analysis is performed on
the results. Deviations in behavior between test runs can significantly
skew a performance test result and make any comparison meaningless.

Document results
The output of performance testing activities should be documented
clearly so that interpretation is straightforward. The results should
be mappable to performance testing criteria and enable the team to
quickly assess whether the requirements were achieved or whether
there are gaps between requirements and results. For example, page
load times can be captured for certain activities and compared to
acceptable page load time requirements agreed to by the business.
It should also be possible to identify when the performance test was
executed and against which version of code if applicable.

Finally, it’s important that the information in the documented report of

results remains consistent with the defined template; the team might need
to compare various performance test run results as changes are made.

Plan performance tests

Document your performance testing strategy and include the following:
▪ Performance testing scenarios including key user activities, back-
ground (batch) processes, integrations, and reports
▪ Normal and peak data volumes for each scenario
▪ Ideal target and maximum acceptable target response time for
each scenario
▪ Day-in-the-life workload to simulate in the most realistic manner,
especially at peak hour
▪ Data preparation activities for each scenario, for example, existing
master or transaction data in the system or creation of file to import
▪ Environment for performance testing
▪ Iterations of performance testing (plan for two or more rounds)
▪ Performance testing tools
▪ Test results documentation tool
▪ Quantity and profile of concurrent users including security
configuration

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 500
 ▪ Guidelines and workflow for investigating and fixing performance
issues
▪ Risks such as delays and budget related to performance testing

We recommend investing the time and effort to automate this process

as much as possible. For example, with relatively low effort, you can set
up performance tests to run through Azure DevOps release pipelines,
providing a way to execute test runs with a single button click or auto-
matically after a deployment. You can even run them overnight.

Network conditions during testing

Be aware of network conditions during testing. Performance tests are
often conducted in isolated environments with the best network condi-
tions, so they usually show favorable results. However, when the solution
is moved to real-world use, performance issues might arise due to actual
network conditions and the solution being implemented at scale.

To benchmark what’s feasible for the solution when all network conditions
are favorable, ensure that the performance testing environment is located
in the same Azure region in which you plan to deploy production for
business users. As mentioned earlier, latency adds overhead to every oper-
ation, and minimizing testing latency can help identify potential location
network issues if the test results vary with the location testing.

Address
performance issues
Analyzing performance problems is complex and situation specific and is
beyond the scope of this book. However, let’s discuss some broad consid-
erations that might help you structure an approach to resolving issues.

Deal in facts
When faced with a performance problem, implementation teams
often begin to frantically point fingers at certain parts of the system

501
 and develop ideas to find the magical quick fix to solve the problem.
Unfortunately, this approach often causes more problems than it
solves because teams make significant changes based on instinct, often
degrading performance further or causing regression bugs as a result.

It’s crucial to understand the root cause of a performance issue in detail

before making any changes. This can be a time-consuming process
depending on the complexity of the issue, but it’s necessary to ensure
that any optimizations made are meaningful. Ideally, the team will
understand the issue in enough detail to have a reasonably accurate ex-
pectation of the differences in performance prior to testing the change.

Expectations
Keep in mind that performance issues are rarely due to a single issue.
Suboptimal performance is most commonly the result of a number of
factors working together, so a single fix is typically unrealistic. Generally,
performance is an iterative process of applying incremental improvements.

In addition, as discussed throughout this chapter, a performance

issue in a business’s Dynamics 365 solution doesn’t necessarily imply
a performance issue in the Dynamics 365 application or its underlying
infrastructure. Microsoft support teams are available to help with
performance issues, but keep in mind that when a performance issue is
isolated to a single environment, investigations typically start by ex-
amining anything unique in that environment, such as customizations,
rather than the application code or infrastructure.

Knowledge is power
Performance issues can be complex and difficult to resolve, so it’s vital
that the implementation team has sufficient knowledge to be able to
ask the right questions and analyze issues meaningfully. The imple-
mentation team is often able to assist with performance issues, but
issues can surface after go live and expiration of any warranty period.
It’s therefore crucial to transfer knowledge to allow business as usual
(BAU) teams to resolve performance issues.

Identify problematic customizations

To identify the source of a performance issue, we recommend that you

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 502
 first assess whether the issue occurs in a specific customization. Many
Dynamics 365 customization options can be enabled and disabled,
and strategically disabling customizations often helps identify whether
the issue is customization related, and if so, which customization is the
source of the issue.

Proceed with caution

When you identify the root cause of an issue, it’s important to take a
risk-averse approach to optimization. Performance tweaks can involve
complex code changes and have significant risk of regression issues, so
it’s crucial to proceed with caution.

Low-hanging fruit
It’s advisable to identify smaller opportunities for performance gains,
rather than consider reworking large and complex areas of the system.
For example, for a poorly performing piece of code, there are usually
several options for optimizations with varying risk of causing issues and
varying performance gains. In some situations, it might be advisable to
make a number of low-risk tweaks; in other situations, it might be better
to make a more complex change and manage the risk. Which approach
to take is entirely situation dependent, but it’s important that the team
agrees on the approach and that the changes are managed carefully.

One change at a time

It’s crucial to assess the impact of a change as soon as it’s made, if
multiple optimizations are made at the same time, it’s impossible to
understand the impact of each change after a performance test run.
For example, if two changes are made and one increases the perfor-
mance by 10 percent and the other decreases it by 10 percent, the net
result will be no performance change. By testing the same changes
individually, the team can use the performance test results to reject
one change and accept the other one.

Workshop strategy
FastTrack runs a solution performance workshop focused on solution

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 503
design that covers the impact of additional configuration and cus-
tomization on the overall performance and end-user experience. The
workshop emphasizes the importance of performance prioritization,
goals, and testing during the stages of a project.

Workshop scope
The workshop includes the following topics that address how to
incorporate performance activities into the overall delivery plan and
allocate sufficient performance expert resources to the project:
▪ Data volumes  Projected workload and integration volumes to
ensure expectations are within limits and aligned with intended
product usage
▪ Geolocation strategy  Physical locations of users and servers to
identify any network-related challenges
▪ Key business scenarios  Main areas of the business for which
performance is particularly important
▪ Extension performance  Planned customizations to understand
how implementation is aligned with recommended practices
▪ User experience performance  Modifications to the user expe-
rience in conjunction with best practices
▪ Performance testing  Performance-related goals and testing
strategy to ensure performance is measured

Timing
We recommend that you conduct the performance testing strategy
workshop before solution design or as soon after as the team is able
to provide detailed information about performance requirements and
the performance testing strategy. Scheduling a workshop later in the
implementation is risky because any findings and recommendations
from the workshop could cause significant rework.

504
 Product-specific guidance
Finance and Operations apps
Following are recommendations for achieving optimal performance in
your Finance and Operations apps solutions:
▪ Use Tier-2 or higher environments based on business objectives.
Don’t use a Tier-1 environment.
▪ Keep the solution up to date with hotfixes, platform updates, and
quality updates.
▪ Identify and maintain a log of performance-related risks.
▪ Use DMF to import and export large volumes. Don’t use OData for
large volumes because it isn’t natively built to handle large payloads.
▪ Use set-based data entities and parallelism to import and export
large volumes.
▪ Build your own data entities to avoid potential standard entity
performance issues. Standard entities contain fields and tables that
you might not need for your implementation.
▪ Configure a batch framework including batch groups, priority, and
threads.
▪ Define batch groups and assign a batch server to each batch
group to balance batch load across AOS servers.
▪ Design heavy batch processes to run in parallel processing.
▪ Use preferable non-continuous number sequences with
pre-allocation.
▪ Perform cleanup of routines and jobs regularly.
▪ Avoid record-by-record operations; use set-based operations such
as insert_recordset and update_recordset where applicable.
▪ Be aware of the implications of security on performance. An ad-
ministrator role, for instance, will have better performance than a
user with limited access. Execute performance tests with users who
have appropriately configured security settings.
▪ Use the Optimization advisor workspace to identify business
processes to be optimized.
▪ Be aware of Priority-based throttling introducing service pro-
tection settings that prevent the over-utilization of resources
to preserve the system's responsiveness and ensure consistent

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 505
 availability and performance for environments running Finance
and Operations apps.
▫ Configure priorities for the OData and custom service-based
integrations, depending on your business-critical need for
these integrations.

Performance tools
▪ Trace Parser
▫ Diagnoses performance issues and various errors
▫ Visualizes execution of X++ methods as well as the execution
call tree
▪ Lifecycle Services Environment Monitoring
▫ Monitors server health metrics
▫ Monitors performance by using the SQL insights dashboard
▪ Query Store
▫ Reviews expensive SQL queries during defined intervals
▫ Analyzes the index used in queries
▪ PerfSDK and Visual Studio load testing
▫ Simulates single-user and multi-user loads
▫ Performs comprehensive performance benchmarking
▪ Performance timer
▫ Helps determine why a system is slow
▪ Optimization advisor
▫ Suggests best practices for module configuration
▫ Identifies obsolete or incorrect business data

Customer Engagement apps

The following are guidelines taken from the Fast Track Performance
Optimization Tech Talk for optimizing solution performance:
▪ Configure before customizing. Achieve a goal by configuring stan-
dard application components if possible. Customizations such as
plug-ins and JavaScript are useful, but they can carry a performance
overhead and can also make future upgrades more challenging.
▪ Avoid OptionSet attributes in a Quick Find search.
▪ Stay current on deprecation announcements to align with the
product roadmap.
▪ Reduce form load JavaScript for a better form load experience.

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 506
 ▪ Display the minimum required number of fields required in the forms.
▪ Design forms and pages to display the most important information
at the top.
▪ Minimize the number of controls in the command bar or ribbon.
▪ Use collapsed tabs to defer loading content.
▪ Avoid unsupported customizations such as direct Document
Object Model (DOM) manipulation.

Use the Solution Checker to detect potential performance issues in

code and configuration.

In high throughput scenarios, such as data migrations, follow the

guidance in the Service Protection API Limits documentation to
maximize throughput.

Conclusion
This chapter discussed why performance is expected and critical for
user adoption, the customer experience, and project success. We noted
that although Dynamics 365 projects are built to perform well at scale,
their flexibility means it’s crucial that implementation teams consider
performance as an iterative process throughout the solution lifecycle.

Performance often involves tradeoffs, so project teams need to work

closely with users and business stakeholders to determine and align
performance expectations.

Good solution performance starts with good design. This means using
Dynamics 365 products for their intended purposes and fully considering
the performance implications of design decisions early on in the process.
Many performance issues occur due to poor customization, configuration,
and design choices, so it’s important to proactively work to avoid this
by clarifying requirements and implementing best practices.

Performance testing is crucial to ensure performance expectations

are achieved during implementation. However, performance tests are
valuable only if they’re meaningful and provide a clear indication of

507
 how the system will perform in the real world.

Performance issues are difficult to resolve and place unnecessary strain

on the implementation team during the most critical points of the
project. With the right prioritization and planning, performance can
be baked into the implementation strategy, leading to a much more
confident launch with greater project success.

References
Product-specific guidance Product-specific guidance
(Finance and Operations) (Customer Engagement apps)
Monitoring and diagnostics tools in Lifecycle Services (LCS) Performance tuning and optimization

Performance troubleshooting using tools in Lifecycle Services (LCS) Optimize model-driven app form performance
Tips and best practices to improve canvas app performance
Work with performance and monitoring tools in Finance and
Operations apps Introducing Monitor to debug apps and improve performance
Performance considerations with PowerApps
Query cookbook

Monitoring performance by using the Query Store

Optimization advisor overview

Cleanup routines in Dynamics 365 Finance and Operations

Priority-based throttling

Priority-based throttling FAQ

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 508
 Checklist

Performance focus
Establish that the responsibility to deliver a performant
solution on the SaaS platform is shared by the cloud
service provider and the implementor who is customiz-
ing and extending the out-of-the-box application.

Design, develop, and test for performance in accordance

with the service boundaries.

Include performance as part of the requirement con-

versation and ensure it is reflected in the design and the
development of the solution, with specific documented
goals that are tracked through future change requests.

Establish development, user experience, and Information

Architecture guidelines that create a positive perception
of performance, prioritizing application responsiveness
and user experience.

Understand that the end-user performance is accumu-

lative of several factors, including the network latency,
device performance, browser add-ins used, firewalls,
and proxies. Establish the performance baseline in the
context of the actual user environment and location.

Ensure the performance testing considers realistic loads

and actual usage scenarios.

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 509
 Case study
Corporate travel company
learns how testing is critical
to measure and optimize
performance
A global corporate travel company was implementing Dynamics 365
Customer Service to drive a call center transformation. Several pro-
cess flows required integration with legacy systems, many with high
transaction volumes, and there was a business-mandated customization
that the project team agreed couldn’t be achieved using out-of-the-box
functionality. This complex solution was projected to support 4,000 users
at scale and needed careful planning to ensure optimal performance.

Despite the complex customizations and integrations, the customer

had decided performance testing was out of scope due to budget
constraints, assuming that because Dynamics 365 is a cloud service, its
performance is owned by Microsoft.

While still in the design stage, the project team decided to use the
best practices from the Success by Design framework and their own
experiences to highlight the risks of leaving performance testing out
of scope. The project team pointed out potential negative outcomes if
performance wasn’t considered:

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 510
 ▪ Users wouldn’t adopt the system if its performance affected user
productivity.
▪ Costs would be higher if reactive measures were required to ad-
dress performance issues later.

The project architect worked with the company’s stakeholders to explain

that while Dynamics 365 is a cloud service, factors such as additional
customizations, data volumes, and network latencies each play a role
in driving the performance of the application. Their recommendations
were summarized as follows:
▪ Review the application design and code for scalability.
▪ Establish baselines and review the areas that need to be optimized,
including network configurations.
▫ Formulate a baseline for form load times to understand perfor-
mance and align a new solution.
▫ Ensure performance with integrated systems meets the com-
pany’s expectations.
▫ Agree on realistic key performance indicators (KPIs) for the
business.
▫ Assess current end-to-end connectivity to determine network
performance and latency.

Based on the project team’s observations, the project steering commit-

tee was advised to approve adding solution performance testing to the
test cycle, and agreed to do so.

The customer’s architect worked with the implementation partner’s

architect to plan for performance testing by:
▪ Defining the key criteria for testing, including integration scenarios.
▪ Assessing the concurrent user loads and peak usage.
▪ Examining the data volumes.
▪ Establishing the tools for testing.

Performance testing helped the team identify several gaps before

deploying the solution to production. In addition, the team:
▪ Optimized their network configuration.
▪ Identified sync calls that were blocking resources, and made the
necessary design changes.

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 511
 ▪ Reinforced and incorporated basic patterns (such as retrieving
only the required attributes) across their code components.

As mentioned earlier in this chapter, “Prevention is better than cure.”

Performance testing ensured optimal performance and stakeholder
alignment before the system was made available for end users. For
overall success and higher levels of adoption, a performance plan is
fundamental throughout an application’s lifecycle.

Implementation Guide: Success by Design: Guide 17, A performing solution, beyond infrastructure 512
Prepare
18 Prepare for go live
19 Training strategy

513
18 Guide
Prepare
for go live
Implementation Guide: Success by Design: Guide 18, Prepare for go live 514
The final countdown for
the start of a journey.

Introduction
Go live is the process through which a new solution
becomes operational. It’s a critical milestone during
the deployment of a new business solution.
This is the stage in which all the parts of the project come together and
are tested and validated to ensure everything works as expected, not
just on their own, but as a whole.

A good plan strives to ensure a successful go live by minimizing un-

foreseen issues and project risks not identified during previous phases.
A good plan also has processes in place to deal with times when the
unexpected does happen, which makes for a more resilient project.

The go-live process goes beyond getting the production environment

The go-live readiness plan addresses all
the activities to minimize unforeseen
ready and deploying the solution. It includes a variety of activities such as
issues and project risks and build executing data migration and communication plans, completing training,
processes to help address possible
unexpected issues. finalizing testing, signing off on the solution, and onboarding users.

Preparing the solution for a successful go

live focuses on activities such as training, For scenarios in which there are phased rollouts, the plan should also
preparing the production environment,
ensure that additional deployments are considered in advance. It’s
conducting user acceptance testing (UAT),
and all the other activities described in important to keep refining the plan and include the learnings obtained
previous chapters. Activities should be
planned well in advance and completed
during each rollout to improve the plan and build resilience.
and signed off on before the cutover starts.

And finally, the cutover, the last sprint! Go-live readiness includes all the key activities identified in the initial
stages of the project and refined throughout the project to ensure a

Implementation Guide: Success by Design: Guide 18, Prepare for go live 515
 smooth go live. These activities encompass getting all required re-
sources to get the solution ready for production and ensuring that end
users are trained, engaged, and part of the process to drive adoption
and stickiness to the new solution.

When going live with a new solution, there’s often a transition period,
also known as cutover. The cutover process involves several steps that
need to be planned, executed, and monitored to ensure the completion
of all the essential activities critical to the transition to the new solution.

Go-live readiness
All the tasks and efforts undertaken during an implementation project
are preparation for the biggest milestone of the project: go live. In
the Success by Design framework phases, completion of these tasks is
when the project reaches the Prepare phase.

At this point, if you followed all the previous guidance and recommend-
ed practices, the solution should have sufficient maturity for going live.
You should perform an evaluation on the preparedness of the people,
processes, and systems for the solution to ensure no critical details have
been overlooked for go live. While you’ll never be in a position of zero
risk for go live, the go-live readiness review is a qualitative method to
determine how fully prepared the new solution is to run your business.
When the go-live strategy is aligned with best practices, requirements
for going live have been successfully completed (including testing, code,
and configurations), and there’s a concise and agreed-to plan for the
next actions required for the transition to the new system (cutover), then
the project is ready to go live. Figure 18-1 shows the Success by Design
implementation phases and when go live occurs.
Fig.
18-1

Initiate Implement Prepare Operate

Get ready to start Design Build Deploy Live

Preparation for go live Go live

Implementation Guide: Success by Design: Guide 18, Prepare for go live 516
 To get ready to go live, you need to verify completion of the most
important project tasks. Once confirmed, you can perform the cutover
activities. The next sections discuss these topics in detail.

Assess go-live readiness

The implementation team should assess go-live readiness early and
frequently. This provides clarity on the status, readiness, completeness,
and maturity of each activity and helps project stakeholders and the
steering committee make a go/no-go decision.

The assessment can be performed and reviewed multiple times in

distinct phases of the project. When the project moves to the Prepare
phase, the implementation team evaluates the preparedness of the
solution, which serves as a quality gate to deploy the new solution.

This evaluation is done with the go-live readiness review.

Go-live readiness review

To assess the go-live readiness of a project, Success by Design pro-
motes the go-live readiness review, a deep-dive review that can be
performed as a workshop. The review validates whether a solution
meets an organization’s needs and expectations, both functionally and
nonfunctionally. It also validates the plan that has been established
to ensure a smooth transition in the available cutover window and
to avoid any surprises during and after go live. The go-live readiness
review summarizes all the tasks that have been completed and covered
from previous reviews.

The review uncovers potential risks and issues that could imperil go live
and provides a set of recommendations and actions to mitigate them.

Start early to prepare for the go-live review. Schedule time to complete
the go-live checklist and conduct the review, accounting for time to
mitigate possible risks and issues, especially go-live blockers. Keep
in mind that time is needed to prepare the production environment.
Starting the review just a few days away from the go-live date risks a

Implementation Guide: Success by Design: Guide 18, Prepare for go live 517
 delay. However, starting the review too early can also be detrimental
to the process because there might not be sufficient information about
the project to complete the go-live review with reliable results to
properly identify potential risks of the solution.

To execute the go-live readiness review (shown in Figure 18-2), you

should complete all required activities for going live and have the business
sign off on them. The main requirements for going live are as follows:
▪ System integration testing (SIT), UAT, and performance testing are
completed and signed off on.
▪ Code is complete and deployable packages are prepared to move
to production.
▪ A data migration plan is prepared.
The go-live readiness review is a holistic
view of the implementation project: ▪ A cutover plan is ready and signed off on.
looking back on the completed and
▪ End user training is complete with the correct security roles.
signed-off on activities required for going
live, looking at solution preparations, and ▪ The correct number of licenses are available and assigned.
looking forward to the plan to roll out and
maintain the new solution.
▪ A production support plan is established and signed off on.
▪ Any critical open issues from the Solution Blueprint Review or
other previous implementation reviews are mitigated.
Fig.
18-2

Initiate Implement Prepare Operate

Get ready to start Design Build Deploy Live

Go live readiness workshop Go live

Check your readiness

for go live
Once the solution is considered ready and the go-live prerequisites are
fulfilled, it’s time to execute the go-live readiness review and evaluate how
well prepared the solution and the organization are for the next step.

Use the go-live checklist to perform the review to help ensure a

smooth go live. Create a mitigation plan to address identified issues
and risks and identify an owner for each one.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 518
 Keep the review objectives in mind
When executing the review, make sure you understand the responses
on the checklist. If something isn’t clear, ask questions and validate
the responses.

Identify all the possible risks and issues and get the recommendations
on each key area. All issues and risks must have a mitigation plan.
Identify workarounds for any go-live blockers. Follow up on mitiga-
tions of issues identified for critical activities.

Stakeholders must be kept informed about the results of the go-live

review and resulting mitigation actions.

After the review is complete, the implementation team determines

whether the go-live date is feasible.

Go-live checklist The go-live checklist

Solution Scope The go-live checklist is a list of requirements for go live. Use it to assess
to be released
and validate the preparedness of your solution. The checklist includes
Acceptance of mandatory activities as well as recommended practices.
the solution

SIT completion In the next section we discuss in depth the main topics in the checklist.
Refer to the “Product-specific guidance” section for specific product
UAT completion requirements.

Performance testing
completion
The Solution Scope to be released
Data migration readiness What you check
and validation
The Solution Scope conceptualizes the final solution that will be ready
Confirm external for go live. It includes all the requirements designed and developed
dependencies
during the Implement phase, ensuring their alignment with the busi-
Change management for ness needs to be able to operate once live.
initial operations readiness

Operational support Consider the following during the Solution Scope review:
readiness
▪ Business processes supported by the solution

Implementation Guide: Success by Design: Guide 18, Prepare for go live 519
 ▪ Different solutions and applications used
▪ Whether the solution contains customizations and independent
software vendor (ISV) solutions
▪ Type and volume of integrations
▪ Number of go lives or rollouts
▪ Number of users at go live and in future rollouts
▪ Solution Scope signed off on by the business

Why you check it

It helps the stakeholders understand the complexity and magnitude of
the solution. In addition, even though the Solution Scope was defined
at the beginning of the project and you’re running the project accord-
ing to it, reviewing the Solution Scope at this time in the project can
help determine additional actions and identify potential risks.

Consider the scenario of a big-bang go live containing a heavily

customized solution in which multiple integrations are delivered. The
more functionalities going live on the first day, the greater the risk
that issues will occur. If the risk is higher, be sure to assemble a good
support team and the right escalation path. The Solution Scope must
be representative of the scope of your go-live event. The recommen-
dations you follow for a big-bang go live might be different from those
for a smaller implementation.

Expected outcome
Solution Scope is aligned with the solution that’s going to be live. It has
been communicated, shared with stakeholders, and signed off on by
the business, agreeing that the expected scope is covered.

Risks of not reaching expected outcome

Unless a Solution Scope is clearly defined and shared, stakeholders
won’t have an understanding of what will be included in the solution.
In addition, the solution might not include all requirements or might
include developments that don’t meet business expectations, which
might cause delays. A well-defined Solution Scope, on the other hand,
significantly increases the probability of project success.

520
 Mitigation plan
Compare the Solution Scope with the solution delivered for go
live, share the Solution Scope and comparison results with the key
stakeholders, and determine whether the solution is complete or if
functionalities are missing. If any are missing, prioritize them and
assign level of risks, owners, and expected completion date.

Acceptance of the solution

Go-live checklist
What you check
Solution Scope Verify that the complete solution has been fully tested with all required
to be released
types of testing and accepted by the business stakeholders.
Acceptance of
the solution Why you check it
The outcome, quality, and coverage of testing are reliable indicators
SIT completion
of the readiness level for go live, ensuring that unexpected scenarios

UAT completion don’t come up after end users get their hands on the new system.

Performance testing As described in Chapter 14, “Testing strategy,” testing accurately gaug-
completion
es the readiness of the solution. Testing measures a solution’s quality
Data migration readiness and effectiveness because testing simulates how the solution will oper-
and validation
ate in real life. As a result, testing builds a solid foundation on which to
Confirm external determine whether the solution is acceptable, enabling the business to
dependencies make a go/no-go decision. The business is required to sign off on the
Change management for solution when there’s objective and rigorous testing to prove that the
initial operations readiness solution fulfills the business vision.
Operational support
readiness Expected outcome
Unit testing through end-to-end testing must be completed success-
fully and the results signed off on by the business. By doing so, the
business states that the solution as built meets their end-to-end pro-
cess needs and that those processes are ready to be executed on the
new solution.

Risks of not reaching expected outcome

Going live with incomplete testing or without having sign-off from
the business means that you cannot validate you have a reliable
solution and therefore cannot be certain that it will work properly

Implementation Guide: Success by Design: Guide 18, Prepare for go live 521
 in production. There’s a high risk of issues such as rollbacks, slow
performance, unstable integrations, security holes, last-minute devel-
opments, and platform churns.

Mitigation plan
Following the best practices in Chapter 14, “Testing strategy,” you’ll
ensure the integrity and effectiveness of the solution, minimize any
go-live blockers, and avoid rushing into fixing unexpected and critical
bugs close to go live. You’ll complete the testing and ensure the integ-
rity and effectiveness of the solution.

Several types of testing are covered in the chapter about testing.

Some elements from these testing types are included in the go-
live checklist and need to be validated. The main testing types are
System Integration Testing (SIT), User Acceptance Testing (UAT), and
Performance testing.

Go-live checklist System Integration Testing completion

Solution Scope What you check
to be released Validate that System Integration Testing (SIT) is successfully completed.
It’s important to test on expected peak volumes and have the business
Acceptance of
the solution sign off on it.

SIT completion
Why you check it
SIT validates that the behavior of the solution, including integrations
UAT completion
between different applications and with external systems, works
Performance testing properly. During SIT it’s important not only to test and verify the inter-
completion actions between all parts of the solution but also to plan and prepare

Data migration readiness

for the unexpected.
and validation
Expected outcome
Confirm external
dependencies The testing should include simulations of external systems that are down,
how those scenarios are communicated, and the impact on the user
Change management for
initial operations readiness experience. Chapter 14, “Testing strategy,” has more details describing SIT.

Operational support
readiness Risks of not reaching expected outcome
Don’t wait until UAT to perform SIT because this could cause significant

Implementation Guide: Success by Design: Guide 18, Prepare for go live 522
 failures during UAT. If you execute SIT during or after UAT and discover
that the incorrect integration pattern was used and you need to rede-
sign your integrations, this will put your go live at risk.

Going live without fully testing the integrations might result in un-
steady and ineffective system connectivity, performance issues, flawed
and weak interfaces, and data flow issues such as data inconsistency
and data not being available on time.

Mitigation plan
Complete SIT with peak volumes that are close to actual peak volumes
and get sign-off from the business that the integration testing strategy
works for the go-live. For guidance on which integration pattern to
choose, see Chapter 16, “Integrate with other solutions.”

User Acceptance Test completion

Go-live checklist
What you check
Solution Scope Verify that User Acceptance Test (UAT) is successfully completed (or
to be released
almost completed if there aren’t any outstanding significant cases) and
Acceptance of signed off on by the business.
the solution

SIT completion Why you check it

UAT helps validate the inclusion of all the expected functionalities. The
UAT completion solution might still have some minor bugs and need some fixes, but
you can ensure that no critical part of the solution is missing.
Performance testing
completion
Often during UAT, edge cases that were overlooked might be discov-
Data migration readiness ered, resulting in continuous fixes and improvements to the solution
and validation
during this period.
Confirm external
dependencies
UAT enables users to use the system and understand how to perform
Change management for activities. It also provides an opportunity for users to bring up scenari-
initial operations readiness
os that might have been missed.
Operational support
readiness UAT should include different types of personas, a variety of geographic
locations, and different environments (such as virtualized, physical, and
remote) and devices (such as laptops, mobile devices, and tablets).

Implementation Guide: Success by Design: Guide 18, Prepare for go live 523
 Expected outcome
UAT should include the following requirements:
▪ Test cases should cover the entire scope of the requirements,
happy path and edge scenarios.
▪ Use migrated data for testing because it validates how real data
will behave in real life.
▪ Perform UAT with the correct security roles assigned to users; don’t
use a general security role or assign the System Admin role to all users.
▪ The solution must comply with any company-specific, country- or
region-specific, and industry-specific regulatory requirements.
Document all features and testing results.
▪ UAT must be conducted in an environment in a Microsoft apps
subscription to ensure that the environment’s properties approxi-
mate your production environment as much as possible.
▪ Obtain approval and sign-off from the customer.

Risks of not reaching expected outcome

If UAT isn’t completed close to go live, the live operations might be
delayed due to various issues including unsuitable processes, perfor-
mance issues, and poor training.

If the decision is made to go live with an incomplete UAT, the project

team might discover during live operations that the solution doesn’t
work properly, and recovery could be difficult, costly, and take a lot of
time and effort to stabilize. This can be represented in new and un-
foreseen developments, redefinitions on design, extending the time of
resources, or even in extreme cases, production shutdown.

Mitigation plan
Typically, UAT is considered complete when all the cases are covered
and there are no blocking issues and any high-impact bugs are iden-
tified. Establish a mitigation plan to address open items identified
during UAT, with an owner and an estimated time of completion.
Without a mitigation plan, there’s a substantial risk of delaying the
go-live date.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 524
 Performance testing completion
Go-live checklist
What you check
Solution Scope Validate that performance testing is successfully completed and signed
to be released
off on by the business.
Acceptance of
the solution Why you check it
Overlooking performance testing might result in performance issues
SIT completion
post–go live. Implementations often fail because performance testing is

UAT completion
conducted late in the project or not until the production environment
is ready, when it’s used as a testing environment. A production envi-
Performance testing ronment shouldn’t be used as a testing environment. It is easier to fix
completion
performance issues in a test environment than in a live environment.
Data migration readiness Also, bear in mind that fixing issues in production may require downtime
and validation
and it also might result in production downtime when trying to fix it.
Confirm external
dependencies Any open issues critical for production need to be addressed before
Change management for going live.
initial operations readiness

Operational support Performance testing covers the following:

readiness ▪ Load times with peak volumes of transactions
▪ Data search performance
▪ End-to-end processes and specific business processes
▪ Performance from different devices, browsers, and virtualized
environments
▪ Integrations
▪ Environment for performance meets the standards to ensure it’s
reliable for testing

It’s important Expected outcome

The solution delivered is performant and meets the business perfor-
to identify the mance criteria: it supports the load of expected transactions and user
root cause of concurrency and usage, and the speed, system response time, stability,
issues so as not and scalability are acceptable.

to replicate them Risks of not reaching expected outcome

in the production A common misconception is that a production environment is more
environment. performant than a test environment, and this belief is sometimes used to

Implementation Guide: Success by Design: Guide 18, Prepare for go live 525
 justify skipping performance testing. However, performance is not only
about infrastructure, as discussed in Chapter 17, “A performing solution,
beyond infrastructure.” Therefore, we highly recommend that you
conduct performance testing before going live in a test environment. A
production environment cannot be used to conduct core testing; it’s used
only for mock cutover, cutover, final validations, and live operations.

If performance checks are done just during UAT, that might not be a
good representation of the actual usage post–go live. For example, if
during UAT there’s good coverage of roles and personas, regional ac-
cess, devices, and environments but those are for a small percentage of
users, UAT can validate the solution readiness in terms of functional and
even end-to-end performance of each business process, but it doesn’t
represent the full load and concurrency of actual usage post–go live.
Therefore, consider defining a separate performance testing strategy
that can simulate stress on the solution and concurrency usage.

Mitigation plan
Go-live checklist Execute performance testing in parallel with UAT. Establish a mitigation

Solution Scope plan to address issues and assign owners and expected completion
to be released dates. Identify the root cause of issues so as not to replicate them in
the production environment.
Acceptance of
the solution

SIT completion Data migration readiness and validation

What you check
UAT completion
It’s essential to validate that there’s a well-designed data migration
Performance testing plan for go live.
completion

Data migration readiness

Why you check it
and validation Data migration is a key activity during the cutover. When you are
migrating from a legacy system, you need to make sure to have the
Confirm external
dependencies accurate and complete data on day one of your live operations.

Change management for

initial operations readiness When executing the data migration, it is important to verify the data
quality. You may not want to import bad data from your legacy system
Operational support
readiness into the new one. So then it is also crucial to have a plan to conduct
data validation after migration.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 526
Expected outcome
Data migration is tested several times prior to execution during the cu-
tover to validate the strategy and processes, identify data corruption or
duplicated data issues and address them, and make sure performance
is measured, validated, and fits within the cutover time window.

All scripts and processes planned for the cutover migration are tested
and signed off on by the business.

Any migration issues and risks are logged and a mitigation plan estab-
lished. This information helps inform the go/no-go decision.

Risks of not reaching expected outcome

It’s important to carry out all the steps of a data migration plan.
Otherwise, there’ll be a lack of confidence in the information because
it might not reflect accurate data. In addition, concerns will arise about
the ability to complete the go-live migration in a timely manner,
especially if record quantities are exceptionally large.

Data quality is a key factor in the user experience. For example, if a

user’s leads are imported but a critical field is missing or has corrupted
information (for example, truncated, not completed, or without the
correct currency or value), this severely impacts how the user perceives
the new solution, reduces the trust level, and might even eventually
lower solution adoption.

If incorrect data is flowing in your new system, that could also be re-
flected when communicating with your customer through the systems
once you are live.

Mitigation plan
Execute multiple dry runs of how to migrate the data and prepare the
migration plan. Follow more recommended practices to prepare your
data migration by reviewing Chapter 10, “Data management.”

527
 Confirm external dependencies
Go-live checklist
What you check
Solution Scope Verify that external dependencies such as ISVs and third-party systems
to be released
and services are aligned with the timelines and scope for go live.
Acceptance of
the solution Why you check it
External dependencies are outside the direct control of the project team.
SIT completion
This means it’s even more important for these dependencies to be ac-

UAT completion
counted for when building the project plan and managing the schedule.

Performance testing Expected outcome

completion
If the scope of the solution includes external dependencies, the
Data migration readiness implementation team should coordinate with them to ensure that
and validation
expectations are aligned, requirements are met, and their solutions are
Confirm external aligned with the roadmap timelines.
dependencies

Change management for All external dependencies should be documented and monitored.
initial operations readiness

Operational support Risks of not reaching expected outcome

readiness Lack of coordination and communication with external dependencies
might result in a delay in go live.

Mitigation plan
It is good practice to have regular meetings to review dependencies
status because problems can cause delays and other project issues.

Change management for initial

operations readiness
What you check
It’s important to include change management at the beginning of the
project and assess the level of change readiness.

Why you check it

When delivering a business solution, the success of the technical im-
plementation and the alignment with the business needs and goals are
paramount but not sufficient to deliver the return of investment. The

Implementation Guide: Success by Design: Guide 18, Prepare for go live 528
 true value of the business solution is realized when end users actively
Go-live checklist adopt and engage with the processes implemented.

Solution Scope
to be released It’s natural to be resistant to change even if it’s positive and can sig-
nificantly improve one’s experience. It’s important to have a change
Acceptance of
the solution management plan that prepares users for changes. There are several
resources and strategies available to achieve successful change and
SIT completion adoption that include ways to communicate, reduce resistance, and
allow feedback.
UAT completion

Expected outcome
Performance testing
completion Our goal here is not to explore the plethora of change and adoption
strategies but to highlight the key principles and activities that are
Data migration readiness
and validation critical before the rollout of a new business solution.

Confirm external
dependencies It’s important to assess early in the project, and continually reassess,
the changes in an organization to ensure there’ll be a smooth adoption
Change management for
of the system.
initial operations readiness

Operational support The following are key elements to drive change management close to
readiness
go live.

Training  For go live to be successful, it’s vital that users know how
to use the system on day one of operations. Therefore, successful user
training is key to go live.

Ensuring that users are trained helps them achieve the desired results from
the new solution and leads to higher productivity and adoption. Failure to
establish a training strategy can negatively impact usage and adoption.

The most common training strategy is a formal set of sessions deliv-

For go live to be ered in a classroom setting or online. Refer to Chapter 19, “Training
successful, it’s vital strategy,” for additional training strategies.
that users know
how to use the When possible, end-user training should be delivered before the
go-live date because this prevents delayed access to the new solution.
system on day one It can also contribute to higher adoption because the users feel better
of operations. prepared and a part of the program and had the opportunity to share

Implementation Guide: Success by Design: Guide 18, Prepare for go live 529
feedback before the new solution went into production.

User engagement strategy  A user-centric mindset helps address

not just the functional requirements but also the unspoken needs of
the user. Providing a story of empowerment to users, how the solution
will help them be more efficient in their jobs, for example, is a message
that strongly resonates with end users.

Your program must have an effective way to engage the end users of
the solution to help drive adoption and also eliminate the risk of build-
ing a solution that doesn’t necessarily meet user requirements.

A key activity to perform early in the project is the identification of the

super-users, or influencers, within the community who are willing to
actively engage with the project team. Provide opportunities through-
out the project for these people to provide input and feedback.

Following are sample project milestones for which super-user feedback

is most beneficial:
▪ Early minimum viable product (MVP) or proof-of-concept
demonstration
▪ Sprint demonstrations during implementation to highlight busi-
ness functionality
▪ User experience and information architecture design
▪ Data mapping and validation that data migration activities are
involved in the project
▪ Baseline performance metrics for frequently accessed components
▪ Expectations for end-user devices and accessibility needs
▪ Review of the timing of key milestones such as go live and cutover
▪ Backlog prioritization and setting expectations for gaps, open
issues, and future functionality

There needs to be a well-structured nurture program for the business

users that focuses on effective communication to keep them up to date
and engaged in the program. Although most of this work happens
throughout the implementation project, the culmination of all this
planning at go live should be validated and signed off on as ready to go.

530
 Business sponsorship  The business sponsor, or executive sponsor,
is a trusted and influential leader who enables cultural changes and
plays an essential role in championing transformation throughout the
organization by communicating the value of the new solution.

Creating an awareness of the key executive sponsors and bringing

them in front of the user community to articulate the value of the new
solution can positively influence adoption and usage.

As part of the go-live activities, the executive sponsor’s communication

strategy needs to be well planned. This includes the types of communi-
cation, the stakeholders, the timing, and recognizing the contribution
from end users.

To help overcome adoption challenges, it’s crucial to get leadership

support to encourage innovative technology use. End users are less
resistant to adopt novel solutions when business sponsors serve as a
role model.

User support  Creating a community that involves the super-users

and providing resources such as videos, tutorials, knowledge base
articles, and business processes will help with user adoption post-go
live and encourage ongoing user engagement.

A well-defined channel to communicate to users about upcoming

changes, open issues, and service notifications in case of planned and
unplanned maintenance and outages should be established.

Risks of not reaching expected outcome

A poor change management plan might cause ineffective execution of
processes on the first day of go live. Users would therefore likely have a
lack of confidence in the new solution, which would be counterproduc-
tive for a successful rollout.

Mitigation plan
Reassess the change management plan throughout the implementa-
tion project.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 531
 Operational support readiness
Go-live checklist
What you check
Solution Scope Verify that there’s a monitoring and maintenance plan for the solution
to be released
in the production environment as well as for transitioning the plan to
Acceptance of support teams.
the solution

Why you check it

SIT completion
Verification ensures the health and availability of the solution once it’s

UAT completion
live. Before go live, there must be a plan for the transition to the solu-
tion’s support teams.
Performance testing
completion
Expected outcome
Data migration readiness Support teams can be from the partner or customer organization, and
and validation
they need to have the necessary resources, tools, and access to sub-
Confirm external ject matter experts, business analysts, process specialists, and change
dependencies management champions. Support teams should also be trained on the
Change management for solution; this ensures that tickets that come through the help desk are
initial operations readiness properly addressed.

Operational support
readiness Notify stakeholders when all parties agree that the system is ready to
go into production and send them a list of end users who will use the
new system.

Users should be notified of any unresolved issues or required work-

arounds and provided with a support contact.

Include the following in the operational support and maintenance plan.

Continuous deployment plan  It’s important to go live in a service-

able system version in order to take advantage of latest updates, have
a better quality of the solution, and have the support available if a
hotfix is required.

It’s also important to understand the update cadence of your new

solution so that you can incorporate updates and new features and
replace deprecated ones quickly. This should be part of your mainte-
nance plan; it’s vital for the continuity of your operations in production.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 532
 For discussions of continuous deployment plans, refer to Chapter 20,
“Service the solution,” and Chapter 21, “Transition to support.”

Monitoring  Establish a monitoring strategy for performance issues,

errors, feature usage, and adoption. This makes the post–go live
process easier to manage and makes it easier to provide updates to
stakeholders about the new solution.

Hypercare  Consider including Hypercare after golive because it

supports and expedites the mitigation of issues, reducing time for
stabilization of your operations.

For additional details about monitoring and Hypercare, visit Chapter

21, “Transition to support.”

Risks of not reaching expected outcome

If there isn’t a well-defined plan to support production or if the plan
isn’t properly communicated before go live, when issues start surfacing
during the operations there won’t be a way to communicate them and
assign resources and severity levels.

There’s also a risk that the new solution could go live in a system ver-
sion that’s out of service. In such a scenario, if an issue is uncovered in
production, you’ll need to update the solution to the latest version to
be able to apply the hotfix. In addition, automatic updates that install
unexpectedly might affect the deployed solution and cause outages,
A support plan unavailability, and blockings.
enables support
teams to be more Mitigation plan

proactive and As discussed earlier in this section, it’s important before go live to plan
the transition to the teams who will support the solution. A support
preventive rather plan enables support teams to be more proactive and preventive
than reactive. rather than reactive.

533
 Production environment
readiness
It’s critical to prepare the production environment before go live. There
are different approaches to achieve this, depending on the applications
in the scope of the deployment.

For some applications, the production environment can be created

from an initial phase without any dependencies. For other applications,
a formal process needs to be followed to create or deploy the produc-
tion environment. Additionally, several milestones must be completed
prior to deploying the production environment to help with the sizing
of the environment and impact on factors such as performance and
scalability of the environment.

The “Product-specific guidance” section later in the chapter details

the steps for each application. For a discussion of production environ-
ments, see Chapter 9, “Environment strategy.”

Establish a plan to prepare the environments for production and

execute the following activities before the cutover. This prevents any
last-minute issues and provides more time to ensure that tasks are
completed successfully.
▪ Deploy applications and update them to the latest version.
▪ Ensure that users were created ahead of time and assigned appro-
priate security roles.
▪ Assign security groups to environments and add users to the
groups to ensure that users are created only in the environments
for which they should have access.
▪ Ensure sign-off on the security and compliance strategy including
penetration testing when applicable, compliance requirements,
and data sovereignty requirements.
▪ Tune the environment when applicable.
▪ Create, import, and validate configurations.
▪ Run and validate integrations.
▪ Ensure that any specific configurations for different applications are

Implementation Guide: Success by Design: Guide 18, Prepare for go live 534
 completed ahead of time, especially configurations that take time to
complete. For example, setting an integration with an email engine
such as Outlook might depend on the size of a user mailbox, and
the bigger the mailbox, the longer the process takes to complete.

Finally, keep in mind that production environment readiness is also a

big dependency for other activities during the cutover. For example,
the final data cannot be migrated and validated until there’s a produc-
tion environment to start those activities. And before data migration
can occur, other activities need to be completed, such as creating the
users and assigning the proper security roles.

Cutover, the last sprint

The last mile before going live is the cutover, which marks the transi-
tion to the new solution. A project cutover entails the execution of a set
of tasks required to successfully go live with a well-defined scope on a
production environment. In the past, it was common to do a big-bang
go live after a multi-year solution implementation period. Now, how-
ever, businesses frequently follow a more agile approach. For example,
a company might plan an initial go live in a small country or region or
just for the core financial processes and complete the rollout to other
countries or more complex functionality at a later point. A cutover
process needs to be defined and executed for each rollout.

In most businesses, only a brief period is available to complete the

cutover activities because during this time, users cannot use the old
business system any longer, and the new business system is not ready
yet. Often the cutover period is referred to as “the cutover weekend”
because it’s usually no longer than 48 hours. For businesses operating
all the time, this time window might be even shorter.

The cutover process, shown in Figure 18-3, involves several steps

that need to be planned, executed, and monitored to ensure that all
activities necessary to transition to the new solution are successfully
completed. The cutover process includes strategy, planning, execution,
monitoring, and control.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 535
 Fig.
18-3

Pre-cutover Cutover Go live

Strategy Plan Rehearse Cutover execution Post–go live

Timing Actions Mock cutover Runbook execution Support

Resources and Timelines Go/No-go

competencies
Owners
Communication plan
Rollback steps
Roles and
responsibilities Resources

Rehearse

Communicate

Go/No-go

Cutover strategy
The cutover strategy begins in an early stage of project planning, is
refined throughout the project, and is completed before the cutover
plan is created.

The cutover strategy ensures alignment of the cutover plan with orga-
nizational objectives and includes the following aspects:
▪ Timing, market conditions, and other environmental aspects
necessary to go live with the new business solution
▪ Organizational readiness requirements in terms of available com-
petencies and resources
▪ Setting up the communication plan
▪ Defining roles and responsibilities
Plan to go live
when there’s a Time your go live
It’s challenging to establish a realistic and achievable go-live date be-
slower flow of cause this date is set early in the implementation. It’s always important
activities in the to have some buffer time when planning the milestones dates, but this
business. is especially true for the go-live date.

To set the go-live date, consider the time required to complete testing
and resolve any issues that might arise, in addition to time for the
preparation of the production environment and the cutover activities.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 536
 Preferably, plan to go live when there’s a slower flow of activities in the
business. For example, in the case of a seasonal business, like for some
retailers, choose the season when sales are slow to lessen any negative
impact on the business. Also consider avoiding going live during any
busier times, for example, month or quarter end. In addition, take hol-
idays into account, they might mean low activity for some businesses
and for others might be the busiest times of the year. It’s also import-
ant to make sure that all team members and users are fully available
during that time.

In addition to those busy times, consider system release dates. Microsoft

continually makes improvements in the system to deliver a better prod-
uct and has established release dates that usually entail a mandatory
update. You’ll need to decide whether to set the go-live date before
or after a release. This consideration also applies for any other external
To determine a realistic go-live date, applications you might be integrating with your new solution.
consider the amount of time required for
each pre-go-live activity, including both
the internal and external factors impacting External factors might also affect the timelines in a project, especially in
your project. Then add some buffer time,
bearing in mind that adjusting dates and long projects during which unforeseen events such as a natural disaster,
resources along the project is expected
political changes in the country of implementation, company fraud, or
because there might very well be delays,
unforeseen events, loss and replacement a pandemic might occur. It’s important to create a list of possible miti-
of resources, and external or internal
dependencies that don’t complete on time.
gation plans for unforeseen events when planning your project. External
factors are part of the risk management of your project.

Resources and competencies

During the process to define the cutover strategy, program stakehold-
ers should identify which roles and competencies are required for the
activities in the cutover plan and assign ownership for each activity.

The stakeholders should verify that all necessary resources are avail-
able not only for the requested duration of the cutover plan activities
but also to support post-go live.

These resources might need to be requested from external companies

if they’re not internally available for the project team.

Communications plan
The communications plan is an essential part of the cutover plan. This

Implementation Guide: Success by Design: Guide 18, Prepare for go live 537
 plan identifies all the stakeholders involved in the cutover. The plan
should include at least one communication channel and appropriate
distribution lists, depending on the type of communication required.
It’s important to identify and document the different communications
required for go live, who is responsible for triggering the communica-
Effective communication helps avoid
tions, and the recipients list. Having a communication plan as part of
uncertainty and provides visibility to
stakeholders about project status and the the cutover plan enables better visibility into who the stakeholders are,
results of each activity, which is important
for a successful cutover.
at what point in time they should receive communications, and who
the points of contact are.

Roles and responsibility

Effective delegation of tasks is key for a successful cutover. Therefore,
it’s important to define which roles are needed, based on the required
activities, skills, and responsibilities, and assign them to members
involved in the cutover. Also assign an owner to each activity and
ensure that all owners understand what’s required of them during the
execution of the cutover.

Team members working together

effectively, including during the The necessary skills and responsibilities will help better define which
cutover as we’ve been discussing, is the
roles need to exist and to whom they’ll be assigned. It’s important
foundation of a successful project. All
those roles functioning together, doing that those with a role assigned know what to do and how to do it, it’s
the right thing at the right time, will
result in a successful cutover.
expected that each person contributes with their tasks, delivering them
on time and well executed.

Cutover plan
The cutover process is a critical and complex step that must be planned
and practiced in advance. The center of the cutover process is the cutover
plan, which lists in detail every step that needs to be executed and moni-
tored to prepare the production environment for business processes to be
executed once the cutover is complete. Cutover activities include system
configuration, data migration, data validation, and decommissioning of
legacy systems when applicable. These steps must be precisely orches-
trated to minimize disruption and ensure a successful deployment.

When creating the cutover plan, it’s important to consider and

document all dependencies, the timing of each task down to the
minute, who’s responsible and accountable for each task, instructions,

Implementation Guide: Success by Design: Guide 18, Prepare for go live 538
verification steps, and any additional documentation associated with
each task. To arrive at the correct timings for the cutover plan, activities
must be practiced and tested multiple times. It’s recommended to
perform a “mock cutover,” or dress rehearsal, simulating the activi-
ties of the real cutover in sandbox environments. Depending on the
complexity of the solution, it might be necessary to conduct multiple
rounds of mock cutover. The goal is that no matter how complex the
cutover procedures are, the project team has high confidence that final
execution will work flawlessly.

A successful cutover plan is when there are zero or few disruptions

during the execution of the cutover and a plan is in place to mitigate
any breakdowns.

Components of a cutover plan

A successful cutover activity always begins with a detailed and well-de-
fined cutover plan. The cutover activity itself is performed as part of
go live, but planning for the cutover should start at an early phase of
the project, and the business should sign off on it. The cutover plan is a
living document that is updated throughout the project as the solution
is implemented and evolves.

All project stakeholders should be involved in each activity. Project

stakeholders should sign off on activities, with a commitment that all
necessary resources are available for the requested duration to support
the go-live activities.

A clearly defined plan including detailed actions, timelines, and owners

needs to be established. The plan should also include a rollback sce-
nario, which adds to the value of the cutover plan. Planning a rollback
is especially important to ensure that the business is not affected by
the outcome of the cutover plan if it isn’t successful and that there’s a
backup plan to retry the cutover at a later point.

The implementation team should rehearse all the cutover plan steps to
ensure that the cutover strategy and communication plan are ready. By
doing so, the predictability of the outcome increases, backup plans can
be created to avoid dependency issues, the duration of each activity is

539
 validated, and any identified issues with the plan can be corrected. Figure
18-4 is an example of a cutover plan and shows the list of activities, the
different elements for good traceability, and execution of the cutover.

Fig.
18-4
Cutover plan example
Go-live date: 8/1/2021
Today: 6/28/2021

Estimated Estimated Validated Actual Actual %

Cutover step Area Environment System Countdown Status Duration Owner
start end by start end complete

Finalize configuration environment

Customization
10 IMP Config Dynamics 3D Planned 1D 7/1/2021 7/1/2021 Team 1 PM 0%
installation
Configurations
20 IMP Config Dynamics 5D Planned 2D 7/2/2021 7/3/2021 Team 2 PM 0%
validation
Run data sharing Member
30 IMP Config Dynamics 6D Planned 1D 7/4/2021 7/4/2021 PM 0%
policies 1
Member
40 Take backup IMP Config Dynamics 7D Planned 1D 7/5/2021 7/5/2021 Tech lead 0%
2
Move to sandbox
50 IMP Sandbox Dynamics 7D Planned 1D 7/5/2021 7/5/2021 Team 3 Tech lead 0%
(UAT)
Backup applied Member
60 IMP Check TIME Dynamics 7D Planned 1D 7/5/2021 7/5/2021 PM 0%
to UAT 1

70

80

All key stakeholders should attend go/no-go meetings to review any

open issues, risks, and unresolved defects to collectively decide wheth-
er the application is ready for production.

The decision should be made according to the success criteria defined

during the cutover plan. For example, one of the go/no-go decision
KPIs might be the number of blockers or high-impacting bugs identi-
fied from UAT.

The cutover plan should also specify whether there are workarounds or
additional plans that might prevent any issues from delaying go live,
for example, if a blocking issue was identified during UAT but there’s
a workaround that can be implemented until a permanent fix can be
applied. The workarounds should take into consideration the impact to
end users. After all, users judge a solution by how well they can use it,
regardless of whether there’s a workaround. This is another reason why
all key stakeholders need to participate not only to define the success
criteria but also to make the final go/no-go decision.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 540
Cutover execution
The cutover execution is the series of activities carried out in adherence
to the cutover plan. Cutover execution includes the following steps:
▪ Communicate activities, milestones, and results to stakeholders in
accordance with the communication plan.
▪ Ensure that all activities are executed and completed and that any
unforeseen issues are addressed or can be addressed after go live,
have been communicated, acknowledged by stakeholders, and
documented.
▪ Execute the activities.
▪ Monitor the execution of the cutover activities.

Cutover entrance criteria

Complete the following activities before initiating the cutover execution:
▪ Complete test cycles (performance, SIT, and UAT) with exit criteria
satisfied.
▪ Build and test data migration strategy including configuration data.
▪ Identify all mock cutover and cutover resources.
▪ Complete internal and external communication plans.
▪ Ensure that the production environment is ready.
▪ Complete the final cutover plan.
▪ Ensure the final go-live support plan is in place.
▪ Ensure end-user training plan will be completed by go live.
▪ Complete data validation plans.

Cutover exit criteria

Execute the following activities to complete the cutover execution:
▪ Execute all cutover steps in sequence.
▪ Ensure the steps fit in the time window established for the cutover.
▪ Document all failures and issues and resolve or identify solutions
to resolve them.
▪ Execute data validation plans successfully.
▪ Obtain sign-off at cutover go/no-go checkpoint.

If you execute a mock cutover, it’s crucial to validate the duration time
of the different tasks and their sequences so that you can achieve a re-
alistic plan for the actual cutover. This validation also helps you identify

541
 the need for mitigation plans and whether you need to run additional
mock cutovers.

Successful go live
and transition to support
Once the cutover execution is completed, the new solution is released
to end users. At this point, the solution becomes operational, and
it’s crucial to ensure a smooth handover to the operations teams,
following the operational support plans discussed in the “Operational
support readiness” section in this chapter. For additional information
on support, visit Chapter 21, “Transition to support.”

Product-specific
guidance
Finance and Operations Apps
Operations must follow the go-live readiness review process to go live.
This review process acts as the quality gate for releasing the produc-
tion environment, which will be deployed only if the go-live readiness
review has been successfully completed.

For Finance and Operations Apps, consider the following.

Lifecycle Services (LCS)

LCS is the workspace for managing all aspects of your operations proj-
ect, including project milestones, environments, updates, deployments,
support, and project administrators. As part of the go-live readiness
review, the LCS workspace needs to be prepared with the correct con-
figuration and information such as milestone dates and project users
and accounts. In addition, create the Subscription estimator for sizing
the production environment. Familiarity with the LCS functionalities for
go live, including deploying a package, requesting a database refresh,
and requesting your production environment, will help you efficiently
run all the preparations for the go live.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 542
 ▪ LCS also contains tools such as LCS environment to monitor, diag-
nose, and analyze the health of the environment and troubleshoot
issues if they occur.
▪ Continuous updates Ensure that your environments comply with
the One Version policy. This policy specifies that an environment
must have a version no older than four monthly service updates
from the latest available one. Set your Update Settings in LCS for
continuous updates.
▪ Review the targeted release schedule to verify that the version of
your environments hasn’t reached the end of service. It won’t be
allowed to deploy production in an older version.
▪ Regression testing is needed when updating or deploying fixes or
making changes to the solution. The Microsoft Regression Suite
Automation Tool tests regressions due to continuous updates.

Upgrades
For Operations, upgrades from Dynamics AX 2012 are common. It’s
important to perform an assessment for go live with the specific re-
quirements when upgrading your solution, such as the following:

▪ Use the collation (SQL_Latin_General_CP1_CI_AS) because it’s

the one supported by the Finance and Operations apps kernel/
runtime.
▪ Run and execute all pre-upgrade checklist steps on the source
system before beginning any database upgrade activity.
▪ Store documents and attachments in Microsoft Azure Blob
Storage.
▪ Freeze all code and application configuration changes in the AX
2012 environment.
▪ If you’re using a mobile solution in the warehouse to interact with
Dynamics 365, whether you’re using the standard Warehouse
Mobile Devices Portal platform in AX 2012, you should move to
the Dynamics 365 warehouse mobile application.

Cutover
Figure 18-5 shows the ideal cutover sequence for bringing data
into the production environment and validating the data. This works
for Operations.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 543
 Fig.
18-5

Opening
Move Master data
Configurations balances in Mock go live Cutover
configurations in production Live
complete production complete complete
to production complete
complete

Configured Back up and Additional data, Load opening Simulate real-life After running From this point,
manually or with restore DB in like master data, balances with operations for a the mock go no more database
data entities/ sandbox and is added on data entities small period to live, restore movements
packages. production. top of restored to reflect a confirm solution production DB allowed, just data
configurations real operation stability in to any previous increments using
manually, with scenario or as production. reusable data entities or
data packages or final load. milestone. manually created.
integrations.

Configurations
environment
Production environment

Dynamics 365 Customer Engagement

To go live with Customer Engagement apps, key stakeholders should
perform a go-live readiness review to understand whether the plan
covers all key activities, activities that can be executed ahead of time,
and the plan to transition to operations.

Go-live readiness consists of understanding the plan that helps ensure

all recommendations and best practices are being followed to antici-
pate any unforeseen issues during the last sprint.

When engaging with the FastTrack team, this review is conducted in a

workshop format and includes all parties.

This section provides key information that enables customers and

partners to perform their own go-live readiness review.

Test and acceptance

In the Prepare phase, the functionalities should be completed (code
complete), integrations should be tested with all external systems (SIT),
and a high-level deployment execution plan should be in place. This
will trigger the solution test and acceptance activities. In this phase, the
nonfunctional requirements (NFR) of the solution that were previously
agreed to and signed off on, for example, form load times, search

Implementation Guide: Success by Design: Guide 18, Prepare for go live 544
 performance, and integration performance under the realistic produc-
tion load, should also be validated.

Several tools are available to aid with building testing automation. One
such tool is Easy Repro, an open-source library intended to facilitate
automated UI testing.

Training and adoption

To help with in-app contextual training, guided help functionalities
are available to give users a custom, in-app help experience tailored to
specific usage and business processes.

The Dynamics 365 adoption guide is an easy-to-follow how-to guide

on the best way to roll out Dynamics 365 to your organization.

Capacity management
Management of your solution is important, especially post-go live,
when migration activities are complete and users are actively using the
system. Storage usage and growth need to be validated; using storage
above the limits has an impact on some features, such as on demand
and restoring backups.

Monitor API limits because after go live, there might be integrations or

bulk operations that cause peaks in the usage, resulting in API limit errors.

Data migration
You can use several tools, like out of the box, ISVs, and custom built, to
migrate data. It’s important to follow best practices and include them
as part of your migration strategy.

Various factors can impact data migration activities, for example, the
service protection API limits that ensure consistency, availability, and
performance for everyone. Keep these in mind when estimating the
throughput and performing testing and the final data migration activities.

Requests limits and allocations can also have an impact and should be
taken into consideration.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 545
 Production environment readiness
When using Customer Engagement apps, the production environments
can be initiated at any point in time. It’s important that apps in production
are the same version as those in the development and test environments;
otherwise, issues might occur that impact the release process.

Several tools help you prepare for go live, including the following:
▪ Power Apps checker helps validate model-driven apps as well as
canvas apps and should be used throughout the implementation.
▪ Microsoft Dataverse analytics enables access to metrics that help
you monitor the solution, especially during the operations phase.

Continuous updates
In the words of Mo Osborne, Corporate Vice President and Chief
Operating Officer of the Business Applications Group at Microsoft, “To
enable businesses everywhere to accelerate their digital transformation,
we are continuously enhancing Dynamics 365 with new capabilities.”

Service updates are delivered in two major releases per year, offering
new capabilities and functionalities. These updates are backward com-
patible so that apps and customizations continue to work post-update.

Upcoming features and capabilities can be validated months in advance

with early access, before they’re enabled in production environments.

To support customers and partners throughout the release schedule,

the release plans and early access are documented and available well
in advance.
Post–go live,
management of
your solution is Conclusion
important when As mentioned at the beginning of this chapter, go live is a critical mile-
migration activities stone during a deployment, every plan and test is validated at this stage.

are complete and This chapter isn’t intended to explain how go-live readiness and the
users are actively cutover plan should be built; there are other resources that cover
using the system. these topics. The aim of this chapter was to provide a summarized

546
 explanation of the key activities for go-live readiness and their purpose
and the important takeaways, best practices, tips and tricks, and common
pitfalls that can happen during this phase and their impact on how the
new solution is received and adopted by end users.

We discussed the importance of training, adoption, and change man-

agement plans and how these plans contribute to the user experience
and adoption of the new solution.

Planning and preparing are key to avoiding delays that can impact
the deployment execution. All resources must be available in time for
activities to be executed and completed. Environments must be created,
configured, and operational on time.

Data migration can be complex. Realistic goals and good throughput are
important to ensure that there are no delays that might impact the go
live date and that all tasks are executed according to the data migration
plan. Ensure that the recommended practices to improve throughput are
in place. Validate that throttling limits aren’t being reached.

There should be coordination between the implementation team and

external dependencies to make sure that expectations are aligned
and requirements met. Any external solutions must work as expected
with the current version of the solution and platform and be aligned
to the roadmap.

Support during cutover, go live, and post-go live is essential for all
activities to run in the expected time. It’s crucial to have resources
available to mitigate any issues so that tasks can complete on time
or with as slight a delay as possible. A poor support strategy for a
successful rollout will impact the success of the rollout over time. It’s
important to create a list of responsibilities for first, second, and third
lines of support as well as an escalation path and to transition knowl-
edge from the implementation team to the support team. Ensure that
there’s coverage outside traditional working hours and on weekends
for critical situations.

Ensure that application and platform monitoring is in place, and set

Implementation Guide: Success by Design: Guide 18, Prepare for go live 547
References up appropriate alerts so that the operations team can take necessary
actions and notify users. A strong monitoring and communication plan
GitHub - Microsoft/EasyRepro: Automated UI has major impact on adoption and confidence in the solution. All crit-
testing API for Dynamics 365
ical application components should have a logging mechanism with a
Use solution checker to validate your model- well-defined process to monitor the system and alert administrators.
driven apps in Power Apps

Microsoft Dataverse analytics

A process needs to be in place post-deployment and post-go live to
Microsoft Dynamics 365
manage changes (such as upcoming updates or new versions of ISV
Create guided help for your Unified
solutions) and prepare users for changes and the adoption of new
Interface app
features and capabilities.
Dynamics 365 Adoption guide

Accelerate Dynamics 365 Adoption checklist

Build a champion program

Dynamics 365 release schedule and early access

Prepare for go live

Report production outage

Targeted release schedule (dates subject to

change)

One Version service updates FAQ

Monitoring and diagnostics tools in Lifecycle

Services (LCS)

Evaluate go-live readiness for Dynamics 365

Customer Engagement apps

548
 Checklist

Go-live readiness Ensure the production environment is prepared and

administrators and IT are familiar with monitoring,
Assess go-live readiness early and often through quality
troubleshooting procedures, and avenues of support.
gate reviews to provide clarity on the status, readiness,
completeness, and maturity of each activity and to help Create a cutover plan that considers all dependencies,
project stakeholders and the steering committee make the timing of each activity, roles and responsibilities,
go/no go decisions. instructions, verification steps, and any additional
documentation associated with each activity.
Ensure that the solution scope is aligned with the solution
that’s going to be live and that it has been communicated,
shared with stakeholders, and signed off on by the
Cutover
business, agreeing that the expected scope is covered. Implement the cutover plan by completing all activities
in sequence, satisfying the exit criteria, and including the
Ensure that the solution has completed all test cycles,
sign-off at the cutover go/no-go checkpoint.
which include system integration, performance, and
UAT, with exit criteria satisfied and signed off on by the
business stakeholders.

Ensure all scripts and processes planned for the cutover

migration are tested and signed off by the business
stakeholders.

Ensure external dependencies such as ISVs and

third-party systems and services are aligned with the
timelines and scope for go live.

Plan all change management activities such as training,

user support, user engagement, and communication,
and implement all pre-go-live change management
activities from Initiate through Prepare phases.

Have a plan for monitoring and maintenance routine for

production as well as for transitioning to support teams.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 549
 Case study
The gift:
go live in a cloud solution
A toy company recently embarked on an implementation project and
migrated all their processes from on-premises to Dynamics 365. They
had integrations with Dynamics 365 Customer Engagement, Microsoft
Power Apps, and Power BI. The company plans to implement Dynamics
365 Commerce and Warehouse Management mobile app in Microsoft
Dynamics 365 Supply Chain Management in future rollouts, but they
needed to move all financial data and their CRM system to the cloud in
time for the end of the year, their busiest time.

At the end of October, the team was excited and a little nervous
about the approaching go live. The impact of changing from a system
in place for many years would be huge. The team was used to their
manual and inefficient processes but were ready to move to a system
that would automate and improve processes and increase revenue
growth and profitability along with scalability. Change management
was critical for success.

The implementation team needed to make sure that everything was ready
for going live. With go live four weeks away, it was time to start review-
ing their readiness for the implementation. SIT was completed, and UAT
and performance testing were almost completed. The system seemed
performant, although not optimal, it was considered ready for production.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 550
 The team was ready to start the go-live review, to switch to the Prepare
phase. They prepared the go-live checklist and asked themselves the
following: Everything was thoroughly tested, but was the system really
ready? Were the users ready? Was all the infrastructure ready? Were all
the systems up to date? Was the mock cutover planned? Were the ISVs
working well?

Concerns came up over risks and issues identified during the as-
sessment. For instance, they realized that the mail service and other
services were in another tenant, so they needed to perform a tenant
move to enable single sign-on for their users. In addition, Microsoft
was planning an automated version release by their go-live date, so
they needed to update their environments to the latest version. This
would necessitate conducting another smoke test and verifying that
the ISVs worked correctly with the version. They missed dates because
they were focused on UAT and making fixes and addressing new re-
quirements identified during the testing phase.

The team realized that open transactions weren’t ready to be closed in

the old system and therefore they created them in the new one. The
data migration plan wasn’t solid. In addition, performance was weak,
so the team needed to troubleshoot the code.

Business stakeholders made a go/no-go decision after this assessment.

It wasn’t an easy decision, but they decided not to go live. Despite all
the heavy lifting they had done, they didn’t have a mock cutover plan,
which they realized was necessary to complete the readiness process.

Would they need to change the go-live date that was so close to the
holiday season?

What could they have done better? They couldn’t have changed the
date because they needed to be ready for the holidays, but they could
have planned the go-live date earlier so that they had more time for
the ramp-up and to address any delays or issues. They could also have
had an earlier assessment review, with UAT almost complete. The
timelines were tight, and they had the choice to go live with what they
had, which was risky because that might mean stopping operations in

Implementation Guide: Success by Design: Guide 18, Prepare for go live 551
 the middle of the busiest season, or move the go-live date to January-
December wasn’t an option. A third option was to hire more resources
and work more hours to try to deliver on time, which was risky because
the ramp-up would be during November. A big-bang go live wasn’t
optimal for this situation.

The team learned that it’s important to start the readiness review on
time and to schedule sufficient time and resources. Additionally, it’s
crucial to have a solid support plan for production. The importance of
the Prepare phase also shouldn’t be underestimated; plan with enough
time to mitigate any issues and risks.

Implementation Guide: Success by Design: Guide 18, Prepare for go live 552
19
Implementation Guide: Success by Design: Guide 19, Training strategy
Guide
Training
strategy
553
“There is no saturation
point in education.”
– Thomas J. Watson, founder of IBM Corporation

Introduction
At its core, the goal of a training strategy and training
is to ensure that all necessary users of your system
are educated on the new application. That way their
knowledge of how to complete their work results in
successful user adoption following go live.
Training is not the only factor in meaningful user education and adop-
In this chapter, we define a training
strategy and determine at a high level
tion; empowering users to perform their necessary tasks in the system
what components you need to embark correctly and efficiently should be the ultimate “why” in the develop-
on a methodical approach to a successful
training execution for your Dynamics 365 ment of a training strategy. As part of user adoption, organizations
implementation. Each of these areas is
should strive to train in a way that gives users confidence in the appli-
covered in detail in this chapter.
cation and inspires a sense of delight when using the system.
For a successful training
strategy consider these
In this chapter, we cover several high-level objectives, as well as exam-
main areas:
• Training objectives
ples of more organization-specific objectives, that should be included
• Training plan in your organization’s training strategy. A proper training strategy
• Scope
• Audience should center around the creation and execution of a comprehensive
• Training schedule
training plan. Furthermore, the training plan should align to the broad-
• Training material
• Delivery approach er training strategy of your Microsoft Dynamics 365 implementation.
• Assumptions, dependencies, and risks
• Training as an ongoing process
We discuss how to create an appropriate scope for your organization’s
• Training best practices training as part of this plan, as well as how to confirm the different
groups of users that need to be trained.

Implementation Guide: Success by Design: Guide 19, Training strategy 554

We cover training topics that should be part of your organization’s
overall project plan, and how best to keep your training plan and
project plan coordinated to avoid significant issues. This chapter also
dives into guidance surrounding different types of training material
and their effectiveness, as well as how to create a training delivery plan
and conduct effective training that can serve all your users (before and
after going live).

Dynamics 365 is an ever evolving and growing application. Because of

this, training is an ongoing process that requires constant adaptation
and updates, and we cover how to best prepare your organization for
this continual update process. We also recognize that training strategy
and guidance can vary depending on your choice of Dynamics 365
application. We provide suggestions on how best to successfully adopt
a training strategy, whether your organization is implementing appli-
cations built on Dynamics 365 Customer Engagement, Dynamics 365
Finance and Operations, or both.

Lastly, we discuss best practices by highlighting the behaviors and

strategies that we have seen organizations successfully employ when
developing a training strategy. We share our collective experience
regarding these best practices toward the end of the chapter, in the
hopes that you will carry them forward in your own projects.

By following the guidelines laid out in previous chapters of this book,

your organization should be well on its way to rolling out a successful
Dynamics 365 application that fulfills the needs of your business and
users. Successfully training your users on using modern technology,
as well as understanding business processes that they will need to
incorporate to do their jobs safely and efficiently, is integral to any
organization’s training strategy.

Training objectives
One of the first things that your organization should consider when
beginning to develop a strategy surrounding training users, and a
plan behind it, is the objectives. Defining the objectives of a successful

555
 training strategy is key, and it can help shape the crafting of a training
plan as well as lead to a more successful rollout of training itself.

At a high level, every Dynamics 365 application training strategy

should include some version of the following objectives. Many change
management objectives can also be addressed by ensuring that your
training strategy explicitly addresses these training-focused objectives:
▪ Employees will gain necessary knowledge required to use
Dynamics 365 applications as a part of their day-to-day jobs in a
safe and efficient manner.
▪ Training should be meaningful and not overwhelming, to provide
users with an efficient way to do their jobs.
▪ Training should help support business goals for the application
(in other words, employees understand the “why” behind training
materials and scenarios).
▪ Training should be inclusive of everyone (for example, experienced
users as well as new hires) so that all users of your application
receive the education necessary to do their jobs.
▪ Training should be an ongoing process that does not stop when
the system goes live.

Your Dynamics 365 implementation will have additional training ob-

jectives that are specific to the applications you are developing, and
to your organization. It is important that you work to define these
objectives as they, along with the more general objectives of training
outlined earlier in this chapter, should formulate the “what” of your
training strategy goals. An effective way to evaluate your organiza-
tion’s training goals is to follow the SMART acronym, that is: specific,
measurable, achievable, realistic, and timely. While it might not be pos-
sible for every training objective to meet every one of these metrics,
we recommend that you create training objectives that are relevant to
your organization’s goal of user adoption and fit the outlined criteria.

What not to do
As a counterpoint to the recommended strategies, let us spend a little
time discussing the training objectives that an unprepared organiza-
tion might create.

Implementation Guide: Success by Design: Guide 19, Training strategy 556

 In this example, let us say the organization creates the following objec-
tive: “Objective: Dynamics 365 Field Service users should be trained on
the mobile application.” While this objective captures at a basic level
what needs to be done, it falls short of accomplishing what a proper
training objective should. When rewritten using the previously men-
tioned criteria, and separated for increased detail and specificity, the
objectives become much clearer:
▪ Objective: 100 percent of Field Service technicians attend the
in-person mobile training session or view the complete online
seminar series prior to system cutover.
▪ Objective: All Field Service technicians finalize the Work Order
completion process on their mobile devices without assistance and
are able to complete the lead-generation process on their mobile
devices in under three minutes.

These rewritten objectives also serve as good examples of making sure

your training objectives specifically relate to the application being
deployed. A poorly written training plan often does not include ap-
plication-specific objectives; in addition to not following the SMART
philosophy, these objectives are not thorough enough to capture the
essence of the goals of training. Clearly defined and measurable train-
ing objectives also help an organization determine the effectiveness
of its training. It is significantly easier to judge the success of training
in your organization if you can track against objectives that have been
well-written. Later in this chapter we discuss specific ways to assess the
effectiveness of your training; this process starts with creating proper
training objectives.
It is significantly
easier to judge Ensuring that your users are comfortable with the most challenging

the success of obstacles to user adoption is important to achieving training success.

While your training objectives do not need to specifically reference
training in your tasks or challenging business processes directly (as these would be too
organization if you specific and thus not useful), they should reflect knowledge of areas of
can track against your application that may require additional attention from a learning
perspective. The mobile objective in the previous paragraph is a good
objectives that example of this strategy because we recognize that Dynamics 365 mo-
have been well- bile apps can represent a significant change in business processes for
written. users and, therefore, could require additional attention from a training

557
 and user adoption perspective. Thus, a key objective should be that all
users receive proper training on this specific job function.

Creating a training plan

Proper training is critical to user adoption. Organizations must develop
a training plan at the start of the project and engage resources from
the beginning.

It is important to remember that managing a training plan is an iterative

process. As you progress in the project and learn more about your busi-
ness requirements and training needs, you might have to update your
plan. This sample training plan charter gives you something to start with.

Fig.
19-1 Training plan elements We recommend you put together
a training plan that, at a high
level, consists of at least the
Training following elements (as illustrated
objectives
Training
Scope
in Figure 19-1):
resources
▪ Training objectives
▪ Scope
Training as
an ongoing Audience ▪ Audience
process
▪ Training schedule and resource
availability (project planning)
Training ▪ Delivery approach
schedule and
Training
materials Training plan resource
availability
(project
▪ Validation of training success/
planning) objectives
▪ Assumptions/dependencies
▪ Risks
Training Delivery
environment approach ▪ Training environment
management
management

Validation
▪ Training materials
of training
Risks
success/
▪ Training as an ongoing process
Assumptions/ objectives
dependencies ▪ Training resources

Implementation Guide: Success by Design: Guide 19, Training strategy 558

 Just as the project plan is key to project timelines and execution, so too
is a training plan key to training timelines and training execution. End-
user training is one of the last tasks on a project with dependencies
on several key items, such as a near-complete solution, good quality
representative business data, and reserved time from end users. If a
project is running even a little late, the final preparations for training
often get squeezed. A robust, well-thought-through training plan that
reduces the number of items in the critical path is essential to avoid
last-minute compromises and lower-quality training outcomes.

Scope
A crucial element of ensuring that your organization’s project has a
proper training strategy is defining an accurate scope for the training.
Here are different areas that cover the breadth of scope that you will
need to build into your training plan:
▪ Business processes completed by end users, derived from your
user stories and functional requirements
▫ Example: Users must be able to create and process an invoice
from end to end in for Finance and Operations apps.
▪ Any nonfunctional requirements that require training users for the
requirement to be met
▫ Example: Users should have the ability to provide feedback
directly from the application.
▪ Any administrative tasks that users might need to accomplish
while in the application, separate from their business-driving work
▫ Example: Users must be able to configure and use the
Advanced Find tool inside Dynamics 365.
▫ Example: Users should be able to upload templates using the
Dynamics 365 data management tool.

There could be other areas of your application that should fall into the
scope of your training. If there is a task or process that your users need
to accomplish while working directly with your Dynamics 365 appli-
cation, it should probably fall into the scope of your training, as you
should assume there are users who need to be trained on that topic.
An example of this would be training project team members during

Implementation Guide: Success by Design: Guide 19, Training strategy 559

your implementation who will need to work inside the application
from a development or configuration perspective.

Once you have the full set of business processes, requirements, and
tasks, you can use them to create a true scope, consisting of actionable
training areas, for your training. There might not be a one-to-one re-
lationship between them, as there may be overlap between processes
as well as additional training topics that need to be added. This is our
guidance on how to avoid overlap and how to add more training topics.

▪ Avoiding overlap  If you have 10 sets of business processes that

are different in terms of function but almost identical in terms of
system execution, and each takes 30 minutes for a user to com-
plete, consider setting your system training areas to represent the
similarity between completing these in your application, rather
than to represent the disparity between business processes. You
can still reference different business functions and even take time
to walk through any differences in execution, but it is important
that the scope (that is, each training area) is framed in such a way
to be efficient and not duplicate training effort.
▪ Additional training topics  If you require users to switch be-
tween their mobile and desktop devices frequently, with minimal
delay between each, it might be important to cover this transition
even if it is not necessarily covered in any business processes or
other requirements or tasks.

The next step in setting the scope for your training is to categorize
and prioritize the list of training areas that you have created. As an
example, some business processes in your new application might
be different from the “as is” processes in your organization’s current
application. This is important for a few reasons:

▪ System training and business process training are two separate

efforts, even though training might address both types.
▪ Training surrounding changed business processes should be given
special attention in training sessions and in training material.

It is common for business process training challenges to affect system

560
 training effectiveness. As part of your scope of training, and combined
with your training materials and delivery, you will want to separate
system training on the new processes and application from topics that
veer more into organizational change management.

Also, you should prioritize, or at least highlight, any training related

to critical nonfunctional requirements that have impacts on legal
regulations or company policies. For example, if your organization is
legally required to fill out fields on your sales order form in a particular
manner, this section of training should be emphasized so that users are
aware of its importance.

One process that might help determine priority of training areas is

the creation of a change-impact assessment, whereby you detail the
changed areas, the topics and stakeholder groups that are affected,
and the impact that the changed areas will have. This will allow you to
have a consolidated list of training topics, an estimated list of people
that will need training on that topic, and the beginnings of a prioritiza-
tion around each.

Figure 19-2 shows an example of a change-impact assessment.

Fig.
19-2

Change impact Training topic Impact (1-5) Action

The purchase department’s Procure to pay 3 Trainers will hold a series of

legacy system will be replaced business process webinars detailing differences
with Dynamics 365 Supply between the legacy system used
Chain Management and for purchasing and Dynamics
business wants to use the 365 Supply Chain Management
standardized process in
Dynamics 365

Field technicians are moving Using the new 4 Because of the impact this
from a pen-and-paper-based mobile application change has, multiple live
work order management system to complete a work training sessions will be held for
to a mobile application order technicians to ensure proper
education

Implementation Guide: Success by Design: Guide 19, Training strategy 561

 Audience
Now that the objectives for your organization’s training strategy have
been defined, along with the scope of content that training should
contain, we can define the next piece of the training puzzle: who to
train, and what type or types of training they should receive.

Who receives training?

It is critical to identify the different groups of people who need
training. Dynamics 365 offers excellent role-based access control
and application segregation as a means of providing access to your
users based on how they need to use your application. As a result, it
is important that all employees who will need to use the product are
trained in the correct way, on the correct topics.

The following section identifies the core groups of people who should
be trained on your Dynamics 365 application. Note that each group
of users can be broken into subgroups. Breaking down these users
into personas is important. If personas are incomplete or incorrect,
your organization might not complete proper training on the correct
business processes or, worse, it might not complete training at all for
specific user groups.

System end users

System end users are the people who will use the system daily and
drive direct business value. Your application could have multiple types
of end users, separated by licensing, security role, or even the applica-
tion itself (Customer Engagement apps versus Finance and Operations
apps users). Each of these groups will require different training. This
group of users might cover multiple departments and job functions; as
a result, the training should both be general, as to provide high-level
knowledge of your application regarding key core entities that all users
may need to access, as well as specific to aid these users in their specific
job functions.

As part of your training plan, be sure to identify each business role that

Implementation Guide: Success by Design: Guide 19, Training strategy 562

 will need to use your application differently and will require training.
This is related to, but distinct from, creating groups of users based on
security role. Certain personas might have identical security roles in
Dynamics 365 but different job functions; these groups of users will
need to use your application differently and will, as a result, require
different training.
Fig.
19-3
As an example, consider a salesperson and a sales team
leader (see Figure 19-3). Even if your organization
Salesperson Sales team lead has defined that their required access in your
application is similar enough to warrant identical
• Creating records security roles, it is important that these groups of
• Updating records users are treated distinctly during this portion of
• Sending emails your training strategy and receive correct training.
Salespeople will need to be trained more on the
“doing” functions of your application, like creating
• Approving records records, updating records, and sending emails.
• Creating and
viewing charts Sales team leaders, on the other hand, might work on
• Dashboards
similar tasks as salespeople, and they might require some
or all of the same training, but they will also need additional
types of training, such as reviewing and approving records, and
creating and viewing charts and dashboards.

Another benefit of being specific when defining training personas

based on business role rather than security role is the value it can
provide for new hires. Having a new hire in a specific business role un-
In many organizations, certain end user
dergo an exact set of training based on what was identified previously
roles have higher turnover than other
personas receiving training, so your in this chapter can make onboarding significantly easier and reduce
training material should be structured in a
way that can help staff onboard easily.
time in trying to match job titles to security roles, and then back to
business processes in your application.

Trainers
Trainers are a special category of users that require training. They
should be knowledgeable in all areas of the system and are responsible
for developing training material and training other users. “Train the
trainer” is a common approach to onboarding this type of user to your
application. Depending on the size of your organization, you might

Implementation Guide: Success by Design: Guide 19, Training strategy 563

have several trainers. Since the goal of your trainers is to create materi-
al and conduct training for your other users, it is important that at the
end of “train the trainer” sessions, your trainers are aware of and able
to perform their job function (training other users) effectively. The goal
of training these users is different from that of your end users; trainers
should be part of the training strategy process and will be advocates of
the new application.

Super users
Super users are an elite group of (usually) end users who act as cham-
pions of the application. This group is key to driving excitement and
adoption of the new system. These users are often involved in early
feedback cycles with the project team. As subject matter experts (SMEs),
they become educated on the application early on so that by the time
end-user training begins, your super users are experts in various business
processes. As part of your training delivery plan, your organization can
use super users strategically (by having them attend training sessions
and serving as application advocates) to aid your trainers in their job
functions. Additionally, these super users can act as a “first line” to help
answer employee questions, keeping your support desk from being
overwhelmed and at the same time keeping morale high.

We have mentioned several times that meaningful user adoption is

critical to the success of an application and relies heavily on a well-ex-
ecuted training strategy; super users are instrumental in this process.
If your organization is deploying several different applications, it is
recommended to have multiple super users, spread across each appli-
cation, and corresponding to their area of expertise.

Project stakeholders and executives

If your project stakeholders and executives need to be able to access
and use your application, they need to receive training as well. This
training should be tailored to their specific role.

Executive buy-in is important for the success of a new application, and

meaningful user adoption applies no differently to this group of users
than it does to your end users. To take advantage of the additional
effect that executive buy-in may have, you should consider “visibly”

564
 training these users so that other members of your organization are
aware that training is ongoing. You can also notify these stakeholders
of training that other roles are expected to undergo, which can help
from a buy-in perspective.

Support desk and administrators

Another critical group of people that will need training will be your
support and administrative users. While these users might need a
certain amount of training on the application and certain business
processes, they will need to become familiar with the application from
a maintenance perspective. Training for these users will focus on tasks
related to system maintenance, troubleshooting, and administration.

Training scope
Once you have defined the distinct groups (and subgroups) that
should receive training, you should work on matching the different
training areas defined in the Scope section of this chapter with the
groups of users defined earlier. Each group of users will require a spe-
cific set of training (based on training areas). We recommend creating
a matrix in your training plan, with training role against training subject
area, that you can refer to when creating training materials, planning
training delivery, etc.

When determining which personas need which types of training, refer

again to your user stories as a baseline. You should also consider which
groups need training in non-functional areas and administrative areas.
You should structure Note that certain groups of users might require different levels of
your assignment of training on identical subjects. For example, trainers might require only

training topics to basic training on Dynamics 365 administration (to assist other users
during training and to implement training tools), while administrators
users in a way that is and support desk personnel might require advanced training on the
comprehensive but same topic. You should structure your assignment of training topics to
not superfluous. users in a way that is comprehensive but not superfluous.

565
 Training schedule
Successful integration of a training strategy into the main project
effort requires coordination with project management resources. Your
organization’s training strategy does not exist independent of other
ongoing efforts in the project, and it is important that key milestones,
resource requirements, dates, and tasks are acknowledged and ac-
counted for as part of your training plan and are incorporated as tasks
or dependencies in your overall project plan.

Regardless of the size of your organization and complexity of the

application being deployed, you must consider the following train-
ing-related factors. We have included here examples of common
questions project teams encounter as a part of their training initiatives.
If these questions are applicable to your project, you will need to
address them as part of your training strategy and project plan. Note
that this is not a comprehensive list of questions that your organization
will need to answer, but this should help you get started:

Project team resources

▪ What is the availability of the trainers and super users?
▪ Who are the super users, and if not defined, when will that decision
be made?
▪ Are there external resources involved in this process that will
require onboarding or administrative setup before they can begin
their work?

Trainees
▪ Who is necessary to take part in training activities, and when will
these activities happen?
▪ Are there any resource dependencies (e.g., Resource A is required
before Resource B)?
▪ How will training, and attendees, be staggered so that there is no
gap in coverage for day-to-day business functions?

Dates or key milestones

▫ When should training begin for distinct groups of users?

Implementation Guide: Success by Design: Guide 19, Training strategy 566

 ▪ What dependencies do these trainings have?
▪ Are any project-related activities dependent on training?
▪ Is training reliant on the creation of material or development work,
whose schedules could shift?
▪ What milestones should training include, and are there any im-
pacts of these milestones?

Training plan
▪ When will the training plan be completed?
▪ Who will review it, and what will be the plan for updating it if
Fig.
19-4
necessary?

Question
Training execution and feedback
Are any project-related activities ▪ When, and where, will training occur?
dependent on training?
▪ Will training have an impact on other project efforts?
Answer
▪ If there are subsequent trainings, how will feedback be evaluated
and what is the cycle for processing this feedback?
We often recommend conducting training
prior to your user acceptance testing (UAT).
The major benefits of this order are: As you can see, many questions related to a training strategy have an
• During UAT, users are already aware of impact on, or are reliant upon, activities that occur separate from tradi-
how to navigate through the system and
accomplish tasks. And as a result, there tional training activities. In Figure 19-4, we examine one question from
should be fewer items logged as bugs that
this list to underscore the importance of asking and getting answers
should be listed as training issues.

• UAT serves as a measure to validate the

for these questions as part of your project planning.
effectiveness of training. If a large number
of training-related issues are identified
during UAT, it may be necessary to make This example shows why it is crucial that your training plan and overall
updates to your training material and
training strategy be reflective of and connected to your organization’s
even potentially conduct future training
on topics with which users are struggling. project plan, and that updates to either plan are reflected in the other
so as to not create conflicts, which can lead to project delays.

Training materials
There are multiple consumable types of content for Dynamics 365
applications that your organization can create. Depending on your
project, you could create multiple types for use in different trainings.
Here are a few common examples of the types of training materials
that an organization could create.

Implementation Guide: Success by Design: Guide 19, Training strategy 567

Documents
Your organization can create written documents that guide users
through the application and educate them on business processes.
These documents can also serve as reference material. This type of
material is the least interactive but is the easiest to create and update.

Videos
Trainers can create videos that explain key training areas in detail and
walk through the application, providing guidance on how users can
navigate the system and complete their job tasks. Videos can be more
helpful than documents for many users, since they “show” rather than
“tell” certain features of the application. Videos can also be recorded
in small chunks that represent discrete processes in an application, as
opposed to creating one longer video that walks a user through an
end-to-end flow. This benefits trainers from a maintenance perspec-
tive; instead of having to rerecord the entire video when changes are
made to your application (and create rework around having to record
training for concepts that have not changed), they can focus on updat-
ing specific videos around new features and functionality.

Microsoft Learn
Microsoft has published several labs via the Microsoft Learn portal
that provide a wealth of knowledge on Dynamics 365 applications, as
well as other applications on which your organization might require
its users to be educated. Microsoft Learn allows the user to follow any
number of predetermined paths specifically designed for different user
personas. Individuals can also create their own knowledge path that
is specifically tailored to their learning requirements. Users can learn
on their own schedule and can browse videos related to topics when
necessary and engage with other community members.

Guided help
Dynamics 365 includes the ability to create custom help panes and
guided tasks, out of the box, to assist your users in walking through
the application and completing basic tasks in the system. Guided help
is easy to set up and implement and does not require an additional
solution on top of your application. Additionally, Dynamics 365 ap-
plications can install the Guides application, which allows for visual or

568
 holographic displays to show step-by-step instructions on tasks that
your users need to perform.

Virtual material via a digital adoption platform

Built on top of Dynamics 365 and allowing users to directly interact
with the application while training, digital adoption platforms are the
most interactive form of training material. This training material can
be easily updated, and once published, it has an immediate effect for
all users who need to view it. It reduces the need to update physical
documents or re-record parts of video material. Digital adoption
platforms can come in multiple flavors. One effective method is a tool
that lets trainers create multiple “scripts” that walk a user through a
business process in Dynamics 365. For example, say that all users must
learn basic navigation through an account record, but only system
administrators must know how to customize a price list. Earlier in this
chapter, in the “Scope” section, we talked about reducing overlap;
using digital adoption platforms is an excellent way to manage scope
in the creation of training material. Instead of having to explain the
same material twice (in the case of video trainings that cover similar
subjects) or create multiple reference guides or a single complicated
guide (in the case of written training material), a trainer can create one
set of scripts and apply them to different groups of users.

What to consider when

selecting training material
Be thoughtful when choosing the type or types of training material to
create for your training audience. Consider the benefits and drawbacks
of each. For example, if your end-user base consists mostly of users
who do not have considerable experience working with web-based
cloud applications, you might want to consider a more hands-on
approach, a digital adoption platform, to help train your users. On the
other hand, system administrators and support desk users might find
documents more helpful, as they would rather have the large volume
of written information that do not require the more hands-on training
that digital adoption platforms provide. When creating training con-
tent, keep in mind that training is ongoing. Your organization can start
to create training material and training early on in your project, and

Implementation Guide: Success by Design: Guide 19, Training strategy 569

 then refine it later with minimal rework needed. In the world of cloud
applications, where software is constantly changing, it is important
that your training strategy prioritizes being adaptable, so that you can
update materials easily as a response to application changes.

You should also consider language and accessibility needs in your

organization. If you have a multi-geo or multi-language deployment,
any training material you create might need to be available in different
languages. This material must also comply with any federal, state, or
local regulations regarding accessibility so that all users have an equal
opportunity to consume effective training material.

Who will develop and review the content?

During the creation phases of your project plan and training plan, you
will need to identify the trainers, SMEs, and business process owners
who will assist in developing the training content. Trainers will usually
create content, which will subsequently be reviewed by the SMEs and/
or business process owners. Depending on the importance of the
course, as discussed earlier, you might want to consider bringing in
multiple people to review the material.

How long will it take to develop the content?

Your organization must make sure to allocate time in the project
schedule and training plan to design, develop, and review content. The
length of time required to create training content will depend on the
complexity of your application and the type or types of content to be
developed. To estimate the time, we recommend that you identify the
scope of training areas and subsequent material that must be created
based on those areas, and the format or formats the training material
must take. You can then estimate total time to create training material
based on experience or an extrapolation calculated on how long it
takes a trainer to create a subset of that material.

When is the content needed?

Your project plan should list when content creation begins and your
training plan should specify when training material will be available to
certain groups of users. These two schedules should be coordinated
to ensure that content creation begins on time, and to ensure the

Implementation Guide: Success by Design: Guide 19, Training strategy 570

 material is ready when needed. End-user training can begin roughly
when development is completed (that is, when the system is code
complete and no changes to the application are anticipated).

Where will the content be stored?

If more than one We recommend your organization employ a content storage control
person will be method, especially if multiple types of content are being created for
editing content, groups of users. Training content should exist in a single centralized
source instead of being passed around by email or shared links. This also
your organization ensures proper source control and the ability to institute access control
might need a for your user groups. Additionally, if more than one person will be
versioning strategy. editing content, your organization might need a versioning strategy.

Delivery approach
As a subset of your training plan, your organization should create a
training delivery document that contains specifics about the actual
execution of training itself. While many of these topics are covered in
the overarching training plan, it is important to have a specific plan
for training delivery to ensure that your organization is prepared to
execute training.

We will now discuss the topics that your training delivery document
should cover.

Who will be involved, and how?

We have defined the audience for training in the “Audience” section
earlier in this chapter and explained the details of incorporating re-
source scheduling into your project plan in the “Training schedule”
section. To help you create your training delivery document, we will
go a step further in organizing these resources into training sessions.
This is important if your organization will schedule multiple training
sessions at the same time. In addition to allocating your trainers and
trainees, you will also need to assign them to the correct training based
on the content to be covered.

Implementation Guide: Success by Design: Guide 19, Training strategy 571

 What methods of training should the organization conduct?
Like training material, diverse types of training suit different situations
and users, so it is important to consider which training or combination
of trainings best fit your organization. Consider the advantages and
disadvantages of different methods. Figure 19-5 shows common types of
Fig.
19-5 training, and the anticipated advantages and disadvantages of each type.

Type Most effective when Advantages Disadvantages

Live training The content is critical for Better interaction Scheduling challenges.
(in person or business function and you between trainer and
want strict control over participants as well as In-person variants of this type of
virtual)
delivery. between participants. training require a physical room with
computers, since Dynamics 365 is a
Content contains non- Immediate feedback. web-based application.
functional requirements,
or business processes that Collaboration in business In-person variants of this type of
require specific devices. processes is significantly training are limited to the number of
easier (i.e., front office people per room, per session.
and back office need
to work together to
complete an order).

Interactive Content is moderate in Web-based content Trainings are less effective than in-
web-based difficulty. can be built once person, since there is less interaction
and consumed by an and the end user faces more
training
Business processes are unlimited number of distractions.
best learned through users.
repetition and ongoing Trainings are not usually real-time,
training, since users’ People can train on their meaning any questions or challenges
access to the application own. No scheduling that users face during training might
is necessary. requirements and no not be answered quickly.
physical location needed.
Web-based training, when built
with digital adoption tools, can be
costly and also require technical
maintenance.

Self-paced Content can be easily Written and video This is the least interactive form of
training explained in a user content can be updated training, meaning users who rely on
manual or video series. quickly as future releases interaction either with a trainer or
and updates occur on with a web-based application might
Content is best your Dynamics 365 struggle to learn when having to
consumed as written application. read/watch content.
reference material that
can be used as needed. Training content is easiest
to create.

Implementation Guide: Success by Design: Guide 19, Training strategy 572

 When should training occur?
As a part of your training delivery plan, you should decide when dif-
ferent types of training will occur. In an earlier section of this chapter,
we discussed how super users and trainers should be trained in nec-
essary areas of your application earlier than your end users; this type
of consideration should be built into your training delivery plan. Plan
to conduct training for your end users late enough in the project—
post-development and pre-UAT—so your users do not forget concepts
and tasks that they have been taught during training, but also be sure
to leave enough time after training occurs to accommodate any delays
in your project or training schedule.

We believe that it is never too early to begin training super users and
trainers. Since the main goal of training these users is to help them
become familiar with your application early in your project, you should
start conducting training with them (whether formal or informal)
during development and continue it as an ongoing process. These
early training sessions should familiarize super users and project team
members with basic application functions and serve as a backbone for
future train-the-trainer sessions.

Figure 19-6 is a high-level sample schedule that shows training ses-

sions for each important group that needs to be trained, alongside key
relevant project milestones.
Fig.
19-6

January February March April May June

Super-user training session #1 conducted Train-the- Go live

trainer
Train-the-trainer session #1 conducted session #2 UAT Support desk
on first half of sprints conducted conducted training occurs
based on
Trainers begin work on training material remainder of
Stakeholder/executive training occurs
development
Design, development sprints begin work
End-user training
Project kick-off Code Super-user training session #2 conducted
Demonstrations/Initial training conducted for super users, complete
trainers, and project team members

Implementation Guide: Success by Design: Guide 19, Training strategy 573

 Trainers and project team members should be aware of ways in which
training differs depending on the size of your organization. Larger
organizations with more users are likely to require multiple personas
and an increased number of business-use cases. Larger organizations,
especially international companies, or companies with numerous
locations, might require different training sessions and venues, and
they might need to occur at different times in your project. In addition
to making sure training materials are available in multiple languages,
employing multiple trainers or multilingual trainers is important.

In the sample project plan seen in Figure 19-6, end-user training takes
place over the course of a single week, in a single location. If this or-
ganization were international and multilingual, most of the items in
the plan become more complex. To illustrate the increased complexity,
we consider what end-user training might look like in a multinational
Fig.
19-7 organization in Figure 19-7.

April
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Trainers work to End-user training #1 End-user training #2 End-user training for field
develop/translate Conducted in the US, Conducted in the US, technicians
material into English, France and Spain France and Spain Conducted in France
French, and Spanish
Training material Environment cleanup Environment cleanup
(including translations)
are validated

As this figure illustrates, training schedule and planning complexity

increase dramatically when the scope of the implementation increas-
es. Instead of taking roughly a week, this updated training schedule
now takes almost a month, in three different countries/regions, and
requires drastically increased coordination of resources. If additional
phases with different business process areas need to also be covered,
this would further increase the complexity of training. We hope these
examples are useful in demonstrating the value of planning the when,

Implementation Guide: Success by Design: Guide 19, Training strategy 574

 where, and how of training and why it is important to understand that
your organization should be prepared to make additional investments
to train its users successfully.

How should your organization set up and maintain a

training environment?
If your organization has created a Dynamics 365 environment to train
users as part of a broader environment management strategy (cov-
ered in Chapter 9, “Environment strategy”), using application lifecycle
management (ALM) best practices (covered in Chapter 11, “Application
lifecycle management strategy”), you will want to make sure this train-
ing environment is deployed with code and data that is representative
of production use and is of adequate quality to meet your training
objectives. Both are extremely important points:
▪ If your organization’s training environment does not have up-to-
date code and application features, your users will be left confused
and frustrated, which could decrease the overall effectiveness of
your training. Your users might feel the training was unprofessional
or incomplete, especially if the quality of the training environment
is insufficient to accomplish training objectives.
If your training environment contains
▪ If your organization’s training environment does not have suffi-
only accounts named “Test 1” and cient and relevant test data, your users might fixate on the quality
“Test 2,” it is unlikely that your
users will enter realistic data when of the data and not on the training material. This will lead to
progressing through their training
improperly trained users from multiple standpoints:
scenarios. If users do not understand
the implications of entering fictitious ▫ Because users are distracted, overall training effectiveness will
data, this could have wide-reaching
impacts on production. If users are
decrease. This could have an impact on user adoption.
entering fictitious data in production,
this could influence downstream
business processes, system logic, and Poor data quality in your training environment might impact the
reporting. training’s real-world relevance. One problem we see frequently in
To ensure more realistic scenarios poorly conducted training sessions is that users complain the data is
during training, we recommend your
not realistic. This in turn can have an impact on the quality of the data
dataset be a subset of real data from a
legacy system. that these users subsequently enter during training, which can affect
the quality of the data that these users enter once the system goes live.

How to assess competency and effectiveness of training?

One of the most effective ways to improve training is to collect feed-
back and use it to make enhancements. Training is not meant to be a
static exercise in which material is created once and delivered the same

575
 way in every iteration; it should be a work in progress that is updated
frequently. Once a training session is completed, there is still work to
be done. Many organizations require multiple training sessions, and all
of them present opportunities to revisit, evaluate, and improve future
training sessions.

You can measure the effectiveness of your training through several dif-
ferent recommendations. The first is by assessing your users’ learning
at the end of, or shortly following, training, in particular, prioritizing
the processes that they will use most frequently in their jobs. It is
much more important that sales users understand how to create and
update an account record in Dynamics 365 without assistance if it is
a process they will need to do every day. Such prioritization should
be emphasized during the creation of any assessment method you
choose. Earlier in the chapter, we discussed how creating solid training
objectives would help you assess the effectiveness of your training. We
can now take those training objectives and use them to form the basis
for how we will assess our users, so long as the training objectives we
created contain the most critical topics that we are training users on.

Many larger organizations create exams to assess knowledge of their

users or they might even create a certification program following
training to assess readiness at a user level, team level, or even larger
level, such as a business unit, subsidiary company, or geographical
region. Trainers can create metrics based on exam results and certifica-
tion statistics to evaluate training quality. If certain groups have scores
or statistics that are lower than desired, the organization can work to
Accuracy is improve future training or host follow-up trainings for these groups.
important to ensure
data quality, but it Day-in-the-life testing, which can be used to measure real-world sce-
narios and the effectiveness or speed with which your users complete
is also important tasks, also measures training effectiveness and organizational readiness.
to know if certain Accuracy is important to ensure data quality, but it is also important
business processes to know if certain business processes take your users 10 minutes or 2

take your users 10 minutes to complete. This could be used to measure your organization’s
readiness for the new application. If certain business metrics must be
minutes or 2 minutes met regardless of the application used, make sure these metrics are met
to complete. within the Dynamics 365 application. As an example, if your call center

Implementation Guide: Success by Design: Guide 19, Training strategy 576

 staff is expected to process at least six calls an hour, it is important to
ensure that your staff has been trained to complete a single end-to-end
Question call process in your application within 10 minutes.

What if, during a training session, a

critical bug is discovered that blocks
Another way to measure the effectiveness of your organization’s
training from going any further? training is by analyzing the results of your UAT. As we briefly discussed
in the “Training schedule” section, you can use UAT as a barometer to
Recommendation gauge user competency of your application by looking at the number
We believe your training environment
and types of non-system-related bugs that are logged by users, and
should be treated similarly to a production by measuring (usually informally, since the goal of UAT is accuracy, not
environment during a training session.
Bug fixes should not be made directly in speed) the amount of time users take to complete the processes they
production; likewise, bug fixes should almost
are testing. If you are finding that your testers are logging many items
never be made directly in a training instance
while training is taking place. Depending on that are not system bugs but rather due to knowledge gaps on the
the severity of the bug, a work item should
be created for a developer in their respective
function of your application, you might conclude that training in those
development or hotfix environment, and the areas needs updating. It could also mean that additional user training
bug should be fixed and tested before being
pushed to downstream instances (including in those areas is required.
training), as per standard ALM processes.

If the bug is critical enough to block entire Similarly, after going live, you can use helpdesk statistics to determine
business scenarios and prevent trainees
the longer-term effectiveness of training. If users are submitting help
from accomplishing any tasks, it might
be necessary to fix these bugs directly in desk tickets, or requesting assistance, on specific areas of your applica-
the training environment. However, we
recommend this only in situations where
tion, it is a good idea to review training or training documents related
the bug prevents any workarounds or other to those features to see if they should be updated. Again, if necessary,
sessions that could be taught. Fixing the
bug directly in a “live” environment could additional user training might be needed.
cause additional, unforeseen issues to occur,
which would further lower users’ confidence
in the application. Furthermore, any bug Another common approach to collecting feedback on training is to
fixes that happen in this instance would also
have to be fixed in the lowest instance where
conduct pilot training sessions with a limited audience. You can invite
development is occurring, and regression super users to attend these pilot training courses and gather feedback to
testing would have to be completed and any
potential conflicts would have to be resolved. improve. A side benefit of conducting these pilot trainings is that it helps
super users in their process of becoming experts in your application.

Assumptions,
dependencies, and risks
As discussed in the “Training schedule” section of this chapter, suc-
cessful training relies on many external factors that should be defined
in your training plan. These factors can be assumptions, which are

Implementation Guide: Success by Design: Guide 19, Training strategy 577

 statements assumed to be true of the project that are often critical to
training success. Here are a couple examples of assumptions:
▪ Appropriate trainer resources will be allocated to create training
material and train users.
▪ The tenant management and development team will create and
maintain a training environment for use during the project.

Your training plan and project plan should also include dependencies,
which can be represented in a process workflow. They effectively show
that certain tasks or activities must occur before others as part of your
training strategy. Figure 19-8 depicts a sample process workflow.
We recommend making sure each training
participant has a user account with the
same level of access as their forthcoming Fig.
19-8
Create training documentation on
production user account.
lead/opportunity management
We do not recommend giving system
administrator access to training users. This
is a common mistake we see organizations
make during training. They offer a variety Update Finalize
Review 1 document document
of reasons for doing this, including the Create draft Review 2
of document based on based on
of document of BPO
desire to not run into access issues during by SME feedback feedback
training and the security model having from SME from BPO
not been finalized.

If you are conducting multiple training

sessions in a single training instance, it is
important to take these proper actions at
This tree is an example of a straightforward dependency list for creat-
the conclusion of the training session:
• Any bugs that discovered during ing a specific type of training document. Your plan will contain more
training are properly documented
(with screenshots, if necessary) and
complex dependencies that might rely on multiple, unrelated project
sent for review or triage. tasks. Defining and organizing these dependencies at the start of a
• If restoring from a Dynamics 365
backup, ensure that a backup for the project will help reduce the amount of time your trainers or other
correct date is available. (Work with
project members are waiting for required tasks to be completed, by
your partner or IT staff, as well as
Microsoft, if necessary, to ensure this scheduling upstream project tasks specifically to minimize this down-
is not a blocker.)
• If restoring from source control as
time. This must be done in coordination with the broader project team
part of your ALM process, ensure and project schedule.
all code and data are ready for an
environment refresh.
• Refresh your environment using your Lastly, creating and maintaining an active training-risk log, as part of
chosen method.
• Validate user access and data via a the overall project risk log, is important to ensure proper attention is
smoke test, prior to the start of the
paid to training risks that could impact the success of training or other
next training.
project activities reliant on successful user training.

Implementation Guide: Success by Design: Guide 19, Training strategy 578

 Training as an
ongoing process
As mentioned in the introduction of this chapter, your organization’s
training journey begins long before the execution of user training, and
its ending extends long after those users complete their first tasks in
your production application. Training is an ongoing and constantly
evolving process that is reflective of changes in your organization and
application, as well as changes to the Microsoft Dynamics 365 ecosys-
tem. While developing a training strategy, it is important to consider
the implications of both types of change.

Internal change
Internal change, for the purposes of this section, refers to change that, at
a transactional level, is unique to the organization experiencing it. Each
of these types of change could require training material to be updated
and subsequently conducted on the same or different groups of users.

Personnel change

It’s important that User turnover is inevitable, either when employees leave an orga-
nization and new ones take their place, or when employees change
new users are job roles within the company. All new users are expected to quickly
provided support learn their job roles and how to navigate and complete their work in
during and after the application; to do this, these users need to be trained. Note that

onboarding, as they it is unlikely that an organization would conduct multiple lengthy

in-person training sessions for a single user’s onboarding and would
might receive fewer instead rely more on self-paced or web-based training as a means of
training resources education. It is important that these new users are provided support
than users who were during and after onboarding, as they not only are initially behind more

present during the experienced users who have been using the application for a longer
period, but they might receive fewer training resources than users who
application’s initial were present during the application’s initial rollout.
rollout.
Your organization’s training material for users who joined after go live
will be similar to that used by your end users before go live, but again,

Implementation Guide: Success by Design: Guide 19, Training strategy 579

 be sure that these users are supported and given the assistance they
require to train on the application.

Application change
Many Dynamics 365 projects have adopted a phased rollout approach,
in which features of each application are released to production over
time, as opposed to a “big bang” approach, in which a single deploy-
ment to production prior to go live has all included functionality. While
the rest of this chapter covers the development and distribution of
training regarding the first release of application functionality, future
phases of the project must include updates to training material and
new training sessions.
Phase two training might include
users who are already familiar with the
application, having used it after the phase Your organization might want to treat each phase of a rollout as a
one go-live, as well as employees who
distinct project from a training perspective, wherein each future phase
have never used or been trained on your
system. Any new and updated training contains a distinct, more condensed training plan and training delivery
material should reflect the difference in
experience and skill level of these user
schedule to ensure proper user training is conducted at every stage.
groups. It should not be too simple as to These future phases will focus heavily on making updates to training
be irrelevant to the experienced user, nor
too complex as to confuse the new user. material based on new business processes, and training users, both
new and experienced, on the application.

Consider a building-block approach when designing your training ma-

terial. Phase one training material and training can be the foundation
on which all users can learn to use the application. Future phases can
build on that foundation; training material for subsequent phases can
be updates of earlier material as well as newly created material.

External change
In other chapters, we discuss Microsoft Dynamics 365 as a software as
a service (SaaS) that provides continual updates to organizations, both
from a technical standpoint and a business standpoint. It is important
that your organization prepares for these updates as part of its training
strategy and that they are acknowledged in the training plan and ex-
ecution. Note that we allow customers to activate updates to software
at their own pace.

As updates to the Dynamics 365 application are announced and begin

to be rolled out, your organization should set aside an instance used

Implementation Guide: Success by Design: Guide 19, Training strategy 580

 Fig. for regression testing business processes against
19-9

Performance your custom application. As part of these activities,

trainers should be involved and assess the potential
improvements impact these updates have on business process.
Software update
We have made improvements on the schedule board Certain updates, especially improvements made
that should result in a 40% increase in performance. on the software’s back end, will result in little to no
Traning needs change in the way your users do their work in your
Even though this update is relevant and beneficial to application. Thus, they should not require a signif-
your users, it would not require an update to training icant update to training material, nor additional
materials or necessitate additional training, since there is
training. Other updates that are more focused on
no underlying change in how your users would conduct
their work.
user experience or new features within the platform
might drastically affect certain end users and will
necessitate updates to training material as well as
additional user training. Figure 19-9 contains ex-
amples of how software updates might or might not
Interface require changes to your training offerings.
updates
Software update
We have made updates to the layout of the schedule Training best
practices
board that should make navigation and creation of
bookings easier for your dispatchers.

Traning needs
Keep an eye on release notes to leverage new features. The authors of this book have experience with
Work with your implementation partner and your Microsoft Dynamics 365 implementations of all sizes
internal teams on your heatmap for Dynamics 365 and complexities. We have created and reviewed
capabilities. Have a regular review of future projects, training plans, participated in and hosted training
incorporating new features delivered.
sessions, and evaluated the outcomes of training
Join the D365UG User Group for Dynamics 365 to learn on end-user adoption. From most of these projects,
about member- driven education, networking, and events. common themes have emerged. We share the most
Create a Yammer group or Microsoft Teams channel to important recommendations here as best practices,
continue conversations on best practices and learnings. things to keep in mind during your organization’s
Host engagement events like town halls” or “Lunch and training process. These best practices are by no
Learns” to continue training after go live and drive end- means comprehensive, nor are they set in stone, but
user engagement. they are meant to guide a project team to a more
Share success stories about how people are using your successful training outcome.
application in innovative and impactful ways.

Implementation Guide: Success by Design: Guide 19, Training strategy 581

 Earlier is better
Starting early (with a training plan) is important. Dynamics 365 projects
are often run on tight timelines with little to no leeway. It is crucial that
a training plan, complete with staffing and resource availability, be
completed at the start of a project, to ensure no unexpected resource
requirements or shortages appear midway through implementation.
Give yourself sufficient time to prepare a training plan, develop con-
tent, and deliver training. Content development should be completed
well before the start of training delivery. Additionally, project roles and
responsibilities in cloud solution deployments are rarely separate; often
a single IT resource will wear multiple hats, and one of these hats might
be to create a training plan, review training material, or maybe even
conduct training. Making sure these individuals are aware at the outset
of their contributions to the training plan will help them balance their
workloads and complete certain activities earlier, which could prevent
delays from occurring.

More is (usually) better

In Dynamics 365 applications, a user can often access the same record
or complete the same business process in several ways. Thus, training
that initially seems repetitive is useful in that it teaches multiple ways to
accomplish a task depending on the situation and context. While too
much training can mean a small amount of wasted effort, for an orga-
nization adopting a new application, the cost of time is far preferable
to having users who are insufficiently trained on a sizable portion of
their job. The latter situation occurs in organizations far more frequently
than the former situation does, even for similar amounts of training.

It is also important that your training material does not contain too
much duplicate content. While it is certainly critical that both the web-
based and live content cover the same features (especially if users are
not consuming multiple types of content), too much overlap can result
in training fatigue, and overall apathy regarding training. An organi-
zation would certainly want to avoid a situation in which users skipped
valuable content because they considered it repetitive.

Implementation Guide: Success by Design: Guide 19, Training strategy 582

 Train with a goal in mind
Conducting a variety of training for all users using different methods
is important. But what is much more important is not losing sight of
the end goal: successful user education and meaningful adoption of
your Dynamics 365 application. It does the organization no good if
every technician understands how to use their new mobile application
but none of them uses it correctly because of a failed training process.
Having comprehensive training material is but one step in this process.
As stated earlier in this chapter, conducting effective training that leads
to excitement and high user adoption with accomplishment of critical
KPIs, in addition to the more obvious goals of high training partici-
pation and education, is the key that can separate adequate training
programs from excellent ones.

Consider your audience

In addition to having distinct job roles, your users come from diverse
backgrounds and probably different generations, too. Avoid alienating
groups of people with references, case scenarios, or even pictures or
stories that could confuse or even offend certain groups of people. Your
training should keep users engaged, but be careful not to be divisive.

For example, say you have a junior salesperson in their 20s and a senior
salesperson in their 60s attending the same introductory training. Your
application includes an embedded LinkedIn widget on your contact
entity. Trainers should not assume that both users (or, more divisively,
It’s important to only one) are familiar with LinkedIn and its capabilities. A good rule of

not lose sight of the thumb is to assume all training participants arrive with the same level
of background information and context of the application, plus any
end goal: successful technology knowledge required to use it effectively.
user education
and meaningful Accessibility concerns
adoption of your Your organization must consider accessibility while creating training
Dynamics 365 material. All Office 365 products have an accessibility checker that can
application. help you spot potential issues in your content.

583
 Identify ongoing challenges
As discussed earlier, feedback is an important part of any improvement
cycle. In addition to identifying areas of improvement for future train-
ings and training material, we recommend that the people conducting
training or collecting feedback from the training identify key areas
where users are struggling with new functionality. This information can
be used to create more focused training material that can be distribut-
ed to users, or to create webinars or other sessions specifically focused
on these problem areas.

Product-specific
guidance
Up to this point in the chapter, our training guidance has applied to
Finance and Operations apps as well as Customer Engagement apps.
While both live in the Microsoft Dynamics 365 ecosystem and custom-
ers frequently adopt both systems (often simultaneously), there are
differences between the two, which can mean differences in how each
project should train its users. In this section, we highlight some of these
differences and how to apply them to your training material.

Finance and Operations apps

A number of resources provided in Dynamics 365 Finance, Supply Chain
Management, and Commerce can assist with product help and training.

Help on docs.microsoft.com
The Dynamics 365 documentation on Microsoft’s docs.microsoft.com
site is the primary source for product documentation for the previously
listed apps. This site offers the following features:
▪ Access to the most up-to-date content  The site gives
Microsoft a faster and more flexible way to create, deliver, and
update product documentation. Therefore, you have easy access
to the latest technical information.
▪ Content written by experts  Content on the site is open to

Implementation Guide: Success by Design: Guide 19, Training strategy 584

 contributions by community members inside and outside of
Microsoft.
▪ Content is personalized based on language selected  If, for
example, a user is set up for the German language, then any Help
content they access will be provided in German.
▪ Help is now on GitHub  Customers can copy and create person-
You can find content on docs.microsoft.com
by using any search engine. For the best alized Help content and link it to their application. Customers can
results, use a site search, such as site:docs.
create a personalized and contextualized Help resource for users
microsoft.com Dynamics 365 “search term”.
of their custom business processes.

In-product Help
In the Dynamics 356 client, new users can enter the Help system to
read articles that are pulled from the Docs site’s Dynamics 365 area
and task guides from the business process modeler (BPM) in Lifecycle
Services (LCS). The help is contextualized to the form that the user is in.
For example, if a user is in a sales orders screen and wants to know how
to create sales orders, the Help system will show the Docs articles and
Fig.
task guides related to sales orders (see Figure 19-10).
19-10

Implementation Guide: Success by Design: Guide 19, Training strategy 585

 Task guides
Other useful help and training features are called task recorder and
task guide. Task recorder allows you to record a user’s activity on the
Consider the number of concurrent users
using the training environment as you
UI. You can capture all the actions as well as any UI fields and controls
choose the right tier for the environment. that were used. This recording can then be used in a task guide.
Do not automatically select the default
tier 2 environment if the volume of
expected users exceeds the recommended
A task guide is a controlled, guided, interactive experience that leads you
volume for the tier. For more information
about different environment tiers and through the steps of a task or business process. You can open (or play) a
the volume of users each supports, read
this documentation on selecting the right
task guide from the Help pane. When you select a task guide, the Help
environment for your organization. pane shows the step-by-step instructions for the task (see Figure 19-11).
Localized task guides are available, and you can create custom task
guides. The data you enter in the system is saved in the environment.
Fig.
19-11

To begin the guided, interactive experience, select Start task guide at

the bottom of the Help pane. A black pointer shows you where to go

Implementation Guide: Success by Design: Guide 19, Training strategy 586

 first. Follow the instructions that appear in the UI and enter data as
Fig.
directed (Figure 19-12).
19-12

Dynamics 365 Customer Engagement

Dynamics 365 Customer Engagement has product-specific tools that
can aid in training users. Custom help panes are one of the most com-
mon tools used; we describe them in more detail here.

Custom help panes

Dynamics 365 Customer Engagement lets you create guided help pag-
es that provide your users with in-product assistance. This assistance
can be customized to your application and user base and can include
text, links, images, and video links. For more information about custom
help pages, read “Custom help panes and learning path.”

References
Training plan template

Training plan charter template

TechTalk Series: Training Plans and Content for your Finance & Operations
Project (Microsoft Dynamics blog)

Dynamics 365 training offerings (Microsoft Docs)

In-product Help – Finance & Operations (Microsoft Docs)

Implementation Guide: Success by Design: Guide 19, Training strategy 587

 Checklist

Implementation Ensure trainers and super users are trained well in

advance of end-user training, which should ideally occur
Have training objectives that are SMART (specific,
prior to UAT.
measurable, achievable, realistic, and timely), relevant
to your organization’s goal of user adoption, and fit the Ensure the training plan appropriately addresses
organization’s overall change management goals. accessibility.

Have a process to update the training plan incrementally Define a process for feedback and continuous improve-
as necessary to reflect scope, changes, risks, and depen- ments to training materials.
dencies to ensure adoption and engagement.
Identify a process to provide for continuous training in
Consider what to include regarding system process and alignment with Microsoft updates and changes to the
business process, so that the training provides the best solution as well as changes in roles and responsibilities.
possible foundation to end users.

Clearly define the match between personas and roles

as well as between business processes and user level for
each training session, so users receive relevant training.

Ensure the training schedule accounts for key project

milestones, resource requirements and availability,
dates, and tasks, and is in alignment with the overall
project plan.

Consider several types of content, the time to create

it, the effort to deliver and maintain it, storage, and
the skillset of the authors and trainers when designing
training material.

Set up the training environment with realistic data, user

profiles, and scenarios. Refresh and clean it after training
sessions to maintain quality.

Implementation Guide: Success by Design: Guide 19, Training strategy 588

 Case study
A proven path to
training success
This global fire-protection company has earned a reputation for
providing customers with maintenance and service of the highest
quality. In this mid-size company’s 75-year existence, they built their
business organically through natural growth, marketing, and geo-
graphic expansion, as well as through numerous acquisitions of smaller
fire-protection companies. They recently hired a new chief information
officer to replace the CIO who retired after 15 years in the job. The new
CIO had extensive experience working with Microsoft and SaaS prod-
ucts. The CIO’s immediate focus was to consolidate and modernize the
applications that employees were using to conduct their business.

In the training plan, the company included a mix of high-level and de-
tailed training objectives. The high-level objectives included these goals:

▪ To not allow access to the live system without a solid training program
▪ To prepare a core team of trainers to help support the initiative
▪ To continue receiving feedback and improving the training approach
▪ To develop the training materials early and schedule classes early
▪ To prepare all application users to efficiently use the application
(Dynamics 365 Finance and Dynamics 365 Supply Chain
Management or Dynamics 365 Customer Engagement), as well
as address any key business process changes required in their
job function

Implementation Guide: Success by Design: Guide 19, Training strategy 589

 Specific objectives were centered on business processes like Prospect to
cash, for example, “All sales team members should be able to execute
the standard Prospect to cash flow on their own.” Their specific training
objectives, listed here, helped support business goals for the application:
▪ Increase sales by 10 percent year over year
▪ Improve productivity by 25 percent by measuring Work Order
completion time
▪ Increase user satisfaction (covered in annual poll) by 10 percent
and user competency (measured by metrics within the application)
by 15 percent within the accounting business department

The team understood that for training to be successful and for mean-
ingful user adoption to be achieved, they needed to begin planning
early and set up a strong team to support it.

The team benefited from using a change-impact assessment matrix to

understand the changed business areas, topics, and user groups impacted.
The matrix helped them prepare and prioritize training sessions.

The company had experience developing and conducting trainings for

previous projects. They knew they wanted to incorporate a variety of
learning modalities; written documentation, videos, hands-on learning,
and multiple training labs using real and recognizable data.

Their legacy application infrastructure was a reflection of the way

business had grown. The back-office system ran on an on-premises
application that had been in use since 1990. Front-office users (sales-
people and field dispatchers), as well as field technicians, did not
employ any application; they were still using pen-and-paper-based
methods. Due to the number of acquisitions they had made over the
years, these methods varied, sometimes by business unit, sometimes
by other organizational structure, and there was no overarching gover-
nance model consolidating these different methods.

As a result, employees that were in identical roles used different

methods to accomplish their day-to-day work. This discrepancy in job
execution versus job role could be attributed to the fact that many
of them joined the organization via acquisition and had never been

Implementation Guide: Success by Design: Guide 19, Training strategy 590

 required to change. It was necessary to capture all “as is” processes and
assess the impacts of moving to the new application, from a technical
perspective and from a business process change perspective.

Given the numerous challenges of supporting a wide variety of appli-

cations for all users, the company made use of the BPM, task guide,
and custom guided help features for content development and user
training. Having all business processes documented on one platform
gave the organization a single source of truth. Task guides not only
provided consistency to all the business processes in use across the
different departments, but they also made the employee orientation
and training a much smoother experience. Task guides were also
embedded in the product help experience to provide a 24/7 interactive
help experience.

By clearly defining all these processes, as well as the group of users to be

trained, they were set up for success. Its subsequent training delivery was
a smooth process that included all user persona business processes.

Because of the evolving nature of Dynamics 365 applications and or-

ganizations, as well as the fact that the project was being rolled out in
multiple phases, the company developed an ongoing training process
by which training was executed, reviewed, and updated as a cycle.

Formal feedback was recorded after the trainings and Microsoft Teams
channels were created for employees to continue providing feedback
to the team. Users were encouraged to share knowledge, ask ques-
tions, and suggest improvements to the training materials. The team
was also able to collect feedback and create metrics using help desk
tickets, which helped them identify areas of the application that users
found particularly challenging.

The organization determined that mobile Field Service technicians

were logging more tickets than other application users. By learning the
types of tickets being created by their technicians, trainers were able to
pinpoint a specific area of the Work Order process that needed to be
made clearer during training. Adjustments were made to the training
material, too.

Implementation Guide: Success by Design: Guide 19, Training strategy 591

 The objective “To not allow access to the live system without a solid
training program” was a difficult objective to meet in the initial days,
but the company learned over time that making sure every user
received adequate training drove a significant reduction in business
process issues for the company.

In the first few months, an evaluation of key KPIs showed that the orga-
nization was on track to meet all the detailed objectives set by the team.

Implementation Guide: Success by Design: Guide 19, Training strategy 592

Operate
20 Service the solution
21 Transition to support

593
20
Implementation Guide: Success by Design: Guide 20, Service the solution
Guide
Service the
solution
594
Continuing the business applications journey.

Introduction
Your solution has been deployed to production.
Users have been trained and onboarded. A support
process is in place.
Most implementation projects focus on building and deploying the
solution, but perhaps not nearly as much emphasis is put on educating
the owner of the new system in keeping it optimal and healthy.

But you might be thinking, “Dynamics 365 is a SaaS platform. Isn’t

Microsoft responsible for maintaining these environments?”

For successful project Let’s take a scenario where you rent an apartment in a large building
governance, consider these
complex. The owner of the building is responsible for cleanliness and
main areas:
upkeep of the building and common areas. They make sure to have
Monitor service health
proper plumbing and electricity in each apartment. They may add or
Service updates
improve amenities such as an exercise gym or a garden in the yard.
Environment maintenance But it is each renter’s responsibility to keep their apartment in good

Continue the business

condition. The renter should clean the apartment and throw out
application journey their garbage. They will need to replace the lightbulbs in the fixtures
if it goes out. If there are any issues with the plumbing, they need

Implementation Guide: Success by Design: Guide 20, Service the solution 595
 to contact the building manager so that a repair person can be dis-
patched to help. In a similar way, Microsoft, the owner of the software
as a service (Saas) system, will be responsible for the platform, the
building and property in our scenario. The customer is responsible for
the upkeep of the health of their solution, the apartment.

In this chapter, we first look at how to proactively monitor the health

of the solution. Then we cover the continuous updates provided by
Dynamics 365 and best practices for working this pattern into your
project plan to minimize impact to the changes in your environment.
We also address environment maintenance tasks to maintain a healthy
system. Finally, we challenge you to continue your education in the
ever-evolving Microsoft business application journey.

Monitor
service health Monitor service health
Service updates A key principle for a successful onboarding experience to Dynamics
365 is knowing the health of your environments at all times. Your team
Environment maintenance must be able to troubleshoot issues right away.

Continue the business

application journey Dynamics 365 and the Power Platform have gone through rigorous
testing, so why do customers need to check that it’s running smoothly?
It’s because of each organization’s unique usage patterns and the
extensibility of the platform.

Just about every customer makes changes to their Dynamics 365

environment. It may be as small as adding new fields and forms. Or
your business may require heavy customizations with complex logic or
performance-impacting integrations with other systems. Or your orga-
nization may have more concurrent users than the typical use case, so
data management needs are also growing at a rapid pace.

Another reason to monitor your system is to understand usage pat-

terns for better business intelligence. Through this analysis, business
stakeholders can pinpoint where to invest and prioritize improvements
and focus areas for training.

Implementation Guide: Success by Design: Guide 20, Service the solution 596
Microsoft highly recommends monitoring service health and provides
you the tools to do so. That way, you know when irregularities are
found and action is needed.

As administrators and power users of Dynamics 365, you can focus on

several key areas when monitoring usage. These areas impact per-
formance, security, cost, and supportability, all central to keeping the
system performance at optimal levels.

This section addresses key areas for monitoring the solution:

▪ Performance monitoring
▪ Licensing and operational cost
▪ Storage allocation and usage
▪ API request and service protection
▪ Security and privacy
▪ User access and resource usage
▪ Application logs and errors
▪ Messages from Microsoft

Performance monitoring
Microsoft recommends that Dynamics 365 project teams consider add-
ing performance testing to the project’s test cycle. Performance testing
is an effective way to gauge the impact that your customizations may
have on the baseline Dynamics 365 solution.

Performance testing is typically conducted in a simulated environment.

This provides insight into changes in solution design or configuration
prior to production rollout at scale. By conducting performance testing
early, Dynamics 365 administrators have a baseline performance ex-
pectation of the new system that can be used as comparison over time.

Many factors can influence how performance is affected in live produc-

tion scenarios:
▪ Integrations with other production systems have fluctuating
response times that are different than in the simulated tests
▪ Workflows such as back-office jobs could be running, which stress-
es the systems

597
 ▪ The network traffic can vary throughout the day depending on an
organization’s usage patterns
▪ For remote workers, reliance on individual internet connections
Chapter 17, “A performing solution, beyond could cause different outcomes for each user
infrastructure,” covers the importance of
having a performance strategy that includes
elements such as defining performance Ultimately, the responsiveness experienced by end users is caused by
requirements, establishing baseline metrics,
and ongoing performance testing.
a mix of multiple factors that aren’t limited to the performance of the
Dynamics 365 application itself.

To understand the performance impact in these scenarios, proper

end-to-end logging and collection of telemetry is required. This pro-
You can use many different tools to
vides a complete picture of how long each step of the pipeline takes to
monitor performance. For Finance and complete. This information can reduce the time required to triage and
Operations apps, Lifecycle Services (LCS)
is the first place to go for performance fix any issues.
monitoring.

For Customer Engagement apps, As this data is collected, monitor the performance of the solution and
Dataverse analytics is helpful. With it you
can gauge and monitor performance from
set up notifications to warn you when performance of any aspect of
within the Power Platform Admin Tool. the solution varies from a defined range.
You can also use Azure Application Insights
to monitor applications like Dynamics 365 Poor performing solution leads to low user adoption. But you can stay
for custom telemetry needs.
ahead of this through proper monitoring and alerts.

Licensing and operational cost

As your solution matures and new users are onboarded to your
Stay compliant with user licensing
requirements and view the independent Dynamics 365 solution, keeping track of the rate of licenses being
software vendor (ISV) license status for
consumed will be very important. Insight on the number of available li-
Finance and Operations apps for more
information. censes will help you plan ahead and adjust your purchasing as needed.
Review the business subscriptions and
billing documentation for Microsoft 365 Another aspect is the operational cost. With solutions that include
and the Microsoft 365 admin center to
manage your subscription for Customer cloud computing, usage volume heavily impacts this expense. Staying
Engagement apps the and Power
aware of any changes in your organization’s usage will help business
Platform.
and IT managers make key budget decisions.
The Dynamics 365 Licensing Guide
provides details on licensing
requirements.
Storage allocation and usage
With a Dynamics 365 subscription, you start with a set minimum
amount of storage space for your environment.

Implementation Guide: Success by Design: Guide 20, Service the solution 598
 Depending on your needs, you can add more data to suit your business
size and expected growth.

Administrators must be mindful of the amount of storage that is

available for the organization’s subscription as usage grows. Different
data classifications (such as transactions, files, and logs) have different
allocations depending on your subscription.

Users and integrations aren’t the only cause of storage growth. Logs
from system jobs, indexes created to help performance, and additional
application data added from new modules also contribute to storage
growth. Another scenario that impacts storage allocation is in the copy
and restore operations of an environment. Depending on the type of
copy, the size of the database can be very different. As an administra-
tor, you need to be mindful of who can create new instances and what
Refer to Chapter 10, “Data their true needs are to minimize impact on the storage allocation as
management,” for details on storage
entitlements, segmentation, and impact
these copies are being restored.
to allocations when backing up and
restoring instances.
Administrators should monitor the volume of storage that is currently
For information on storage capacity for used as well as its growth rate. This information will help you budget
Finance and Operations apps, see the
Dynamics 365 Licensing Guide. for any additional storage needs or look into data archiving and de-
letion to free up space. Scheduling and cleaning up data from time
Note that Dataverse storage capacity
entitlements and usage changed in 2019. to time will help as well. This is covered in the “Environment mainte-
nance” section of this chapter.

API request and service protection

One of the advantages of choosing a cloud solution over on-premises
software is that your application is scalable. Among other things, you
can adjust the amount of processing and memory resources based on
the usage. But to provide consistent availability and performance for
everyone, Microsoft monitors and applies some limits to how APIs are
used. These limits are designed to detect when client applications are
making extraordinary demands on server resources.

The intent is to prevent over-utilization of resources to preserve the

system’s responsiveness and performance for environments running
Dynamics 365. When an API limit is hit, the service may be throttled or

Implementation Guide: Success by Design: Guide 20, Service the solution 599
 even prevent API calls to be run. Err or logging shows you when these
limits are exceeded.

For Finance and Operations apps, use

LCS to monitor API usage. For more
Administrators can also pull historical telemetry reports to see if any areas
information, review priority-based of the application are at risk of hitting these limits. Then you can work with
throttling and how to configure
priorities for special circumstances such the implementation team to make appropriate design changes. Or
as integrations. better yet, tools like Azure Application Insights allow you to set thresh-
For Customer Engagement apps, use olds on these API calls so that when they’re exceeded, the administrator is
the Dataverse analytics available in the
notified and can mitigate the risk of throttling or being shut down.
Power Platform Admin center. Refer to
the service protection API limits and the
ways to maximize throughput to avoid
hitting these limits.
Security and privacy
From a security perspective, you want to prevent unwanted access to
your system, applications, and data. The unified Microsoft 365 security
center combines protection, detection, investigation, and response to
threats for your Dynamics 365 system, all in a central portal.

At the solution level, you need strategic planning to create a robust

security policy that can be implemented during deployment. As the
administrator who services this solution, you must also monitor and act
upon other aspects of security.

Take a scenario in which an employee changes roles or leaves the

company. You need a process to change the security role and permis-
sions for that user. When a vendor is granted access to a system as a
consultant, they often continue to have access to it even when they’re
no longer working with the company. Microsoft recommends that you
have a process to address these scenarios.

“Who has access to my customer data?” may seem like a simple ques-
tion, but as the number of users and system usage grows, it can be
daunting to keep track. You need to be able to identify who is access-
ing the system and what they’re doing with the data.

Depending on the industry or the region of the organization, you

may also have regulatory or compliance-related requirements. You
might need to provide reports about customer data handling or delete
customer data upon request.

Implementation Guide: Success by Design: Guide 20, Service the solution 600
 Microsoft recommends that you have a proper auditing strategy to
capture the information needed to track user actions so that you can
satisfy these types of requests. Most common areas are covered by
Refer to Chapter 12, “Security,” for details
Dynamics 365. Because auditing takes up mor e storage and poten-
on security strategy and planning, and
see the security center overview and the tially impacts performance, administrators need to turn some of these
compliance center overview for more
details. The Microsoft 365 security center
capabilities on where they may not be by default.
provides the ability to search through
Dataverse activity logging for Customer
Engagement apps.
User access and resource usage
Your organization may be subject to
rules such as the General Data Protection For any application or solution, it’s important to understand your or-
Regulation (GDPR) that give users specific
ganization’s resource usage patterns. Business sponsors want to know
rights to their personal data. You may
need to respond to data subject requests who is using (and not using) the system and the frequency of business
(DSRs) to delete a user’s personal data.
processes and use cases that are being run.

You can clearly understand user adoption through telemetry on usage

patterns, and you can take proven actions to improve it. For example,
if a certain capability isn’t being used much, you can query the users
Monitoring and Diagnostic tools in LCS
to get their feedback on it. By using telemetry to identify areas with
provide telemetry for usage patterns for
Finance and Operations apps. the greatest impact to the user group, the project team can work with

Microsoft Dataverse analytics provides

business sponsors to prioritize features.
user access and usage pattern telemetry
for Customer Engagement apps and the
Power Platform. You can also use this information when estimating license needs. For
example, some licenses may be assigned to users who rarely access
the system; you can reassign these to someone new. Insights provided
by the telemetry are readily available in Dynamics 365. Use this data
to help improve business processes and allocate resources in the right
areas to maximize your investment.

The platform is constantly adding functionality to the standard usage

reports. If you have any gaps between your needs and these options,
you can tailor tools such as Azure Application Insights to your usage
pattern tracking requirements.

Application logs and errors

Typically, system errors are found by end users who report issues
through an IT service management (ITSM) process. But you may have
unreported issues that can only be found in the application logs.

Implementation Guide: Success by Design: Guide 20, Service the solution 601
 Turning to notifications and application logs to proactively look for
entries is a good way to find trouble spots and address them before
they impact users.

You can use tools such as Azure Application Insights for Dynamics
365 and other applications and services that are part of the overall IT
solution for your organization. Application Insights lets you collect
telemetry both in and outside of Dynamics 365.
Monitoring and Diagnostic tools in LCS
allow administrators to monitor and
query logs for issues detected in the
For example, if a process is initiated by a user in Dynamics 365 that
system for Finance and Operations apps.
calls on an integrated, but external system, Application Insights can still
Trace logging in Dataverse provides
plugin error information for Customer
detect performance and exceptions at different stages of the execu-
Engagement apps and the Power Platform. tion pipeline. You can see these issues whether they occur at the user
You can also use Microsoft 365 service interface level, in Dynamics 365, or the external system. This empowers
health to identify service issues, and the administrators who monitor alerts to react quickly the moment the
administrators can be notified via email
or through the mobile app. exceptions surface. They also have immediate access to information on
the source of the issue.

Messages from Microsoft

As your SaaS provider, Microsoft provides ample information about the
health of your service, upcoming release information, service excep-
tions and interruptions, and other important details. The Microsoft 365
Message center is the central location for technical and licensing com-
munications with our customers. As an administrator of your solution,
this will be a very important area to frequently check for new messages.

Typically, organizations have multiple people who manage Microsoft

services. You may want to provide appropriate team members with
access to these messages.

You can also configure your preferences on receiving emails. A mobile

app is available to manage these communications.

As described in this section, having the information needed to under-

stand the health of your system is key in knowing when to act. Multiple
tools are available to monitor your solution, which can feel daunting.
But with tools to alert administrators when services are running less

Implementation Guide: Success by Design: Guide 20, Service the solution 602
 than optimal, the task will be much more manageable.

Key alerts you can receive from Microsoft include the service updates to
your Dynamics 365 solution. In the next section, we discuss when and
how Dynamics 365 is updated and what you can do as an administrator
to take advantage of the predictable nature of the service updates.

Monitor service health

Service updates
Service updates
To enable businesses everywhere to accelerate their digital trans-
Environment maintenance formation, Microsoft is continuously enhancing Dynamics 365 with
new capabilities. We add product enhancements and performance
Continue the business
improvements at a rapid pace, so it’s important to make sure Dynamics
application journey
365 provides an optimized user and administrator experience. Our
objective is to help you keep your computing environment current in a
consistent, predictable, and seamless manner.

Service updates are continuous, touchless updates that provide new

features and functionality. They eliminate the need to do expensive
upgrades every few years. The service updates maintain backward
compatibility, and contain application and platform changes that are
critical improvements to the service, including regulatory updates.

Microsoft’s push to One Version

One of the benefits of Dynamics 365 is that every customer runs on
the same version of the service. With market demand for increased
agility to gain incremental benefits of these cloud offerings at reduced
operating expense costs, Dynamics 365 took a bold step to refrain
from multiple releases. Now every security patch, bug fix, performance
enhancement, and functional improvement accrues to all implementa-
tions across the globe.

We call it One Version. This is also referred to as Run One, a solution

that is evergreen in the modern SaaS world.

Implementation Guide: Success by Design: Guide 20, Service the solution 603
 One Version promises to bring predictable release management to
these areas.

Predictable updates with continuous deployment

▪ Regularly scheduled intervals for major releases of performance
and reliability improvement updates throughout the year
▪ Using safe deployment practices and monitoring updates closely
for any issues
▪ Knowing the cadence up front so you can work it into your project
schedule

Early visibility of upcoming changes

▪ Access to release notes well before the release for readiness so you
know what the new changes are and the dates to start planning early
▪ Visibility into new features and updates for business sponsors that
can help their team be trained in new features

Refer to “Modernizing the way we update

Dynamics 365” by Mo Osborne, Microsoft Test new capabilities ahead of time
COO Business Applications, for details on
the benefits of continuous deployment. ▪ Ability for customers to opt in for early access to the release
▪ Fully supported, production-ready updates that can be tested in
your sandbox environments
Fig.
20-1 ▪ Turning on features when you’re ready instead of having it forced
upon your environment

Planning Scheduling
effectively to avoid conflict with Opportunities to advance your knowledge
for updates project plan ▪ Faster access to new features
▪ Ease in sharing ideas and collaborating with the greater community
Raising Handling because everyone is on the same version
awareness unsupported
of feature changes techniques
With One Version, Dynamics 365 addresses the tough challenges faced
by customers and partners and reduces common rollout concerns
Staying aware Encouraging (Figure 20-1). The solution automatically enhances the predictability
of what’s coming adoption
of product updates. For example, One Version reduces the effort and
capacity needed to test the update. It makes it easier to manage busi-
ness change and facilitate appropriate adoption.
Understanding Aligning
effort and capacity the team with the
for testing implementation plan
Although One Version greatly reduces the impact of deployments, the
solution owner and the implementation team are still responsible for

Implementation Guide: Success by Design: Guide 20, Service the solution 604
making sure certain tasks are addressed. These include planning for the
update, assigning ownership of the tasks, raising awareness of coming
changes, and supporting adoption. We cover these items throughout
this section.

Release readiness
System administrators and others who have signed up for service
notifications are alerted about upcoming releases, minor updates, and
bug fixes. Being properly prepared is the key to successfully managing
your solution. This isn’t just up to the Dynamics 365 administrator. You
need tight coordination between systems administrators, the Dynamics
365 project team that works on the platform, and the business user
groups and subject matter experts (SMEs) who are the end users and
champions of the system.

These notifications provide the information you need to start your

planning: dates of release availability, release notes, and the process to
opt in for early access.

Release notes and feature deprecation

Great planning starts with education; learning when and what is being
rolled out. Your notifications explain the upcoming changes with a link
to online documentation for more details. This information can help
you gauge the impact to your organization and determine when and
how to take action.

Administrators are notified of upcoming releases through the mes-

sage center and by email. Notes for major releases are made available
months before availability and provide several key pieces of infor-
mation. First, the notes list the features in the release, including a
description of the feature and the general availability (GA) date. The
notes also show if the feature will be available for public preview and
early access. This information will help in your readiness tasks.

Let’s break down the three types of release dates:

▪ Public preview  These capabilities are pre-release features
intended to collect early feedback and aren’t intended for

605
 production use. For planning purposes, these public previews can
give you a good idea of what’s coming up in the short-term road-
map. Not every feature is available through a public preview.
▪ Early access  You can apply these production-ready, fully sup-
ported updates and features to your environment prior to the GA
date. See more details in the following section about the process
of opting in early.
▪ General availability  This is the date that these features are
deployed to your environment if the administrator hasn’t opted in
for early access.

Each release wave includes features and functionalities that you can
enable for different types of users:
▪ Users, automatically  These features include changes to the user
experience for users and are enabled automatically.
▪ Administrators, makers, or analysts, automatically  These
features are meant to be used by administrators, makers, or busi-
ness analysts. They’re enabled automatically.
▪ Users by administrators, makers, or analysts  These features
must be enabled or configured by the administrators, makers, or
business analysts to be available for their users.

If you choose to opt in to early access updates, you get features that
are typically mandatory changes automatically enabled for users. Each
feature in the release notes indicates which category it falls under.

There may also be announcements about deprecated features; those

with a timeline to stop being supported. These announcements are
equally important because they allow you time to gauge the impact
and work with your business sponsor and technical team to address
any concerns.

Deprecation of features is an important part of a solution’s growth. As

business and technology needs evolve, the solution needs to change
to keep up with new requirements for security, compliance, and over-
all modernization. As a result, some features are deprecated and are
replaced by a better solution.

Implementation Guide: Success by Design: Guide 20, Service the solution 606
 Deprecated features continue to work and are fully supported until
they’re officially removed. After removal, the feature or capability no
longer works. The deprecation notes provide information on what
Refer to the Message center for
detailed information on notifications,
features are being removed, when this will happen, why it’s happening,
email preferences, and recipients for and what actions you can take to address the impact. Just like when
service updates.
you’re getting ready for new features, organizations must plan and
For Finance and Operations apps, prepare well before the removal of features to avoid negative impact
refer to the One Version service
updates overview for details on release from the deprecation.
planning, release notes, deprecations,
and release cadence.
Opt in for early access
For Customer Engagement apps and
the Power Platform, refer to the release
At this point, you have the release notification and reviewed the re-
notes, deprecation announcements, lease notes to do an impact assessment. The next step is to work with
and release cadence.
the project team to test the release with your solution. Some features
available through our public preview program may be of interest for
your organization.

Microsoft recommends creating a sandbox instance from a backup

of your test instances prior to update deployment. The environment
should be running your latest solution with enough data for mean-
ingful testing. If you already have automated tests, you can speed up
this process by validating that your solution functions as expected. If
you need to fix any areas, you can flag them and work them into your
project plan.

If your solution has been developed using supported methods for

customization, the release in most cases will have little to no effect on
it. Scenarios regarding deprecation that weren’t addressed previously
may surface from testing and need to be fixed.

Another important area to cover for early access is to work with the
business sponsors to help them understand the impact to the end us-
ers. As part of the release, some updates may impact user experience,
such as user interface (UI) navigation changes. Even small differences
can have a meaningful impact. Imagine users in a large call center
scenario, in which every additional second on a call with a customer
can impact their service goals. In such a case, business managers want
to make sure that the user group receives proper communication and
takes time to provide training if necessary.

607
 After this due diligence, your organization can schedule and apply the
new release to the production environment. You should time this task
for when it’s most convenient to the users and administrators. Your
organization can determine the best time to apply the release, when
there will be the least amount of disruption to end users, other tech-
nical dependencies, and impact to other projects. Don’t wait for the
release to be automatically applied to your environment.

Microsoft has deliberately planned the release cadence so that it allows

you enough time to prepare. Aligning your own internal deployment
planning with this schedule allows you to benefit from the new features
and bug fixes and make sure that any customizations specific to your
environment continue to function as expected. Thorough planning is
key so that there are no unexpected results from the updates and you
have control of when these releases are applied to your environment.

Product-specific guidance: Service

updates for Finance and Operations apps
With continuous touchless service updates in Finance and Operations
apps, the system maintains backward-compatibility, so you don’t need
to merge your code.

Safe deployment practices

Three new innovations provide additional enablement for the new
modern application lifecycle: safe deployment rings, automated re-
gression testing, and the customer Release Validation Program (RVP).
In this section, we explain these and how they apply to you.

Deployment rings are a technique used to decrease any risk associated

with rollouts for Azure and other Microsoft cloud services managed at
global scale.

As Dynamics 365 updates are created each month, they progress

through a series of rings, with each ring providing broader exposure
and usage and validation through system telemetry (Figure 20-2).
Refer to the Dynamics 365 for Finance
and Operations Cloud Application
Lifecycle for more information. The GA update benefits from extensive Microsoft testing as well as
validation through each of the earlier rings.

608
 Fig.
20-2
Safe deployment practice for
Finance and Operations

Finance and Targeted First Standard Servicing

Operations team release release release (GA) N and N-1

Finance and Targeted First

Operations team release release
▪ Extensive validation ▪ Preview early access program ▪ Select customers
▪ Compatibility checker ▪ Preview build ▪ Auto update
▪ Over 100 customer RVPs ▪ No production use ▪ Production ready

Standard release Servicing

▪ Predictability ▪ Hotfix support for critical issues
▪ Customer-controlled ▪ Available for N and N-1 updates
patching window ▪ Cumulative updates
▪ UAT followed by production
on consecutive weeks

Implementation Guide: Success by Design: Guide 20, Service the solution 609
 Fig.
Update cadence
20-3
Customers are required to take a minimum of two service updates per
year, with a maximum of eight service updates per year (Figure 20-3).
8 updates
delivered You can choose to pause up to three consecutive updates at a time to
per year accommodate your project schedule.

Jan Pausing a service update can apply to the designated user accep-
Feb tance testing (UAT) sandbox, the production environment, or both.
Mar If the pause window ends and the customer hasn’t self-updated to a
supported service update, Microsoft automatically applies the latest
Apr
update based on the configuration selection available in LCS.
May
June System updates follow these guidelines:
July ▪ Updates are backward-compatible
▪ Updates are cumulative
Aug
▪ Customers can configure the update window
Sept
▪ Quality updates containing hotfixes are only released for the
Oct current version (N) or previous version (N-1)
Nov ▪ System updates contain new features that you can selectively
choose to enable
Dec

Release readiness
We strongly recommend that you plan ahead and work the updates into
your internal project release cadence. To do so, take the following steps:
▪ Step 1: Plan  Have a good understanding of the release schedule
and a plan to work this into your application lifecycle manage-
ment (ALM) strategy. Because you can pause updates up to three
months, you can set aside plenty of time for testing, impact anal-
ysis, and developing a deployment plan. Use the impact analysis
report from LCS to identify areas of change that may affect your
solution and help determine the level of effort needed to remedi-
ate any impact.
▪ Step 2: Test  We recommend using a non-production envi-
ronment such as your UAT instance to opt in early and apply the
release. You can configure service updates through LCS and spec-
ify how and when you receive service updates from Microsoft to
your environments. As part of the configuration, define the update

Implementation Guide: Success by Design: Guide 20, Service the solution 610
 environment (production) and an alternate sandbox (UAT). Use the
Regression Suite Automation Tool (RSAT) to perform regression
testing to identify any issues. Work any fixes into your ALM cycle
and deployment plans.
▪ Step 3: Deploy  After you define environment and schedule for
the service updates through LCS, the back-end tools automatically
update the system.

Product-specific guidance: Service

updates for Customer Engagement apps
Dynamics 365 is continually updated to provide valuable functionality,
fix any issues, and maintain the overall health of your solution. System
administrators receive communications and are alerted to when
changes will be made to their environments. Organizations need to
plan to enable these service updates in alignment with their internal
ALM process.

Safe deployment practices

When rolling out the major and minor releases, Microsoft Dynamics
365 follows a series of progressive, staged deployments across the
world to adhere to safe deployment practices. At each stage, feedback
is gathered and quality is monitored for continuous improvement.

For example, for Customer Engagement apps and Power Platform

releases and updates, a station-based deployment plan is followed
(Figure 20-4).

After thorough internal integration tests and validations, Customer

Engagement apps and Power Platform updates are rolled out to
Station 1. This first release station consists of select customers and
production-like environments. This is set up for early validation testing
of the service update before it’s rolled out to customer production
environments. Based on the pass rate of the first release, the service
update gets rolled out to the customer environments by Stations 2–6.

Some organizations require instances to be in different regions, which

may place them into different stations than that of their primary

Implementation Guide: Success by Design: Guide 20, Service the solution 611
 Fig.
20-4 Safe deployment practice for
Customer Engagement
Customer First Standard release
Engagement Team release and servicing

Station 1
Station 2
Station 3
Station 4
Station 5
Station 6

Customer
Engagement Team Station 1: First release Station 2 JPN, SAM, CAN IND
▪ Extensive integration ▪ Production quality
testing and validation ▪ Early view of weekly release
Station 3 APJ, GBR, OCE
▪ Solution checker ▪ Select customers

Station 4 EUR

Standard release and servicing

Station 5 NAM
▪ Predictable weekly releases
▪ Hotfix support for critical issues
Station 6 SPECIAL

Implementation Guide: Success by Design: Guide 20, Service the solution 612
 geographies. Therefore, as the deployment cycle of a new release com-
mences, instances in different stations can be on different versions of
the solution. As your project team is developing new features that are
Check the latest release of station
consumed into your ALM deployment process, set up a version check
mapping and their corresponding regions.
Release updates to Station 1 through for Dynamics 365 and make sure that there is a match so that your
Station 6 follow the dark hours defined for
each geography.
project doesn’t encounter incompatibility issues. When the version
of the source environments matches the destination, you can safely
deploy your solutions.

Types of service updates

There are four types of updates:
▪ Major release  These happen twice per year in April and October.
Major releases offer new capabilities and functionalities. These
updates are backward-compatible. New features with changes that
might be disruptive to the user experience are turned off by default.
▪ Minor service updates  Minor service updates are deployed on
a weekly basis, region-by-region. They contain software customi-
zation changes that support new features, product improvements,
and bug fixes.
▪ Planned maintenance  Planned maintenance includes updates
and changes to the ser vice to provide increased stability, reliability,
and performance.
▪ Unplanned maintenance  Applications sometimes encounter
unexpected issues that require changes to maintain availability.
Microsoft strives to provide as much notification as possible
around these events.

Update cadence
Customers receive two major updates per year, in the April and
October GA releases (Figure 20-5). You can get early access and opt
Dynamics 365 apps have a different in months before the GA dates. These updates apply to both Power
cadence from the major releases. For Platform and Dynamics 365 apps. We encourage you to opt in early to
example, Dynamics 365 Marketing
and Dynamics 365 Portals have test and apply the release. The releases are production-ready and fully
monthly updates. Apps from ISVs from
supported even when applying prior to the GA date. Activation for
AppSource, Microsoft’s app marketplace,
may also have a different cadence. You major updates is automatic through safe deployment processes for the
should consult with the ISVs for any
third-party apps you’re using.
region where the Dynamics 365 instance resides, on the deployment
dates specified for the region.

Implementation Guide: Success by Design: Guide 20, Service the solution 613
 Fig.
20-5 Feature release
October October April April
public preview release public preview release

May June July Aug Sept Oct Dec Jan Feb Mar Apr May

Access to latest features (often weekly) Prep time latest release Latest release
(admin opt-in) (admin opt-in) (automatically applied)

F1
F2
F3 Features Features
F4 F1-F6 F1-F6
F5
F6 Access to latest features (often weekly) Prep time latest release Latest release
(automatically applied)
(admin opt-in) (admin opt-in)

F7
F8
F9
F10 Features Features
F11 F7-F12 F7-F12
F12

Continuous updates that can include feature code which has no UI impact (often weekly)

Opt-in to experience all UI features coming in the next scheduled update (preview)

All GA functionality is automatically deployed on a semi-annual schedule

Release readiness
We strongly recommend that organizations work updates into their
internal project release cadence. To do so, take the following steps:
▪ Step 1: Opt in for early access  Before you apply the changes
to existing production or non-production environments (which
may disrupt users and developers), we recommend you create a
new instance. You can’t revert back to the previous version, so all
testing should be done on a new instance. Take a copy of the test
environment that has your latest solution and data and create a
new instance from it. Enable early access to apply the new release
capabilities. After you opt in, some features are turned on by
default; others may require an administrator to explicitly configure

Implementation Guide: Success by Design: Guide 20, Service the solution 614
 them. The details are documented in the release notes.
▪ Step 2: Test  Run regression testing to make sure that the solu-
tion continues to function as expected. Early opt-in features are
production-ready and fully supported, so if you encounter any
errors, you can submit a service request to report and get help
with issues. Another important aspect of enabling the early access
capabilities is that some UI and navigation changes may impact
users. Work with your user group to do a side-by-side comparison
between the current and opt-in versions. Use the release notes
to identify the areas where UI or navigation changes have been
made. Document the change with screenshots to be shared with
the users. Depending on the significance of the changes, it may
warrant some level of end user training as well as communications
out to the group.
▪ Step 3: Deploy  When the testing is complete, you can turn
the early access features on in other non-production instances.
This ensures that new features and customizations by the project
team are developed on the upcoming early release solution. You
may keep one or more instances without the early access features
turned on to support production bug fixes or anything that needs
to be deployed prior to enabling the early access features in the
production environment. When it’s time to deploy your solutions
to production, you enable the early access features on the produc-
tion instance. You should notify business users of the new features
(within the product as well as anything custom built by the project
team) and any changes to how end users will navigate through the
system. Timing the deployment is important. Microsoft doesn’t
Refer to the Dataverse storage capacity
guidance to understand the impact on your recommend opting in at the same time you have a new project
storage allocation for your tenant when
creating new instances from a backup.
release in the production environment. If you encounter deploy-
ment issues, it’s easier to troubleshoot when you’re not deploying
multiple solutions to the environment.
Monitor service health

Service updates

Environment
Environment maintenance
maintenance Protecting your solution and providing continuous availability of
Continue the business service is your primary goal as the system administrator. In a cloud
application journey environment, these maintenance jobs are automated, but it’s critical

Implementation Guide: Success by Design: Guide 20, Service the solution 615
 for an organization to have a strategy so that these routine jobs are
appropriately configured and scheduled. In some cases, you may need
to perform these tasks manually but still in alignment with your overall
planning and strategy.

User access and security

Members of business and technical teams fluctuate. New employ-
ees are onboarded and existing ones may change roles or leave the
company. As an administrator, you need to work with other teams to
manage user access, assign the appropriate licenses, and assign securi-
ty roles that meet their business and technical needs. You may want to
hand over some of these responsibilities to business and IT managers
who are more familiar with their resources’ user profiles and know if
any frequent changes in their roles require immediate changes.

Your organization’s data is likely one of the most important assets

you’re responsible for safeguarding as an administrator. The ability to
build apps and automation to use that data is a large part of your com-
pany’s success. You can use Power Apps and Power Automate for rapid
build and rollout of these high-value apps so that users can measure
and act on the data in real time.

Apps and automation are becoming increasingly connected across mul-

tiple data sources and multiple services. Some of these might be external,
third-party services, possibly even social networks. Users generally have
good intentions, but they can easily overlook the potential for exposure
from data leakage to services and audiences that shouldn’t have access.

Microsoft recommends that you create data loss prevention (DLP)

policies to act as guardrails to help prevent users from unintentionally
exposing organizational data. You can scope DLP policies at the en-
You can manage users and licenses in both vironment level or tenant level, which provides the flexibility to craft
Finance and Operations apps and Customer
Engagement apps. sensible policies with the right balance between protection and pro-
ductivity. For tenant-level policies, you can define the scope to be all
You can also use Azure Active Directory
(Azure AD) groups to simplify user environments, selected environments, or all environments except those
management and provision and
deprovision users.
you specifically exclude. You can define environment-level policies one
environment at a time.

Implementation Guide: Success by Design: Guide 20, Service the solution 616
 Dynamics 365 environment management
Dynamics 365 provides point-in-time restore (PITR) capabilities for
databases. This means that all databases are backed up automatically
by the system and retained for a set number of days. In the event of
Although the backup and recovery accidental corruption or deletion, administrators can choose to restore
operation is dependable, it could also be
time-consuming depending on the size the database from any of the backups taken. The automated backup
of the backup.
system and PITR provides a zero-admin way to protect databases.
In a Customer Engagement apps scenario
when solution imports fail, it’s often
better to fix the import issue instead
If your organizations require a proactive approach to manually take
of restoring from a backup. Fixing the backups (such as before a deployment of a new release of your solu-
import should take significantly less time
than restoring. tion), the administrator may be called on to assist. You should perform
these tasks in line with your organization’s environment strategy.

Data management
Data is central to all applications. It drives business decisions through
analytics and artificial intelligence. It also reveals crucial information
about the overall health of the system and what administrators need
to do for maintenance. The theme of this chapter is to be proactive
in planning and strategizing around the upkeep of your system. Data
maintenance is no different. In this section, we discuss the service
aspects of data management. To explore the broader topic, refer to
Chapter 10, “Data management.”

Refer to Chapter 9, “Environment

strategy,” to explore the importance Data volume grows quickly. Dynamics 365 has routine jobs in place to
of having a strategy for creating and
manage data that maximizes system performance at the platform level.
maintaining environments.
You can also perform maintenance jobs like reindexing databases that
Learn more about database movement
operations in Finance and Operations
adjust to variations in transactional data. But more improvements can
apps and review environments for more always be done to better fit how you use the services. Developing a
information about managing instances
for customer engagement apps. strategy to archive and delete data is an important step toward main-
taining a well-functioning system.

How fast is the data growing?

You first need to understand how fast data is growing in your envi-
ronment. The rate of growth is a key metric to monitor. Set alerts to
let you know when this growth rate exceeds a threshold determined
by your organization.

Implementation Guide: Success by Design: Guide 20, Service the solution 617
 The rate of growth can fluctuate depending on the number of users
or even during certain times of the year if your business practice has
special circumstances that may impact record creation. Monitoring
You can find details on storage allocation
and purchasing additional storage for
storage growth and using historical trends will help estimate data
Finance and Operations apps in the growth. This information can help you determine how often the data
Dynamics 365 Licensing Guide.
archiving and removal process should take place.
Read about the storage capacity model for
Dataverse and how to check storage growth.
What data can be removed?
Planning starts with identifying the types of data that need to be
stored over particular timeframe. Besides cost, there are performance
implications of having a large dataset. Building a data removal and
retention strategy will help determine what to do when data is no
longer useful.

Transactional data may help you make key business decisions, assist
customers, and use AI to determine the next best action. But after
years of inactivity, the data may be taking up storage space and not
providing any value.

You may have logging requirements like auditing customer records

to understand which users have access to read and modify personally
identifiable information. But excessive data auditing leads to rapid
storage growth and can negatively impact performance. Storing large
files and attachments to emails is also an area in which datasets can
quickly expand. Planning around what needs to be stored and for how
long should be done deliberately and conservatively.

You may have logs, notifications, and other system records that you
can delete with no impact to the business. You may also have other
Review the cleanup routines for Finance transactional data that can be deleted. Because Dynamics 365 appli-
and Operations to delete historical logs and
cations have heavy parent-child relationships between records, pay
notifications. You should only run these
cleanup routines after the business has careful attention to how records are deleted and any impact to related
completed a detailed analysis and confirmed
that the data is no longer required.
records. Look for triggers that run extension code or workflows when
a record is modified or deleted. A delete operation could potentially
You can also free up storage space for
Customer Engagement apps. write a new entry in the audit log to record the transaction. You must
account for all these things when planning for bulk deletion.

Implementation Guide: Success by Design: Guide 20, Service the solution 618
 Archive and retention strategy
What happens if the database is growing too large, and you’re seeing
the impact of it, but deleting records isn’t an option? You may need
to retain the data due to organizational, compliance, or regulatory
requirements. In this case, you can archive the data instead.

We recommend the following strategy:

▪ Identify retention requirements  Get a full understanding of
the requirements for the types of records that you need to retain
and how long you need to do so.
Your organization may be subject to rules
such as the GDPR that give users specific ▪ Develop an archiving strategy  Once you have a clear under-
rights to their personal data. To comply
standing of what data to retain, plan for how to archive these
with privacy and security regulations, you
may need to respond to DSRs to delete a records. Where will you store the archived data? How will you
user’s personal data.
access this data when needed? What is the process to periodically
We recommend reviewing the guidelines move records from the Dynamics 365 systems into the archives?
for privacy and personal data for
Microsoft Dynamics 365. ▪ Remove the data  For any data that doesn’t meet the retention
criteria, plan out your process to remove it from the archives.

Monitor service health

Continue the business
Service updates
application journey
Environment maintenance
The importance of continuing your education on the business applica-
Continue the tion platform is often overlooked. In the on-premises software world,
business application you would go live with a version of your solution and it would mostly
journey
remain the same until you decided to develop or purchase add-on
functionality. As mentioned earlier in the ser vice update section, the
benefit of continuous updates in Dynamics 365 is that you always have
access to the newest features to help your business. This also means
you have a constant stream of new and better ways to meet your
business and technical objectives through your solution.

For example, let’s look at integrating Dynamics 365 with external sys-
tems. It wasn’t very long ago that you needed custom code modules in
order to efficiently pass data to and from Dynamics 365. But with tools
like Power Automate and Azure Logic Apps, you can build very power-
ful integrations through configuration with little to no code.

Implementation Guide: Success by Design: Guide 20, Service the solution 619
 Reporting is another good example. The out-of-the-box experience
for reporting had limitations in Dynamics 365—you could only build
reports with data stored in Dynamics 365. Now, with tools like Power BI
and its capabilities to build reports from data in and outside of Dynamics
365, you have much mor e flexibility to quickly build out and embed
powerful reports. Also, some advanced Azure and Microsoft 365 services
coexist with Dynamics 365, and are going through the same type of
evolution. Everything from business functionality to ALM build and
deploy tools are constantly seeing incremental improvements.

The only way to truly understand these benefits and how to best apply
them in your solution for the most return on your IT investment is by
continuing your education. Gain an understanding of where the busi-
ness application domain is headed. Strive to understand what is in the
Dynamics 365 roadmap, why it’s included, and how it can make your
solution and your organization more efficient and robust.

Each year, Microsoft creates many events and learning opportunities to

help developers, IT and business professionals, partners, and educators
extend their expertise. Here are some great resources for furthering
your education:
▪ Microsoft Events consolidate links to stay up to date on upcoming
events for all kinds of audiences, both business and technical
focused.
▪ Microsoft Business Applications Summit helps you maximize on
using business applications through live and on-demand record-
ings and discussions.
▪ Microsoft Ignite is for anyone who is motivated to be on the fron-
tier of innovation and tech. There are live and on-demand sessions
tailored for different audiences.
▪ Experience Dynamics 365 provides access to Dynamics 365 experts
and the Insider program to participate in preview features. You
can also provide feedback to Microsoft engineers and vote on new
product features.
▪ The Dynamics 365 YouTube channel delivers the latest product
news, announcements, events, demos, and customer success stories.

Implementation Guide: Success by Design: Guide 20, Service the solution 620
Conclusion
In summary, take steps to be aware of your solution performance, be
proactive in taking action, and be prepared with a solid strategy for
maintaining your environments. Keep up on new trends and tools that
can help improve your solution and your organization.

It all starts with visibility into the health of the system through proper
monitoring. Having the right telemetry and managing notifications
from the system as well as Microsoft will help you to prioritize and act
to address maintenance needs.

A large part of the overall maintenance is simply understanding

the continuous update model of the Dynamics 365 service. We rec-
ommend that you be proactive in planning, communicating, and
deploying service updates to gain the benefits of new features and
enhancements made available throughout the year.

Appropriate maintenance of your environments and data will help

protect and optimize your solution. This is a team effort, not just the
responsibility of IT administrators. By working across the business and
technology resources, you can make sure that the solution performs
well technically and maximizes results for the user community.

Finally, just as your solution is continuously updated, organizations

also need continuous education in trends and tools related to your
business applications. With awareness of the product roadmap as well
as improvements in ancillary products, you can maximize your organi-
zation’s investment in the Dynamics 365 platform.

621
 Case study
Fruit company learns the
importance of servicing
the solution
An agricultural business that grows fruit implemented Finance and
Operations apps as soon as the cloud version of Dynamics 365 became
available. The company has been a global leader distributing fruit
across different regions, and warehouse operations and transportation
are part of the company’s core business.

Before implementing the Finance and Operations apps, the fruit

company was using Dynamics AX 2012, and was familiar with the ALM
of that on-premises enterprise resource planning (ERP) solution, but
moving to the cloud required a mindset change.

Dynamics 365 has been evolving since its initial release, when the appli-
cation and the platform were released as separate components. By the
time continuous updates and a single, unified version became the norm
for Dynamics 365, the fruit producer’s operations in the cloud were
mature in taking updates under this modality. The company was ready
to adopt the modernized update pattern and take advantage of the
continuous innovations from Microsoft. They also wanted to fulfill one of
their expected returns on investment (ROIs) by entrusting Microsoft to
bring new functionality, instead of developing it on their own.

Implementation Guide: Success by Design: Guide 20, Service the solution 622
 The fruit company’s solution involved standard Dynamics 365 apps,
some extensions, and an ISV. As the direct implementer of the entire
solution, the ISV created a strategic partnership with the company and
provided IT outsourcing services.

In the beginning, the incentive for the company to update Dynamics

365 with new releases was the continued support from Microsoft per
the software update policy. The app functioned properly and applying
the releases was manageable, so the company didn’t have hard re-
quirements to stay up to date with the release cadence.

But the company noticed that while the ISV kept the solution up to
date with the Dynamics 365 releases, the ISV always provided an up-
date of their solution using Microsoft’s last update ring, or missed it
entirely. Because of this, the company had to apply Microsoft’s updates
late in the release cycle or, in some cases, not at all.

To grow and distribute fruit effectively, the company relies heavily on

time-sensitive warehouse operations where every hour matters as fruit
is moved from the fields to climate-controlled warehouses to shipping
containers that are sent worldwide as quickly as possible to avoid
spoilage. They also were expanding into distribution of seasonal fruits
in new markets, which made management of warehouse operations
even more critical.

Then Microsoft notified the fruit company about new functionality for
advanced warehouse operations in an upcoming release of the Supply
Chain Management app. Because of the expansion of the company’s
operations and the complexity of their warehouse management, these
new features were crucial to increasing productivity and saving time
and resources while managing the expansion.

To adopt this new functionality, the company had to test the features
in advance to align their processes. They realized that being late on up-
dating standard features wasn’t workable, and they wanted to optimize
their ALM and validate their solution in early update rings. To be ready
for their peak season, completion of testing was time-sensitive, and
the company would only have enough time to adopt new functionality

Implementation Guide: Success by Design: Guide 20, Service the solution 623
 and do proper regression testing if all solution components were in
place and ready. So, they asked their ISV to provide software earlier
than the general availability of their new version.

The challenge came when the ISV wasn’t aligned to that timeline. The
ISV’s usual practice of adopting Dynamics 365 releases when they
became generally available, in the last ring, meant the fruit company fre-
quently was one version behind on Finance and Operations apps release.

The ISV wanted to honor their partnership and be flexible in meeting

the timing requirements for updates, but due to dependencies on their
internal roadmap and ALM strategy, the ISV couldn’t make the software
available to meet the company’s timeline. As a result, the fruit company
couldn’t take advantage of the Supply Chain Management app’s
advanced warehouse operation functionality in time for their expansion.

Even though the company’s ALM strategy could adapt to Dynamics

365’s continuous update delivery model, reliance on the ISV’s software
prevented them from using the latest version of Dynamics 365.

After conversations with the fruit company, the ISV agreed to align
with Microsoft’s update rings and give the company, and their other
customers, an opportunity to test their entire solution when the stan-
dard release was available.

Servicing the solution involves more than just Microsoft providing

continuous updates. Microsoft customers must synchronize updates to
their own extensions and any ISV solutions they’re using, so they can
benefit from the latest Microsoft innovations as soon as new features
are available.

Implementation Guide: Success by Design: Guide 20, Service the solution 624
21 Guide
Transition
to support
Implementation Guide: Success by Design: Guide 21, Transition to support 625
Poor service is always more
expensive than good service.

Introduction
When introducing a new system into an organization,
we need to think through the various areas that will
influence how the project transitions from project
mode to support mode.
In this section, we discuss how to construct a strategy to help you
prepare, define, and operate a support model.

Fig.
21-1 Organizations that spend the necessary time and energy to construct
a strategy that explicitly addresses how to create a fit-for-purpose
Support scope support organization for their Dynamics 365 application tend to have
▪ Enterprise architecture better user satisfaction, better adoption of the system, and therefore
▪ Business and IT policies
▪ Dynamics 365-specific considerations
higher-quality outcomes for the business.
▪ Business processes
▪ Business continuity
▪ System update cycles If this is the first Dynamics 365 business application for your company,
you may not have experience in setting up the support organization,
Support models support infrastructure, and support procedures for this application.

▪ Support model options

Establishing a support organization for a new system (for example,
▪ Support organization Dynamics 365) needs more due diligence and may require a different
perspective that better reflects the new technologies and new ways of
Support operations working that come with it.
▪ Transition
▪ Requirements management
This chapter is organized to reflect the journey of preparing, defining,
▪ Change management
▪ Hypercare and operating support through the categories of support scope, sup-
port models, and support operations (Figure 21-1).

Implementation Guide: Success by Design: Guide 21, Transition to support 626

 Support scope
A key aspect in preparing for production support is compiling the list
of topics, tasks, impacts, and responsibilities that supporting the new
Dynamics 365 application needs to cover. The support scope defines
the “what” of the support model.

Support models
The scope definition influences the decisions that need to be made to
identify the best support model and the constitution of the resulting
support organization. This helps define the “who” of the support model.

Support operations
Finally, we discuss the distinct support requirements that emerge from
the transition and hypercare project phases.

Support Support scope

scope Support scope activities should be started in the Initiate phase. It may
be tempting to defer defining the scope until much later in the project
Enterprise architecture because it’s not seen as urgent compared to getting the project off
the ground and creating the system solution. However, as this chapter
Business and IT policies
discusses, early attention to this topic enables the support team to be-
Dynamics 365-specific come engaged in the project from an early stage, which pays dividends
considerations during the project lifecycle.
Business processes

Business continuity
Enterprise architecture
System upgrade cycles
In most organizations, the Dynamics 365 system is embedded within
the wider enterprise system landscape. The Dynamics 365 application
has connections to multiple systems and to underlying and coexisting
technologies. When considering the strategy for defining the support
model, the focus is often solely on the Dynamics 365 application archi-
tecture. It’s worth accounting for the changes, constraints, procedures,

Implementation Guide: Success by Design: Guide 21, Transition to support 627

 Enterprise and influences from the surrounding architecture on the Dynamics 365
architecture applications, and vice versa; how will the enterprise architecture and
Business and IT policies environment be impacted by installing and operating the Dynamics
365 application?
Dynamics 365-specific
considerations
Firstly, from the enterprise system architecture documents, confirm
Business processes
that you have identified all the various systems, software, and data that

Business continuity interact with Dynamics 365 applications. You should make a distinction
between new systems being introduced into the enterprise architec-
System upgrade cycles ture and those that may be influenced by the new systems.

During the project, the new Dynamics 365 system will probably be
implemented within a special sandbox (test environment) and not
necessarily be subject to all the influences and rules that the pro-
duction system is subject to. This also applies to the third-party test
systems that are part of the middleware or integrations. For example,
production environments have more limited access to the Dynamics
SQL database, and the process by which custom code is promoted to
production or how the Microsoft system updates are applied isn’t the
same. You should specifically examine the implications of the produc-
tion system environment on support operations and not rely solely on
the experiences of the test environment.

Identify the specific, new, or changed support requirements impacted

by the following:
▪ Enterprise-level architecture configurations in areas such as
Microsoft Azure, operating systems, any existing Microsoft 365
tenants, Azure storage, and Azure Active Directory
▪ Dynamics 365 Azure environments managed by Microsoft (for
example, the Dynamics 365 production environment)
▪ Identity and access management systems, such as operating
systems, browsers, mobile devices, and firewall configurations
▪ If this is the first significant cloud, software as a service (SaaS)
system in the enterprise, the related system architecture, and
servicing needs
▪ Systems to which the Dynamics 365 application interfaces directly
▪ Middleware systems to which the Dynamics 365 application may
connect

Implementation Guide: Success by Design: Guide 21, Transition to support 628

 ▪ Any systems that will be decommissioned but have residual tasks
that may need to be reabsorbed elsewhere by the support team
▪ Any document management systems such as Microsoft SharePoint
or third-party systems
▪ Any code configuration management systems such as Azure
DevOps or third-party systems
▪ Reporting systems to which the Dynamics 365 application interfaces,
including Azure Data Lake or data warehouse

The impact of, and on, the surrounding architecture can be difficult
for the customer project team working on the Dynamics 365 business
application to fully identify. In almost all cases, you need the enterprise
architects from IT and the business to be involved in identifying the
changes. Some changes may also impact the roles of individuals formally
part of a support organization and those in peripheral organizations.
Enterprise architecture

Business and
IT policies Business and IT policies
Dynamics 365-specific
considerations All new systems need to think about how they will operate within the
Business processes wider organization’s policies and standards. When Dynamics 365 is
introduced into the enterprise, it needs to follow the applicable poli-
Business continuity cies already in place and may require new operating policies. In either
case, you need to review the existing policies and standards to deter-
System upgrade cycles
mine which policies to add, which to review, and which to apply to the
support model (Figure 21-2).
Fig.
21-2

In many cases, the policies apply not only to the creation of the support
Group and company policies
and standards model and its scope of responsibility, but also to how it operates as an
Dynamics 365 application-level policies organization and how it addresses the lifecycle of a support request.

Dynamics 365 application-level security

and access management

Data classification and retention Group and company policies and standards
Group and company-level security and When evaluating group-level policies or standards, you may have to
access management
review the impact of business policies and procedures on various levels:
Regulatory compliance
▪ Contracting with and managing third-party vendors 
The support model needs to include some type of contract with
Support scope
technology partners and with Microsoft

Implementation Guide: Success by Design: Guide 21, Transition to support 629

 ▪ Operating hours  Current and future operating hours for the
business can directly impact the working hours for support
▪ Financial period end and seasonal deadlines  This may impact
the level of service expected at times when the business has critical
business activities
▪ IT policies on change control on enterprise IT assets  This
may impact the access and approval chain for changes required
for the Dynamics 365 application troubleshooting or servicing (for
example, creating new Azure virtual machines, copying databases
between environments, and building IP allowlists)

The overall support model on Dynamics 365 business applications

operates across many of the company’s business processes and is
therefore shaped by the relevant business policies.

Dynamics 365 application-level policies

Some policies are managed in the system by the business adminis-
trators, and some may be delegated to the support team. Even if the
administration of a policy is the responsibility of the business leads, the
monitoring and auditing of the business process is often the support
team’s responsibility.

You could set up some of these policies within the Dynamics 365 ap-
plication, such as new vendor approval policies, customer credit limits,
purchase order approval thresholds, and travel and expense policies.
The support team may need to help provide information on compli-
ance or enforce compliance as part of their duties.

Dynamics 365 application-level security

and access management
Administration and review of application-level security is usually an
area of responsibility of the support team. The support team needs to
prepare by gaining an understanding of the Dynamics 365 role-based
application security model and the related access policies.

Implementation Guide: Success by Design: Guide 21, Transition to support 630

 Group and company-level security and
access management
In addition to the security and access policies and procedures that
you need to define for Dynamics 365 applications, you may have to
consider enterprise-level security policies that intersect with the appli-
cation-level security requirements.

In larger organizations, enterprise-level security and access manage-

ment may be handled by a separate IT organization or a separate IT
team. Some of the support team members will need elevated access
not just at the application level, but possibly for other systems within
the enterprise.

The Dynamics 365 support team needs to work with these other
enterprise IT and business teams to define the rules and procedures
for managing some of the security topics that may impact Dynamics
365 applications:
▪ Azure Active Directory groups and identities
▪ Single sign-on (SSO)
▪ Multifactor authentication
▪ Mobile device authentication and management
▪ Authentication and management for custom applications working
on Microsoft Dataverse (such as Power Platform apps), which
requires an understanding of Dataverse security concepts
▪ Application access for third parties (such as vendors and customers)
▪ Application access for third-party partner support organizations
(such as technology partners and Microsoft)
▪ Service account and administrator account use and management
▪ Password renewals and security certificate rotations
▪ Secure and encrypted communications within the enterprise and
Identify any outside the enterprise (such as those involved with integrations with
internal systems, or with external web services or banking systems)
enterprise-level
policies that The Microsoft Trust Center can help your organization consider overall
intersect with security and managing compliance in the cloud. Chapter 12, “Security,”
application-level provides a more detailed discussion of security for Dynamics 365
business applications.
requirements.
631
 Data classification and retention
Consider how the organization’s data classification and retention pol-
icies reflect on and need to be expanded to include the new Dynamics
365 application:
▪ How is the support team expected to enable and enforce these
policies?
▪ What is the impact on the backup, restore, and archive process?
▪ What is the impact on creating and managing database copies?
▪ Do any data classification properties flow between systems, or do
they need to be recreated or audited by the support team?

Regulatory compliance and policies

In addition to any data retention policies, the organization’s businesses
may be subject to national and supranational regulations. Some may
impact business processes that are expected to be assisted by the
support team (such as e-invoicing and digital tax).

The support team may be expected to support, monitor, and audit

other regulations on a regular or ad-hoc basis, such as General Data
Protection Regulation (GDPR) or Health Insurance Portability and
Accountability Act (HIPAA).

All of these different areas of business and IT policies shape the nature,
size, and scope of the support organization. Early examination of these
factors will help the team be effective from the start.

Enterprise architecture
Dynamics 365-specific
Business and IT policies

Dynamics 365-specific
considerations
considerations In this section, we examine some topics specific to Dynamics 365 busi-
Business processes ness applications that we should consider when designing the support
model. These can be broadly grouped by operational and maintenance
Business continuity
topics and business process topics.
System upgrade cycles

Implementation Guide: Success by Design: Guide 21, Transition to support 632

 Operational and maintenance topics
We can further divide operational and maintenance areas into the
management and support of the following:
▪ Dynamics 365 environments
▪ Integrations
▪ System and data maintenance
▪ Performance

Dynamics 365 environments

When Dynamics 365 is in production, various aspects involved in sup-
porting the system environments need to be factored into the support
model preparation:
▪ Maintenance and operational tasks (both scheduled and ad hoc)
such as environment data refresh, database movements between
environments, configuration moves, and code promotion
▪ Support organization roles and resources needed
▪ Skills required to support the environments
▪ Resulting budget impact

Typically, you need to apply these tasks and considerations for the
following environments:
▪ Dynamics 365 application support environments, which are recent
copies of the production system
▪ Test environment for testing the next versions of the application
software
▪ Development environments
▪ Any related data migration, training, or integration environments
▪ Test environments for integrated systems

This is not an exhaustive list of environments, but should help prompt

the right discussions related to supporting the environments. For a
more detailed discussion on the Dynamics 365 environment strategy,
refer to Chapter 9, “Environment strategy.”

Implementation Guide: Success by Design: Guide 21, Transition to support 633

 Integrations
In addition to defining the environments required for supported
operations, it’s worth delving a little deeper into the support require-
ments specific to managing integrations. When defining the support
requirements for all the integrations, consider what areas are in scope
for the support organization:
▪ Data management at either ends of the integration
▪ Security of the integration services
▪ Performance of the integration
▪ Auditing or monitoring of the integration
▪ General troubleshooting analysis and stopping and starting of
integration services

The level of skill and effort required to manage integrations depends

on their complexity, criticality, and robustness. Designs of custom
integrations that are tolerant of variations of data distribution and
variations in volumes, are resilient to errors, and have automated
self-healing mechanisms will require less support.

System maintenance requirements

As part of the definition of the system, the definition of maintainability
requirements is an area that doesn’t always get the attention it de-
serves. This can manifest itself in unplanned activity for the support
organization. So, during the Initiate phase of the project, the current
support organization should look to derive the specific maintenance
requirements for managing the new Dynamics 365 implementation.

The level of skill Because Dynamics 365 applications are cloud and SaaS-based, many

and effort required maintenance tasks and responsibilities that are common to on-prem-
ises solutions are now managed by Microsoft. In general, the burden
to manage of infrastructure provision and maintenance for production systems is
integrations reduced, which leaves more time to focus on managing and improving
depends on business process performance.

their complexity, Define the system maintenance requirements and what is within the
criticality, and scope of responsibilities for the support teams. Typically, these are in
robustness. the following areas:

634
 ▪ Servicing the non-production environments, which can include:
▫ Requesting and configuring new Dynamics 365 environments
▫ Requesting and configuring database copies and restores
between environments
▫ Managing any customer-managed, cloud-hosted environments
▫ Performing specifically requested backups
▪ Managing system operations, which can include:
▫ Assigning users to security roles
▫ Reviewing and running periodic system cleanup jobs
▫ Managing system administration messages and notifications
▫ Batch calendar management
▫ System update calendar

Data maintenance requirements

Data maintenance is a significant part of operating a business system,
and it’s essential that the scope of data management is clearly de-
fined for the support team. Much of the data maintenance will be the
responsibility of data stewards from the operating business, but the
support team may be expected to manage some critical areas.

Typically, in a Dynamics 365 application, a support team may be man-

aging system administrator data, some types of primary data, security
and access data, data related to periodic processes, and so on; for
example, assigning a user to a security role, or adding or revising a new
customer attribute.

Performance management
Performance management for a Dynamics 365 business application is
a mix of tasks and responsibilities for Microsoft and for the customer. In
this section, we consider the implications on the support model.

The support team needs to have some way to proactively monitor and
respond to any questions from the users on performance. Therefore,
the support team needs to be able to do the following:
▪ Understand the impact of the reported performance issue on
the business

Implementation Guide: Success by Design: Guide 21, Transition to support 635

 ▪ Assign the relevant priority
▪ Diagnose the issue across the potential root causes from config-
uration, data, custom code, independent software vendor (ISV)
code, standard code, and so on to determine the root cause, or to
hand it off to a specialist team (internal or external) with a mean-
ingful starting point
▪ Reproduce the steps and circumstances to demonstrate the per-
formance issue (and have the necessary resources to do so)
▪ Communicate the problem to other more technical specialists and
work with them to resolve the issue

To better detect processes beginning to show poor performance, it’s

important to have a baseline definition of required performance for
critical processes, also expressed as performance goals. Having ac-
tionable performance goals allows for more focused monitoring and
proactive actions to stop any decline in performance quickly.

Because performance troubleshooting is a specialist skill, the internal

support team may not have all the skills and experience to examine
and correct the performance of custom development. It’s also a skill
that may only be needed sporadically, so you may need to consider
planning for access to external teams that can use performance tools,
interpret the results, and decide on corrective actions.

Enterprise architecture
Business processes
Business and IT policies
As a business application, supporting the users in the day-to-day pro-
Dynamics 365-specific
cesses is a key function of a Dynamics 365 support organization. The
considerations
support organization is expected to provide coverage across all the key
Business processes business process, or at a minimum, know where to direct questions and
issues that were agreed as being outside their support scope.
Business continuity

System upgrade cycles Based on the definition of the key business processes in scope, consider
the following for each process:
▪ What is the level of expertise needed in the business process?
▪ What are the likely support tasks and queries related to this process?
▪ What is the type of usage, frequency, and volume?

Implementation Guide: Success by Design: Guide 21, Transition to support 636

 ▪ How critical is this business process to the business outcomes, and
what degree of priority do the support issues need?
▪ What is the type and level of skills needed to support this process?
▪ What processes require resources aggregated at a department,
workstream, or value chain level?
▪ What are the interactions with other systems and communication
protocols required?
▪ Does this process require the support personnel to have specialist
security or compliance training or certification?
▪ Do any special requirements need support services outside of the
anticipated support hours (outside of business hours or during
holiday periods)?

An important factor when determining resources and escalation pro-

cedures is the level of ownership and responsibility that the business
process leaders are expected to take from their own organization to
support the day-to-day operations for their processes.

These questions and more will help shape the operating model for the
support team.

Enterprise architecture
Business continuity
Business and IT policies
Many organizations need to have a business continuity strategy and
Dynamics 365-specific exercise regular checks of business continuity in case of a system-down
considerations
disaster. This may be required by external regulations or by internal
Business processes policies. In any case, the support organization is probably expected to
play a major role.
Business continuity
System upgrade cycles Depending on the size and complexity of the system landscape and
the types of disaster scenarios being exercised, this may need a signifi-
cant amount of preparation, coordination, and timely communication
between multiple parties.

As part of preparing for the establishment of the support organization,

you should consider the definition of the business continuity strategy
and what resources, skills, and tooling are required.

Implementation Guide: Success by Design: Guide 21, Transition to support 637

As a cloud-based SaaS service, Microsoft provides the production
infrastructure, platform, and application-level business continuity
services. For Dynamics 365 Finance and Operations apps, the opera-
tions as listed in the Finance and Operations apps service description
provide details of the services and service-level agreements (SLAs)
from Microsoft and the expected responsibilities and actions from the
customer. You should bear this in mind when you build your business
continuity planning, which should reflect your specific systems land-
scape and scenarios.

Of course, the Dynamics 365 business application may not be an

isolated system, it may be connected to other enterprise systems, and
your business continuity planning will need to consider your entire
enterprise architecture when defining your strategy and the related
support requirements.

Dynamics 365 Finance and Operations apps

Dynamics 365 Finance and Operations apps have specific high avail-
ability (HA) and disaster recovery (DR) features for cloud deployment.
To ensure service availability, all production environments are pro-
tected by using default Azure HA features. HA functionality provides
ways to avoid downtime caused by the failure of a single node within a
datacenter, and DR features protect against outages broadly impacting
an entire datacenter. When planning support requirements and proce-
dures, consider the implications if the disaster is severe enough for the
primary production site to be forced to switch over to the secondary
site and the DR features are warranted. In such extreme cases, you
would experience downtime for production access while access to the
secondary site is being established, and the customer team is responsi-
ble for several actions. For more information about business continuity
with HA and DR, see the Dynamics 365 Finance and Operations apps
service description.

If the production system is running on the secondary site (while the

primary site is being repaired), you may have restrictions (such as the
inability to perform package deployments) and limitations (such as
performance reduction) that the support team needs to be aware of and

638
 help mitigate to reduce the impact on the business. You may also need
to apply specific setups (such as IP allowlists) to the secondary site.

In summary, you should have a business continuity plan that includes

how the support team should plan, perform, and communicate during
a service outage, including during a severe disaster.

Dynamics 365 Customer Engagement

For Dynamics 365 Customer Engagement, HA provides mechanisms that
can reroute requests in case of failures, planned downtime (patching
and updates), and more. On the database side (relational storage), the
Recovery Time Objective (RTO) for failover is dictated by the timings
from Azure SQL Database. With diverse storage, the scheme extends to
other storage components. For DR, Power Platform brings the ability to
fail over and fail back in seconds. You can do this at the granularity of an
organization (environment). There is no data loss in the case of planned
DR failovers. In the extreme circumstances of an unplanned DR, well-
defined processes are applied by the administrators of the Dynamics 365
(online) datacenter to recover from a service interruption; however, the
support team needs to help with any potential data loss.

Enterprise architecture
System update cycles
Business and IT policies
Microsoft Dynamics 365 application updates are one of the changes
Dynamics 365-specific that the support team probably needs to manage. From the perspective
considerations
of defining the scope of responsibilities for the support organization,
Business processes you must understand what is involved and plan for the updates.

Business continuity
Creating a calendar for the Microsoft updates helps the support team

System upgrade cycles plan for the resources and effort associated with the following:
▪ Microsoft Dynamics 365 updates
▪ ISV updates
▪ Updates for custom code
▪ Associated testing, including any regression testing and release
management

Implementation Guide: Success by Design: Guide 21, Transition to support 639

 We saw in the previous section that the solution is unlikely to remain
static for long, and the solution functionality can change because of
new and backlog requirements. For details on Microsoft system up-
dates, refer to Chapter 20, “Service the solution.”

Support Support models

models Defining the support model is ideally started in the Implement phase
so that it can be ready and exercised prior to user acceptance testing
Support model options (UAT), can help with the transition during the Prepare phase, and be
fully effective and working by the Operate phase.
Support organization

Support model options

Thus far, we have discussed topics that help define the scope of re-
sponsibilities and actions for a support team. Now we look more at the
“who” and the “how,” starting with an examination of the spectrum of
support models.

One of the foundations of the support strategy is defining the oper-

ational model for the support organization. Some of the factors may
be driven by enterprise-level strategies, including on outsourcing
or existing commercial agreements. Some organizations reference
independent, formal IT support and service frameworks such as
Information Technology Infrastructure Library (ITIL) or ISO 20000, or
other formal standards and methodologies. These are typically adopt-
ed at an enterprise level, rather than system-by-system. Regardless of
the framework, you have some key options to choose from:
▪ Fully outsourced  This support model is more often seen in
smaller projects that don’t have many critical business processes
within their scope or have insufficient internal resources to create
a critical mass for any dedicated internal support members. This
model requires the outsourced support organization to be more
immersed in the business and understand the business well to
navigate the organization when looking for decisions, diagnostic
data, and approvals, as well as have a good understanding of the
business priorities. To succeed, it needs the business and partner

Implementation Guide: Success by Design: Guide 21, Transition to support 640

 to have a common language to communicate business needs and
technical application processes between each other so that expec-
tations remain aligned.
▪ Fully internal  A support model without partner support isn’t
seen very often because it’s a difficult model to sustain on an active
project without a larger internal team that can cope with peaks and
troughs of demand and has the necessary distribution of skills across
business workstreams, technical areas, and servicing, including any
specialist skills. If internal resources are constrained, you run the
risk that the business stops being ambitious and outward looking
in driving improvements and encouraging innovation. The solu-
tion can become stagnant, where assumed stability can come at
the expense of extracting optimum value from the application.
▪ Mixed  This is probably the most common model. It has a bal-
anced blend of internal resources that develops expertise in the
application and retains a deep understanding of the business.
Partner organizations can offer strategic advice based on their
wider experience across multiple implementations and provide
support for any custom code, resources to cope with peaks, and
specialists when required.

Considerations per support level

In this section, we break down the business considerations by each
support level (Figure 21-3).

First level
Many customers have a concept of identifying and nominating super
users within a business unit. Typically, the super users gained their
deeper knowledge of the system and processes in their previous role as
subject matter experts (SMEs) during the project cycle.

Super users can provide the following benefits:

Super users help ▪ Provide immediate assistance to a user to unblock someone in

organizations scale most simple cases

▪ Filter out the simple questions such as “How do I do this?” or “Is
out application this what I should expect?”
support. ▪ Provide ad-hoc training at a business unit level

641
 ▪ Triage and communicate issues in a way that makes it easier for
the helpdesk or resolving authority to understand the issue, repli-
cate it, and fix it rapidly
▪ Collate best practices, FAQs, and issues, and provide the core team
better data on areas of priority for them to tackle
▪ Provide early warning on issues that have the potential to escalate
Fig. and help with better adoption by the business
21-3

Fully outsourced Fully internal Mixed

Super users per business unit Super users per business unit
1 or function
1 or function
1 Super users per business unit or function

Internal helpdesk to route to Internal helpdesk or dedicated Internal helpdesk or dedicated

2 the partner
2 Dynamics 365 support team
2 Dynamics 365 support team

Dynamics 365 CoE or Dynamics 365 Dynamics 365 CoE or Dynamics 365
3 Partner 3 project core team
3 project core team

Partner via support agreement,

4 Partner 4 Dynamics 365 CoE 4 or partner for custom code via
support agreement

5 Microsoft support or ISV support 5 Microsoft support or ISV support 5 Microsoft support or ISV support

This model helps organizations scale out application support without

having to create a large central support team or pay external parties
to resolve day-to-day queries on internal business process or simple
application questions.

For larger business units, super users may be further divided by

functional area, such as Finance or Operations. These super users are
typically identified during the project phase and participate in formu-
lating and implementing testing (such as UAT) and may also become
the trainers for cascade training. This allows them to gain the necessary
experience and skills to provide first-line support.

Consider formalizing the super-user role with funding and consider

formal titles or dedicated time to perform that role. Formal performance
objectives and compensation also help ensure this is not just one more
assigned task. Informal super-user roles don’t always succeed because

Implementation Guide: Success by Design: Guide 21, Transition to support 642

 day-to-day operational tasks tend to be given a higher priority.

Second level
Almost all customers have some internal helpdesk function for IT
systems, and many also have them for business applications. The size
and nature of the helpdesk varies depending on the model adopted.

For a fully outsourced model, the internal helpdesk registers the issue
and assigns it to the resolving authority (internal or external). This fully
outsourced model is tougher to deliver for highly-customized business
systems compared to less customized ones. Determining the correct
resolving authority can be tricky in business systems, even assuming
the super-user roles have eliminated the cause as a business process
or training issue, you may have many different system layers, such as
an internal network, infrastructure software managed internally or by
Microsoft, standard Dynamics 365, custom code, ISV code, or integra-
tion software at either end.

Determining how to prioritize and who should drive the issue to

resolution is a typical deficiency seen in this fully outsourced model,
because very often the issue can bounce around various parties. The
other consideration for business systems is managing what business
data and process steps should be made available to the next resolving
authority. Additionally, if the partner is driving this in a fully outsourced
model, you need a clear definition of the access the partner is given to
users, the business data, and what can be further passed on the next
resolving authority.

In the fully internal model, the internal helpdesk also routes the issue
to a small internal 365 Dynamics team that triages the issue and helps
determine the best resolving authority. This team is also responsible for
ensuring the issue is driven to resolution, regardless of the number of
parties (internal or external) that may need to be involved. The differ-
ence is that the next resolving authority may be the internal Dynamics
365 Center of Excellence (CoE).

For multi-company or multi-project organizations, many create a CoE

that works with group IT to ensure the following:

Implementation Guide: Success by Design: Guide 21, Transition to support 643

 ▪ Common standards on the Azure environment
▪ Operational and maintenance policies on backups, business conti-
nuity, and disaster recovery
▪ IT security and policies (such as password policies, MFA, SSO, and
firewall rules)
▪ Regulatory needs, including access and security, data retention

In the mixed model, which many customers adopt, the internal

helpdesk routes the issue to a small internal Dynamics 365 team that
triages the issue and helps determine the best resolving authority. This
team is also responsible for ensuring the issue is driven to resolution,
regardless of the number of parties (internal or external) that may need
to be involved.

Third level
In a fully outsourced model, the partner is responsible for triage and
resolution. In the mixed model, the most common scenario is for the
internal Dynamics 365 support team to determine if they can resolve
the issue; if not, they work with the Dynamics 365 CoE and the partner
to ensure it’s well understood and correctly prioritized.

In the fully internal and mixed model, the Dynamics 365 CoE includes
functional and technical experts (including developers) who can
resolve issues with custom code. If the issue is seen to be with stan-
dard Dynamics 365 or an ISV, the CoE logs the issue and works with
Microsoft or the ISV to get it resolved.

Fourth level
In the fully outsourced model, the partner manages the process and
the customer is involved as necessary. Most customers tend to have
some parts of the solution written or managed by their internal IT team
(such as integrations), so you still need a definition of how the partner
should work with the customer.

In the fully internal model, the Dynamics 365 CoE takes on the diagno-
sis and resolution if it’s within their control. They only involve external
parties if the issue lies with standard Dynamics 365, platform hosting,
or an ISV.

644
 In the mixed model, the internal Dynamics 365 CoE or core team
typically fixes the simpler issues, but involve a partner for issues that
require deeper Dynamics 365 knowledge or for complex issues.

When the issue is resolved, the method to get the fix back into produc-
tion also requires a clear definition of the expected standards, testing
regimen, and deployment process. Even in a mostly outsourced model,
the partner drives this to a pre-production environment and the inter-
nal team deploys to production. Most customers don’t give partners
admin access to production environments.

Fifth level
Registering the issue with Microsoft support tends to be the common
escalation point, after the determination is made that the most likely
root cause is the standard Dynamics 365 software or service.

Similarly, most customers have a support and maintenance contract

with their ISV suppliers. In the fully outsourced model and in some
mixed mode models, the partner is contractually responsible to work
with the ISV for resolution.

Support model options

Support organization
Support organization As noted in the previous section, multiple operating models exist, and
within each model a spectrum of sub-options. Therefore, the specific
organization structure is influenced by the variant of the model that
is chosen. In this section, we explore some common patterns for roles
and structures.

Support team responsibilities

Any organization looking to support a business application system of
a reasonably-sized footprint needs to make sure that they have the
necessary skills and resources, not just across the business process
areas and technical topics, but also to manage data quality, maintain
and improve user proficiency in the system, and help the organization
remain compliant with processes and procedures.

Implementation Guide: Success by Design: Guide 21, Transition to support 645

 Each of these areas needs to be defined in more detail to ensure the
planning for the support model is sufficient.

Support for business process areas

Support for business processes is usually defined by the main workstreams
and includes the following responsibilities:
▪ Liaising with process owners on the overall health of the process
▪ Keeping up to date with the current business priorities in this area
▪ Gaining and maintaining a deep understanding of the steps and
relative criticality of the various functions within a business process
▪ Gaining and maintaining a deep understanding of how the
Dynamics 365 application and related systems deliver this process,
so they can provide how-to advice and analyze potential problems
▪ Analyzing, triaging, and recording potential issues
▪ Communicating with the originator of an issue throughout its life-
cycle to maintain user satisfaction, promote business awareness,
and manage expectations
▪ Curating issues through multiple resolving authorities until resolu-
tion within agreed SLAs
▪ Liaising with internal and third-party resolving authorities (such as
any partner organizations, ISVs, or Microsoft)
▪ Monitoring the lifecycle of a defect resolution from registering to
promoting a fix

Support for technical areas

Support for technical areas is usually defined by the following categories:
▪ Service operations (database copies, new environment creation,
managing system downtime windows, maintaining Dynamics 365
system quality update calendar, etc.)
▪ System operations via services (integrations, batch processes,
reporting, business intelligence, etc.)
▪ System maintenance and monitoring (system health monitoring,
Responsibilities in security role maintenance, security certificate rotations, periodic
technical support system cleanup of log files etc)
▪ Dynamics 365 application management related to break or fix
roles may overlap, errors reported by users
but with important ▪ Dynamics 365 application management such as updating system
distinctions. administration parameters

646
 ▪ Dynamics 365 application system update management (code
branching, any software development, unit testing new changes, etc.)

Responsibilities may also overlap with those for business process support,
with the following distinctions for the expected areas of understanding:
▪ Gaining and maintaining a sufficient understanding of the relative
criticality of the various functions within a business process
▪ Gaining and maintaining a deep understanding of the underlying
technical processes involved in the system functions (especially
integrations, custom code, and critical processes)

Support for data quality maintenance

In many organizations, some of the business data maintenance respon-
sibilities are passed to the support team by the business data stewards
or owners. This type of support includes the following areas:
▪ Core business data and configuration in the system as well as IT
and system administration parameters
▪ Liaising with business data stewards to determine priorities
and actions
▪ Periodic audits and reports to confirm the data quality

Support for user proficiency in the system

In some organizations, the functional experts within the support team
act as a default team to provide user training both formally and on an
ad-hoc basis. This level of support includes the following:
▪ Monitoring user proficiency within the system by reviewing the
number of how-to queries and issues created by not following the
agreed business process in the system
▪ Creating training content and conducting formal end-to-end
training of new users
▪ Creating training content and conducting informal specific functional
Plan for the training of users on request or as needed
support team’s ▪ Conducting change management activities to help improve user
involvement in adoption

more than just the New features and developments

break/fix stage of In addition to managing the day-to-day and periodic maintenance
the lifecycle. tasks and addressing queries and issues in the preceding areas, the

647
 support team may also be expected to keep track of new features and
developments (positive and otherwise) that may impact the system.

Often, the planning for the break/fix part of a support team’s orga-
nization is well thought through, but the planning for other areas of
responsibility may be neglected. For example, many support team
members may also need to be involved in helping with the lifecycle of
the next system update. This may include the following duties:
▪ Helping with assessing the impact of proposed changes on the
current system
▪ Providing insights into areas that need reinforcement based on
statistical analysis of the areas with the most issues
▪ Testing the changes
▪ Updating documentation to reflect process and system changes
▪ Training in any new processes
▪ Sending communications of the system update
▪ Managing the next release through the development, test, and
production environments
▪ Refreshing the support environments with the latest production
environment

Dynamics 365 business applications are cloud-based SaaS applications,

so Microsoft handles a lot of the traditional effort (for on-premises
systems) of managing all the layers of the production infrastructure
hardware, operating system, and application software, and much of the
monitoring of the underlying system health.

The duties and responsibilities outlined in this section aren’t meant to

be exhaustive, but rather prompt you to make sure you have sufficient
definition of the coverage, skills, and effort needed to maintain and
support a business application. Later in this chapter, we discuss the
special considerations in the transition period and in the early days of a
new system implementation (sometimes called hypercare).

This section concentrated on the tasks and activities that your support
organization may need to cover. For details on operational consider-
ations, refer to Chapter 20, “Service the solution.”

Implementation Guide: Success by Design: Guide 21, Transition to support 648

 Support organization structures and roles
When considering the support organization, each business should
consider the level of effort that is implied by the range of duties
expected as part of support and determine how to distribute the
various responsibilities.

The following is a typical set of roles (depending on the size and com-
plexity of the implementation, sometimes multiple roles may be filled
by a single individual):
▪ Business super user(s)  As we discussed earlier, super users serve
as first-line support with some form of part-time role. The level of
formal recognition of the role varies from customer to customer.
▪ Business process expert(s)  Also called SMEs or functional
leads, these experts in the support team usually have a full-time
or near-full-time role. A lead (technical architect) usually has
oversight on the full solution architecture across all the business
process workstreams.
▪ Technical architect  Also called system architect or solution
architect, this role within the support team ensures the enterprise
and Dynamics 365 technical standards are met. They have over-
sight across all the technical tasks and duties mentioned earlier in
this chapter. The technical experts, developer, and release manag-
er roles often report to the technical architect.
▪ Technical expert(s)  This role has expertise in the technical
servicing and support duties that we discussed earlier.
▪ Developer(s)  This role is responsible for bug fixes to custom
code and writing code for new custom functionality.
▪ Release manager  This role is responsible for confirming that a
release has been appropriately validated for release to production.
They’re also often responsible for other servicing and support tasks.
▪ Support manager  This manager is accountable for the support
services for Dynamics 365 applications (and often other related
systems).
▪ Business and IT stakeholders  You should consider these roles
significant because their sponsorship, direction, prioritization, and
decisions are key parts of establishing the right support organiza-
tion, defining fit-for-purpose processes for support, and meeting
business expectations.

Implementation Guide: Success by Design: Guide 21, Transition to support 649

Depending on the support model chosen (fully internal, fully external,
or a mix) and regardless of the organization responsible, you will likely
need these roles.

Organizations that are implementing across multiple business units or

companies with an incremental rollout plan should consider setting
up a Dynamics 365 CoE (Figure 21-4). This allows the organization
to bring together specialist knowledge and make it available across
different business, where each individual business or implementation
may not have been able to justify it. This model gives economies of
scale on constrained and specialist resources, and helps disseminate
good practice across implementations. In a CoE model, you may have
permanent internal roles or create sub-roles that have deep specialist
knowledge in business and IT processes. Because the Dynamics 365
CoE can employ its own specialists, it can also help manage other
specialist third-party organizations providing support and service.

Fig.
21-4

Executive
• Provides executive guidance • Clears roadblocks
• Sets priorities • Secondary escalation point
• Authorizes budgets sponsors

• Provides goals and strategy

• Approves expenditures
• Provides day-to-day leadership Program lead
• Primary escalation point

• Day-to-day project management

• Responsible for deliverables and Planning/ Business Technical/ Architecture Change Support
own individual outcomes Release Development management
• Brings operational expertise

A common approach is to outsource some of the Dynamics 365 appli-

cation functional and technical support to a Dynamics 365 partner. This
is especially true if the application has a lot of customizations that were
developed by that partner. In these circumstances, the customer may
still elect to keep a small internal support team of a technical archi-
tect, developer, and release manager to help manage the outsourced
services and maintain final control over the quality of the changes to
the production system.

The size and nature of the support organization should be determined

650
 by analyzing the scope of the requirements for the support organiza-
tion (as discussed earlier in this chapter) as well as the size, complexity,
functional spread, and geographical (time zone, language, local data
regulations) distribution of the implementation.

Standard Dynamics 365 platform and application support and servicing

for production environments is also available (Figure 21-5). Microsoft
offers different levels of support for your Dynamics 365 products. Refer
Fig.
21-5 to Chapter 20, “Service the solution,” for more details.

Support provided by Microsoft

Infrastructure Application platform
• Storage and database capacity management • Diagnostics, patches, and updates
• High availability and disaster recovery • Network management
• Platform security • Application monitoring and
• Infrastructure capacity, scale up and down first-line support

Implementation
• Infrastructure management and deployment
• Datacenter networking, power and cooling
roles and
responsibilities Customer leading, supported by
implementation partner
User/Data Application
• Security, identity configuration, and • Define and test business processes
management
• Develop and test customizations
• Monitor sandbox environments

Budgets, contracts, and agreements

When a new system is introduced into an existing enterprise architec-
ture, some changes can impact related systems, such as the changes
required to budgets and support agreements. A common issue that
can occur is that the budgets and agreements aren’t modified to reflect
the new system environment and processes. This can cause delays
when addressing issues that may span departments, or if the resolving
authorities or third parties have their own budgets based on previous
processes and agreements.

Implementation Guide: Success by Design: Guide 21, Transition to support 651

 After you have a reasonable draft of the new system solution blueprint,
to-be enterprise architecture, and support organization, you should start
defining the budgets related to the new tasks and areas of responsibility.

Consider the full set of tasks and activities in scope for the support or-
ganization (as discussed earlier) and map these to the various roles and
resolving authorities over the full lifecycle of a support job or request
so that no gaps appear in the flow. You can use this to make sure that
the specific responsibilities of all the roles and resolving authorities can
be mapped to agreements and service-level expectations.

For internal parties, this may mean defining budget and resource splits
and less formal service delivery agreements. For third parties, that may
mean formal contracts with formal SLAs. Mismatched expectations
between the customer and a third-party support provider in the mid-
dle of dealing with high-priority issues are not uncommon. A contract
created based on a deeper understanding of the expected tasks and
service expectations is much more likely to avoid misaligned expecta-
tions and provide a faster path through the process.

Support Support operations

operations The preparation and design work allows you to provide the level of
support expected by the business. There are many aspects to consider
when planning for the practicalities of operating the support services.
Transition
Preparation for support is necessary, but not sufficient. The support
Requirements management team also needs to learn and practice to ensure they are ready for the
extra pressures of the period immediately following a go live. The next
Change management
section explores this in more detail.

Hypercare

Transition
Support organizations for business applications are rarely fully formed
on the first day of go live. You normally have a transition period from
project mode to support mode. However, the transition period often
starts later than it ideally should, and the quality of the transition is
often not sufficient to provide the level of service that is expected. This

Implementation Guide: Success by Design: Guide 21, Transition to support 652

 results in poor user experience, frustration from the support team, and
a drag on the project team that is called into support requests long
after the transition is supposed to be complete.

What is a good time to start the transition?

Learning a new business application system, even one that is as user-
friendly as Dynamics 365, can take time, and actively using the system
is one of the best methods. For the business process-related support
roles, if the SMEs are expected to move into the support role at go live,
training happens naturally as part of the project, assuming the project
approach includes a significant amount of hands-on system involvement
by the SMEs. It’s important that the business process experts don’t just
become consultants to drive business requirements and have very little
exposure to how the system design meets the requirements.

If, however, the existing SMEs supporting the legacy system are
expected to support the new system, consider formalizing their in-
volvement in the project during the Implement phase, from the early
build steps and especially in shadowing the business functional leaders
at critical points in design and software playbacks.

Furthermore, consider involving the support business process roles

in the formal testing phases as authors of some of the test cases or in
conducting the testing. Other opportunities to get the team to directly
experience using the new system can come from creating task record-
ings and task guides in Dynamics 365 Finance and Operations apps, or
other forms of training documents.

You can apply a similar approach to the technical roles, they can be
involved in shadowing the project team at critical points and be given
project tasks to complete (under the supervision of the project team).

Consider setting intermediate targets for transition readiness through-

out the project so it doesn’t become a last-minute process. It may
help to exercise the whole support organization prior to go live and to
review the effectiveness of the readiness for supporting the production
system. Consider using the formal support processes during the later
stages of formal testing (such as UAT) to review readiness. This way, a

Implementation Guide: Success by Design: Guide 21, Transition to support 653

 representative sample of issues raised during this test period can be
recorded, triaged, and resolved through the support organization tools
and procedures. The project team will be expected to participate as
resolving authorities to help train the support team and ensure that
issues are addressed in a timely fashion. The learning from this type of
exercise can help make the transition more effective than last-minute
handover training sessions.

If the SMEs aren’t being exposed regularly

and extensively to the system during its Most organizations run lean support teams, which doesn’t offer many
build, a gap will likely appear in their
opportunities for the team to participate in the transition activities we
understanding and ability to support,
which means increased reliance on the described without explicitly planning these activities so that the sup-
partner functional consultants for support
(and testing).
port team gets time away from their day-to-day operational duties to
participate fully during the project implementation.

Transition
Requirements management
Requirements When the application is in production, it doesn’t mean that the sys-
management tem solution can now be considered static for the next few years. The
Change management solution will need to deal with new requirements for several reasons:
▪ Some projects implement a minimum viable product (MVP) on
Hypercare their first go live with the goal to incrementally add the lower-pri-
ority requirements over time
▪ Some projects have planned multiple rollouts that may impact the
already live areas
▪ Some changes to Dynamics 365 are driven by changes in connected
systems in the enterprise
▪ Businesses need to react to the changing world around them
▪ In a modern cloud SaaS world, incremental functional and tech-
nical updates from Microsoft and ISVs help keep the customer’s
solution secure and updated

Some organizations will continue to have a project team that is respon-

sible for reviewing, prioritizing, and delivering the backlog, but many
won’t have a standing project team, or the project team will be purely
focused on the next rollout.

In such cases, the support organization needs to have the resources

and means to keep the solution up to date for the users currently using
the live system.

Implementation Guide: Success by Design: Guide 21, Transition to support 654

 In any case, the support teams are often at the front line of the feed-
back and requests for new requirements. At the very least, they need
a way to capture and channel the new and emerging requirements to
the right parties.

Transition Change management

Requirements management We saw in the previous section that the solution is unlikely to remain
static for long, and the solution functionality can change because of
Change management new and backlog requirements.

Hypercare
However, support teams may need to accommodate other sources of
change in the production system, such as the following:
▪ Core data changes (new customers, suppliers, items as part of
normal business operations)
▪ System parameter changes (such as in response to the growing
volume of data)
▪ Configuration changes (such as in response to regulatory changes)
▪ Approval process changes in the system (such as in response to
internal policy changes)
▪ Bug fixes

Some of these changes can be made within the agreed change control
process enforced and facilitated by the system itself. In other cases,
proposed changes need to be reviewed, prioritized, and approved by
business or IT stakeholders and any other relevant parties.

This process should be considered as part of defining the scope of the

support teams’ responsibilities. It should also prompt the organization
to create the appropriate change management boards for changes to
the system in operational mode.
Transition

Requirements management Hypercare

Most projects define a short period of time after go live in which addi-
Change management
tional resources and extraordinary attention are given to supporting
the business. This hypercare period is necessary to ensure that the
Hypercare
business can operate within acceptable thresholds while addressing

Implementation Guide: Success by Design: Guide 21, Transition to support 655

 any initial growing pains with a new system and new business processes.

The role of the support team during this period is critical, and it’s import-
ant that the expectations are clearly set out and agreed with all parties:
▪ Define a clear exit strategy for the end of the hypercare period
based on meeting explicit criteria. For example, criteria might
include no open P1 issues without an acceptable action plan, key
business processes are operating at an acceptable efficiency rate,
over 90 percent SLA targets being met for P1 and P2 support re-
quests, or the support team can resolve over 95 percent of support
requests without using resources reserved for hypercare only.
▪ Decide on an end date for the hypercare period and have an
action plan in place to meet the exit criteria.
▪ Define expectations and an explicit plan for the project team (and
partner resources) during the hypercare period so you can get reli-
able help from the team. Otherwise, you may find that the project
team is committed to working on the next phase of the project
and will resist being drawn into support issues.
▪ Make sure the support team gets the necessary documentation
and top-up training from the project team and implementation
partner during the hypercare period.

Conclusion
During operational use of the Dynamics 365 system, the support
operations are expected to function efficiently and evolve alongside
the expanding use of the system. For that to happen smoothly, the
preparation and definition of the support operating model are essen-
tial precursors.

Scope of support operations

The sections in this chapter are designed to help define the potential
scope of support operations; you need to convert each of the areas
covered from the “what” to the “what in detail,” “who,” “when,” and “how.”

You should convert the enterprise architecture alongside the business

Implementation Guide: Success by Design: Guide 21, Transition to support 656

 and IT policies and business continuity requirements into practical
reference architecture and detailed procedures with reliably repeatable
steps that the support team can implement day to day.

You should convert the Dynamics 365-specific areas and business

process requirements into the support organization charter or terms
of reference with a further detailed definition of the tasks, the roles
responsible, and the standards expected to keep the system operating
safely and efficiently.

Use the system update cycles section, including managing new re-
quirements and changes, to define the means by which the Dynamics
solution can stay updated and evolve with the additional and im-
proved features. This should deliver a detailed definition of the tasks,
the roles responsible, and the standards expected to continuously
improve the business value from the Dynamics 365 system. This is im-
portant to ensure that the support organization can keep the current
operational system functioning optimally while also working in parallel
on the next update.

Support operating model

The sections discussing the support model and support organization
should help take the definition of the support operations scope and
use the chosen support model to create the shape and size of the
support organization. This resulting design should be a good fit for the
requirements of support operations.

Finally, transition guidelines can help you make sure that the transition
from project to support happens incrementally and through practical
experience of the system during implementation. We also encourage
you to validate the readiness of your support processes and organiza-
tion prior to go live.

Operating considerations
The nature and details of the support tasks are influenced by the spe-
cifics of the business application because they are so embedded in the

Implementation Guide: Success by Design: Guide 21, Transition to support 657

 core business. The details of the support operations are further affect-
ed by the details of the circumstances of the business being supported.
We discuss some of the circumstances next.

Routine maintenance and service operations

Consider defining a support and servicing shared calendar that lists the
regular, periodic tasks. This can help make sure those tasks are reliably
implemented and have the right visibility across the team.

Service hours, language, and geography

In addition to defining the normal hours of operation, consider how to
define out-of-hours service, especially for urgent issues.

Some larger organizations, especially multi-national ones, may need to

provide support in multiple languages and across multiple time zones
based on the geography of the users.

Furthermore, consider business-critical periods when support may

need to provide out-of-hours support; for example, during seasonal
peaks or period-end or one-off events such as company acquisitions.

All of these topics have a bearing on the operational patterns that the
team needs to support.

SLAs, priorities, and escalations

Support organizations need to provide some form of SLA for the
different types of support request, if these terms are based on genuine
business criticality, they will help drive the right prioritization.

Well-defined Establishing a formal escalation procedure with defined criteria will also
help prevent frustration on slow-moving cases and help mitigate the
scope and support impact of those that escalate issues beyond their true business impact.
operations will
yield an efficient Tools and access
Most support organizations have some type of internal helpdesk
support model software that is used to manage support requests across all IT ser-
that can evolve as vices. Most continue using their existing helpdesk for Dynamics 365
needed.
658
 applications but need to consider new internal and external resolving
authorities. Third parties such as partner support organizations and
ISVs have their own helpdesk systems. When the issue is diagnosed as
likely to be in standard Dynamics 365, the recommended method of
registering the issue with Microsoft is to use the in-application process:
Dynamics 365 Finance and Operations apps Support and Dynamics
365 CE Power Platform Support. Tracing a support request across
multiple technologies and especially across third-party organizations
requires some deliberate planning and possible configuration so that
the internal helpdesk has a reference from start to finish of any ticket.

Policies on access for partner resources to the customer systems and

customer data need to be defined as part of the support organization
so that there are no delays in the middle of dealing with high-priority
issues while policies are reviewed, and special approvals are obtained.

For example, if the policy only allows for partners to view anonymized
customer data, having an automated way to copy the production
system that includes anonymization will help reduce delays when
troubleshooting or testing.

Avoid delays by
defining access
policies for partner
resources.

659
 Support operational
Checklist considerations
Establish a transition strategy that drives the support
team to participate in the emerging solution and im-
prove their support readiness over the project lifecycle,

Support scope rather than a limited handover close to go live.

Confirm the readiness of the support team’s operating

Establish a practical reference architecture and detailed
model, resources, knowledge, and procedures during UAT.
procedures with repeatable steps for the support team
that can be operated day to day. Have a defined strategy to manage the critical period
just after going live (hyper care) with explicit entry and
Identify the impact of the business and IT policies on the
exit criteria.
scope and the impact on the policies after the Dynamics
365 application is in operation. Define how to keep the solution up to date and evolving
with additional and improved features and capabilities.
Identify the Dynamics 365 application-specific functions
and procedures as well as business processes that need Determine a support and servicing calendar for regular
to be supported. and periodic maintenance activities.

Identify the normal hours of operation across different

Support model and locations, as well as out-of-hours service windows,

the related support multi-language and time zone requirements, seasonal

organization peaks, period-end, or one-off events.

Establish formal escalation procedures with defined

Define the type of support operating model to adopt
criteria, SLAs, and different tiers of support, including
and the nature of the role of the internal and external
internal and external support.
support teams.
Determine the tooling that the support organization
Create a support organization charter with a detailed
requires.
definition of the tasks, roles, responsibilities, and stan-
dards expected to keep the system operating. Identify the policies to access data and environments,
either by internal support teams or by external support
Consider the implications of supporting a cloud and
teams.
SaaS application, and potentially including a Dynamics
365 Centre of Excellence (CoE) in the support model.

Implementation Guide: Success by Design: Guide 21, Transition to support 660

 Case study
A story in two acts
Over the last few years, IT project implementation methodologies have
been changing to try to keep pace with the speed of innovation of
software in the cloud, especially Dynamics 365 applications. IT organi-
zations within businesses have similarly been evolving as the cloud is
changing the role and areas of responsibility for IT.

The following case study is an illustration of the dangers of assuming

that an existing, well-functioning support model will continue to
perform well in a different architecture and with changed business
processes in new applications, without deliberate planning and action.
This is doubly true in a cloud world.

Act I: The journey of learning

This is a story of a medium-sized organization (fewer than 1,000
employees) with multiple offices across three European countries and
their corporate headquarters in the UK. Their core business includes
professional services, field services, and bulk distribution from their
own warehouses.

The group was undergoing rapid business growth. To support and

accelerate this growth, they chose to replace several specialist local
applications and multiple on-premises ERP and CRM systems with
Dynamics 365 Finance, Dynamics 365 Supply Chain Management, and
Dynamics 365 Customer Engagement applications across the enterprise.

Implementation Guide: Success by Design: Guide 21, Transition to support 661

 They planned their first rollout to locations in two countries that had
significant interdependency (accounting for approximately 15 percent of
their business) and to the other businesses incrementally, as part of the
next phases of the project. Subsequent phases would also include addi-
tional functionality related to Human Resources, Payroll, the Customer
Service Center, and shared services on contract management.

The existing support teams were fragmented across various systems,

and sometimes also by business unit. These multiple support teams
were planned to be organized under a unitary structure in IT before
the transition to Dynamics 365. All support activities and teams were
planned to be managed by a director of support services, and who
would report to the CIO.

The project was planned to go live approximately 10 months after

project initiation. At the exit criteria review of the Initiate phase, they
identified the following:
▪ The to-be end-to-end business process flow diagrams and de-
scription were adequate in most areas.
▪ The solution blueprint document was judged to be sufficiently
well-defined to confirm the scope of the overall design.
▪ Budgets and project resources to complete the project were
revised and approved in light of the latest solution blueprint and
project plan. The project scope and go-live date had been re-
viewed and reconfirmed by the steering group.
▪ In readiness for the implementation of enterprise-wide Dynamics
365 applications, the ambition was for the support team to be
reorganized by main business areas: Finance, Sales and Marketing,
Operations, Field Service, Warehouse and Shipping, IT business ap-
plication, and IT servicing, rather than by legacy system. However,
in reality, the teams continued to be tied to the legacy systems
they had existing expertise in supporting.

There was an informal understanding that the support team would be

trained during the project. However, during the Implement phase, the
project team stated that they were behind schedule and too focused
on project delivery to help educate the support team. The mitigation
plan was for the support team to be part of an extended, one-week

Implementation Guide: Success by Design: Guide 21, Transition to support 662

 training and hand-off period after UAT was complete.

As the implementation progressed, the support organization contin-

ued to focus on the day-to-day needs of the currently operating legacy
systems. There was a general feeling in the project and in the support
organization leadership that the risk was low because the existing
support team was experienced and some of the project’s SMEs would
join the support team, thereby seeding knowledge into the team. The
underlying assumption was that the support team would be able to
pick up the new duties without too much fuss, given that they had
been working in a support role for a few years.

During the Prepare phase, when some members of the support team
were asked to assist with the testing, the team didn’t feel ready to par-
ticipate until they had been through the training, which was scheduled
after the testing. The project SMEs didn’t feel they could conduct the
internal training because they also had very little hands-on experience
in the working system.

When UAT was complete, the project team had other unplanned
activities that were considered to be higher priority, and so the training
and handover to the support team was reduced to two days.

The go live went ahead on schedule, but the support team struggled
to adequately support the business. The initial assumptions that the
support team would just pick up the new system support were found
to be mistaken, because the underlying processes had also undergone
significant changes. The support team hadn’t been informed about
all the business process decisions and weren’t familiar with the new
control and approval requirements configured in the Dynamics 365
application. The shortened training and handover were conducted
without reference to the significantly changed business processes or to
the new enterprise architecture, so didn’t provide the necessary context
to support a production system.

The support operating model had been revised, in theory, from the
distributed, system-level team structure to be more business process
and enterprise IT-based, but it wasn’t exercised during the Implement

Implementation Guide: Success by Design: Guide 21, Transition to support 663

 or Prepare phase; the gaps and deficiencies were only exposed once
the teams had to operate in production. The new operating model had
also not adequately considered the implications of changed require-
ments of the new business processes, architecture, and policies.

In the previous support operating model, all queries and issues came
directly (and with little structure) to an individual support team member,
and the full lifecycle of the issue was performed with little external visi-
bility. In the enterprise-level support, even with the first limited rollout,
the support team was receiving large numbers of administrative queries,
which reduced time to address the more business-critical tickets.

The team was also forced to revise IT servicing procedures to reflect

the new technology implications in a cloud-based SaaS model. They
had to revise policies related to data retention and backups because
processes based on the legacy, bespoke on-premises processes weren’t
reviewed and updated for the new cloud-based SaaS architecture.

The project team had planned to start working full-time on the next
phase of the project after only two weeks of hypercare. The budget
hours for the implementation partner hypercare support were planned
to last four to six weeks, but were exhausted within two weeks or so
because the support team needed far more of their time. Many SLAs
were missed in the early weeks, and the business was forced to formally
extend the hypercare period for both the project team and the imple-
mentation partner because trying to get the timely attention of the
implementation teams was difficult given that they had other planned
priorities for the next project phase.

After the P1 and P2 support tickets were resolved and the volume of
new tickets was down to a manageable level, the post-go live
retrospective review made the following key recommendations to be
enacted as soon as possible, to improve the support outcomes during
the next rollout phase:
▪ The scope of the activities, tasks, and actions expected from the
support team should be explicitly defined, in the context of the
new system, processes, people, and architecture.
▪ The support operating model needs to have a more formal

Implementation Guide: Success by Design: Guide 21, Transition to support 664

 definition of the different level of support and a clearer agreement
on the expectations from third parties. For example, using tools
such as Microsoft Teams channels where super users can help
address the higher-volume, simpler questions quickly. The super
users would also help to better formulate issues and questions that
needed to be addressed by the next support level.
▪ The support team needs to be formally allocated time to be in-
volved in the implementation project to ensure their learning is
sufficiently deep and is embedded in the new process and archi-
tectural context to be usable in production. The transition should
not rely solely on a last-minute, half-hearted handover.
▪ The transition should be formally planned within the project plan
with specific tasks (such as process design reviews and test case
preparation) so the team has sustained and multiple opportunities
to have hands-on experience of the developing system.

Act II: Fewer problems = shorter story

These recommendations were approved by the stakeholders and
enacted over the following months. Even though the next rollout (six
months later) had a much larger user base, it was a much smoother
operation with a shorter hypercare period, higher user satisfaction, and
markedly less cost than the first go live.

This story exemplifies the adage “Poor service is always more expensive
than good service.”

Implementation Guide: Success by Design: Guide 21, Transition to support 665

 Conclusion

Implementation Guide: Success by Design: Conclusion 666

 Conclusion
As we have seen in numerous examples
throughout this book, decisions taken during
the project implementation lifecycle can have
dramatic impact on overall project success.
For example, the case study in Chapter 14, “Testing strategy,” describes
the negative business impact resulting from deprioritizing perfor-
mance testing. What I love about this example is that it describes a
situation we see all too often. Perhaps it’s logical to think that perfor-
mance testing is optional because the first deployment is a limited go
live, but as the case study shows, it’s important to prioritize perfor-
mance testing before Phase 2 when more users are onboarded onto
the system.

Every decision taken during a project implementation ultimately comes

down to risk; the decision is either unknowingly adding risk or helping
intentionally manage risks during delivery. What we have attempted
to do with the Success by Design framework is provide you with a very
clear set of guidelines and practices that will help you manage risk and
ensure that you have approaches to mitigate all risks that show up as
part of a solution implementation. These practices and the prescriptive
approaches are based on years of experience gained by our team of over
200 Dynamics 365 solution architects, having overseen thousands of
complex implementations over the last many years.

Implementation Guide: Success by Design: Conclusion 667

At Microsoft, our passion for building the most innovative products
is rivaled only by our commitment to helping our customers and
partners bring these products to life so we can jointly realize the op-
portunities presented by digital transformation. The Success by Design
approach is a framework, and like all frameworks, it’s far from perfect.
We invite you to take this guidance and apply it to your projects—by
making it your own and adding to it what we may have missed. In
the spirit of jointly building the body of knowledge that drives our
Implementation Guide, we invite you to engage with this content and
let us know how it compares to your experiences in implementing
Dynamics 365 solutions.

Your feedback is critical to helping us evolve this guidance,

so please don’t hesitate to reach out. You can email us directly at
[email protected].

We’re immensely proud of the body of knowledge that our FastTrack

for Dynamics 365 team has formalized through Success by Design, and
even more excited to share this knowledge in the first iteration of this
book. What you’ve just read is only the beginning—we will ensure that
this guidance continues to reflect product truth and includes the latest
best practices gleaned from the thousands more implementations that
we will run in the coming years.

While we may never fully eliminate the nail-biting anticipation that

accompanies most go lives, I hope that our unique take on the fun-
damentals of successful implementations will bring more certainty,
predictability, and stability to your digital transformation journey.
Thank you for partnering with us on your journey. I look forward to
hearing from you as you put this guidance to practice.

Faisal Mohamood
General Manager, FastTrack for Dynamics 365

668
 Acknowledgments
This book celebrates Microsoft’s desire to share the collective thinking
and experience of our FastTrack for Dynamics 365 team, which is
currently made up of more than 140 solution architects who serve
Microsoft Dynamics 365 customers and partners around the world.

For being the kind of leader who purposely carves out time in your
busy schedule to think about, write about, and share your vision for
Dynamics 365 and the Microsoft Power Platform, thank you, James
Phillips. We hope that the way you’ve inspired us comes through on
every page of this book.

For your tireless efforts in making Dynamics 365 a better product and
for making us better solution architects, thank you, Muhammad Alam.
We’re also grateful for the time you gave us to round out the content
in Chapter 1.

For encouraging us to embrace James Phillips’s vision of making

Dynamics 365 Success by Design for all, and for supporting us in
bringing this book to life, we thank Faisal Mohamood, Swamy Narayana,
and Jason Kim. Successful leaders prefer details over executive
summaries—the reason we know this is because of you. Thank you.

No one who Thank you to the book’s authors and reviewers.You kept your promise
to write the truth into each chapter, and to make sure the message
achieves success they contain is valuable to our readers.
does so without
acknowledging To the entire Steyer Content team for your professional and highly

the help of others. competent effort in advancing the book’s content and design. Thank you.

The wise and Finally, a special thanks and debt of appreciation is due to all our
the confident current and former Dynamics 365 customers and partners from whom
acknowledge this we learn and by whom we’re inspired with each encounter. We marvel
at what you do.
help with gratitude.
-Alfred North Whitehead —The FastTrack for Dynamics 365 Team

669
 Appendix

Implementation Guide: Success by Design: Appendix 670

Phase
Initiate Regardless of which implementation methodology
you use, here are the typical implementation
deliverables, activities, and recommended roles that
need to be performed by the project team during
the Initiate phase of the project.

This table is for reference only; a missing deliverable,

activity/task, or role might not be an indication that
something is missing in your project.

Note that the names of deliverables, details

regarding activities and tasks, and roles and job titles
might change depending on your project.

When a deliverable or activity is application-specific,

it is specified as either:
▪ F&O for Finance and Operations apps; or
▪ CE for Customer Engagements apps

Implementation Guide: Success by Design: Appendix 671

Topic Deliverables Activities/Tasks People

Program strategy Program strategy/ Define: Customer project

charter ▪ Business vision sponsor, program
▪ Business goals/business director
drivers Project managers
▪ Business objectives
Customer business
▪ Identify key success
stakeholders
metrics or KPIs
▪ Project budget Customer IT director
▪ High-level project and Customer IT architect
solution scope Solution architect
Project plan Define the project plan, Functional consultant/
scoping all activities across architect
project phases. Change manager
Risk and issue Define the risk and issue Project management
register register (with risks, severity, office (PMO) team
mitigation, consequence, Resource manager
issues, escalation process).

Project Define the governance

governance controls through
establishment of a PMO
and steering committee to
manage the triple constraints
(schedule, budget, resources),
project risks, and change
management/change control
for the project.

Solution Define architecture review

architecture board (ARB) or design
management authority (DA), which governs
key architectural decisions
made during implementations.

Implementation Guide: Success by Design: Appendix: Initiate phase 672

Topic Deliverables Activities/Tasks People

Program strategy Project Install applications and tools Customer project

continued technology that will be used to deliver the sponsor, program
and tools project. These could be: director
▪ Lifecycle Services (for F&O Project managers
projects)
Customer business
▪ Azure DevOps
stakeholders
▪ SharePoint
▪ Microsoft Teams Customer IT director
▪ Conduct implementation Customer IT architect
team training on Solution architect
methodology and tools.
Functional consultant/
Familiarization Conduct initial familiarization architect
training training on the Dynamics Change manager
product(s) that will be
Project management
implemented, as well as
office (PMO) team
independent software vendors
(ISVs), add-ons, and every Resource manager
other software element.

Train the trainer Plan train-the-trainer strategy, Functional consultant

which includes content Technical writers
preparation for training;
Trainer SMEs
identifying the customer
trainers; number of training Customer trainers
sessions; plan of execution of Customer business
the training. consultant

End-user training Plan end-user training, which Customer IT team

includes number of users,
number of training sessions,
and the plan for executing the
training.

In-context help Plan in-context help

documentation document/manuals that
Dynamics 365 application will
provide for users.

Implementation Guide: Success by Design: Appendix: Initiate phase 673

Topic Deliverables Activities/Tasks People

Test strategy Test plan Plan the testing strategy. Test lead
Test consultant
Define test activities such as
unit testing, functional, user Functional consultant
experience (UX), integration, Solution architect
system integration testing Customer IT team
(SIT), performance, and time
Customer business
and motion study.
consultants

Provide high-level definitions UX consultant

of test cases, scenarios, Performance test lead
alternate scenarios, and Performance test
expected results. consultant

Test execution/ Establish KPIs for measuring

metrics test execution outcomes.

Performance/ Plan performance benchmark

scalability goals testing to assess the scalability,
concurrency loads, network
configurations (latencies,
bandwidths), and browser mix.

Usability Elicit accessibility,

compatibility, and ease-of-use
requirements.

Testing Confirm testing automation

automation tool tool that will be used for
testing, including regression
testing.

Business process Process Define process architecture, a Solution architect

strategy architecture commonly understood, shared Functional consultant
view of all business processes
Customer business
that an organization performs
consultant(s) (SMEs)
to deliver a product or service
to their customers. Customer IT architect

Implementation Guide: Success by Design: Appendix: Initiate phase 674

Topic Deliverables Activities/Tasks People

Business process Requirements Define requirements Solution architect

strategy traceability traceability matrix (RTM), a Functional consultant
continued management key deliverable and success
Customer business
measure. Every requirement
consultant(s) (SMEs)
needs to be linked to at
least one process. It can also Customer IT architect
depict the to-be processes as
requirements in the form of
user stories/use cases. The RTM
also includes nonfunctional
and negative requirements.

Fit-gap Assess which requirement

assessment and respective process can be
addressed by the system or
requires some customization.

Gaps evaluation Evaluate and estimate gaps with

and estimation the minimum information of:
▪ Related process
▪ Related integration
▪ Criticality
▪ Complexity
▪ Dependencies

Organizational Define organizational

structure structure based on feedback
from executives and senior
managers from functional
areas. Model how the business
operates. It might consist of:
▪ Geographical structure
▪ Organizational groupings
▪ Line of business
▪ Reporting lines
▪ Segregation of duties

Implementation Guide: Success by Design: Appendix: Initiate phase 675

Topic Deliverables Activities/Tasks People

Business process Country/region Validate country/region

strategy localization localization availability for
continued
availability each country/region in scope.

Application Solution Defines high-level view of Solution architect

strategy architecture the solution architecture, Technical lead(s)
blueprint which depicts how Dynamics
Customer IT architect
365 apps interact with other
systems and users. UX consultant
Customer business
ISV/add-ons Define need for considering
consultant(s)
any ISV solutions to meet the
business requirements.
Include all considerations in
the relevant strategies.

User experience Prototype high-level

(UX) wireframes from UX point of
view to derive the new user
experience.

Data strategy Data migration Define data that needs to be Customer IT architect
plan and migrated to Dynamics 365 Solution architect
execution apps, along with data volumes,
Functional consultant
strategy approach, and tools to be used.
Data migration
Define high-level plan of Lead/architect
data migration activities,
such as data extraction,
transformation, and load;
migration runs; initial and
incremental migrations;
data validations and data
verification; and production
migration.

Implementation Guide: Success by Design: Appendix: Initiate phase 676

Topic Deliverables Activities/Tasks People

Data strategy Data management Outline strategy, including: Customer IT architect

continued strategy ▪ Scope Solution architect
▪ Key requirements
Functional consultant
Approach, including:
Data migration
▪ Governance
▪ Compliance Lead/architect
▪ Migration
▪ Master data
▪ Configuration
▪ Sharing and promotion
Technology and tools

Configuration Define plan and approach for

management how to manage:
(F&O) ▪ Configurations
▪ Shared data
▪ “Golden” configuration,
including promotion
▪ “Golden” configuration
environment

Integration System Elicit the integration needs, Customer IT architect

strategy integration such as: Solution architect
requirements ▪ Identifying lines of
Functional consultant
business (LOBs) to be
integrated. Integration lead
▪ Inbound and outbound
integration requirements
from Dynamics 365 apps
▪ Integration types such as
Sync, Async, Batch
▪ ACID (atomicity,
consistency, integrity, and
durability) requirements
towards integration calls

Implementation Guide: Success by Design: Appendix: Initiate phase 677

Topic Deliverables Activities/Tasks People

Integration Integration Outline the strategy, including: Customer IT architect

strategy strategy ▪ Scope Solution architect
continued ▪ Integration approach
Functional consultant
▪ Landscape
▪ Technology and tools Integration lead

Integration Define the integration

interfaces mapping between Dynamics
identification 365 and other systems. Plan
the dependencies.

Dual-write Consider using the dual-write

infrastructure infrastructure if customer has
both Dynamics 365 CE apps
and F&O apps.

Intelligence BI and reporting Outline the strategy, including: Customer business

strategy strategy ▪ Scope consultant
▪ Reporting structure Reporting lead
▪ Key requirements
Functional consultant
▪ Roles and responsibilities
▪ Approach Solution architect
▪ Technology and tools AI SME

Printing strategy Outline the strategy, including: Customer IT architect

▪ Scope
▪ Key requirements
▪ Approach
▪ Technology and tools

Personal, ad-hoc Elicit the reports for personal

reports or ad-hoc needs and data
volumes

Operational Elicit the reports for managers

reports at various levels (e.g., team,
business group)

Implementation Guide: Success by Design: Appendix: Initiate phase 678

Topic Deliverables Activities/Tasks People

Intelligence Organizational Elicit the reports for broader Customer business

strategy and executive scope, consolidated data consultant
continued reporting across the enterprise systems, Reporting lead
and KPIs for executives (CxOs).
Functional consultant
AI/ML needs Elicit the requirements that Solution architect
are fulfilled through artificial
AI SME
intelligence, such as customer
churn analysis, and customer Customer IT architect
lifetime value.

EDW (enterprise Define the high-level view of

data warehouse) Dynamics 365 data integration
to customer’s EDW.

Common data Define the high-level

model (CDM) and considerations to use and
Azure Data Lake structure data using CDM.
Storage

Security strategy Security strategy Outline the strategy, including: Customer information
▪ Scope security group (ISG)
▪ Key requirements team
▪ Approach Solution architect
▪ Technology and tools
Identity SME
Federation, SSO Define the needs for single Customer
sign-on and federation with infrastructure SME
other identity providers.
Customer IT architect
Security roles Define all the roles required
and assign these roles to
business process activities so as
to map them to business roles/
personas.

Identity Elicit the requirements for

integration integration with other identity
providers.

Implementation Guide: Success by Design: Appendix: Initiate phase 679

Topic Deliverables Activities/Tasks People

Security strategy Azure Active Define the requirements for Customer information
continued Directory Active Directory or Azure security group (ISG)
(AAD) access Active Directory, identity team
management integration needs, and cloud Solution architect
identities.
Identity SME
Information Elicit security, privacy, and Customer
security compliance requirements; infrastructure SME
regulatory needs; and audit
Customer IT architect
events.

Mobile security Define BYOD requirements,

MDM, MAM, and/or
conditional access policies.

Compliance Define the privacy, regulatory,

management compliance, and security
procedures.

Run risk assessment.

ALM strategy ALM Define the required DevOps consultant

environments environments for ALM setup Customer IT architect
(e.g., dev, test, golden, pre-
Solution architect
prod, prod).
Technical lead
Solution Define the solution
Customer IT team
management segmentation strategy

Configuration Define the source control and

management branching strategy.

Build process Define the processes for build

management and automation.

Deployment Define the processes for

process deployment management
and automation between the
environments.

Implementation Guide: Success by Design: Appendix: Initiate phase 680

Topic Deliverables Activities/Tasks People

ALM strategy Release process Define the processes for DevOps consultant
continued
release management of the Customer IT architect
solution.
Solution architect
ISV ALM Define the process for Technical lead
managing the DevOps process
Customer IT team
for other components and
solutions (such as Azure/non-
Azure components) that make
up the solution.

Environment and Tenant, Define tenants/environment/ Customer IT architect

capacity strategy environment and instance needs and the Solution architect
geo strategy locations.
Technical lead(s)
Environment Define required environments Customer support
planning to be set up. team

Deploy required environments Technical lead(s)

to enable the team to start PMO
working. Customer information
Platform release Define the plan, maintenance Security group team
process and schedules for the platform Customer
updates release and upgrades. infrastructure team

Regulation and Define the regulatory and

compliance compliance checks that
must be adhered for the
implementation.

Service-level Plan the post-go-live

operations operational procedures, such
as application availability,
backup, archive, audit,
maintenance registers,
supportability, and service
updates.

Implementation Guide: Success by Design: Appendix: Initiate phase 681

Topic Deliverables Activities/Tasks People

Environment and API limits Define the capacity-planning Customer IT architect

capacity strategy requirements for integrations Solution architect
continued and user data access.
Technical lead(s)
Storage Define the capacity planning Customer support
requirements for storage. team
Deployment Plan the deployment strategy, Technical lead(s)
planning locations, required hardware PMO
and software prerequisites, etc.
Customer information
Rollout planning Create the high-level plan for Security group team
staggered rollouts. Customer
Continuance plan Create the continuous update infrastructure team
architecture and update policy.

Regression Confirm the regression

automation tool automation tool that will be
used.

Implementation Guide: Success by Design: Appendix: Initiate phase 682

Phase
Implement Regardless of which implementation methodology
you use, here are the typical implementation
deliverables, activities, and recommended roles that
need to be performed by the project team during
the Implement phase of the project.

This table is for reference only; a missing deliverable,

activity/task, or role might not be an indication that
something is missing in your project.

Note that the names of deliverables, details

regarding activities and tasks, and roles and job titles
might change depending on your project.

When a deliverable or activity is application-specific,

it is specified as either:
▪ F&O for Finance and Operations apps; or
▪ CE for Customer Engagements apps

Implementation Guide: Success by Design: Appendix 683

Topic Deliverables Activities/Tasks People

Program strategy Project scope and Following all related activities, Customer PMO
requirements list the project scope and RTM to Customer business
signoff be signed off on. stakeholders
Cutover strategy Complete this deliverable by Project manager
and plan the start of the build. Solution architect
Solution design Once design is documented, Functional consultant/
signoff it needs to be signed off on. architect
All changes that are required
follow review, approval, and
version control.

Train the trainer Prepare the training materials Functional consultant

for customer training teams, Tech writers
such as functional admins,
system admins, and identified
functional users.

Test strategy Develop unit test Build unit test cases for the Test lead
cases defined functionality to ensure Test consultant(s)
TDD (test-driven development)
DevOps consultant
practices are adhered to.
Technical consultant(s)
These unit test cases should Performance test
be executed as part of CI/CD consultant
pipelines.

Build verification Conform to DevOps practices

tests (BVT) and build the BVT scripts that
decide on the validity of the
build for accepting to the test
environment.

Execute test Execute test cases manually or

cases/test automatically.
scenarios

Implementation Guide: Success by Design: Appendix: Initiate phase 684

Topic Deliverables Activities/Tasks People

Test strategy Automation of Build automation scripts for Test lead

continued test cases high-impact processes. Test consultant(s)
DevOps consultant
It is recommended to build
automation scripts for Technical consultant(s)
processes that are critical and Performance test
prone to changes. consultant

Performance Develop and unit test

testing performance test scripts and
execute the scripts, per the
load requirements.

Business process Financial structure Finalize financial structure Solution architect

strategy design with the components: Functional consultant
chart of accounts, dimensions,
Finance SMEs
and account structure.

Test cases/ Write test cases/scenarios Functional consultant

scenarios supporting the functionalities Test consultants
to be tested for various
identified test types, such as
functional testing and SIT.

Design the automation testing

needs for performance testing,
UI testing, and more.

Business Define business processes

processes and (linear and cross-functional
personas process flows) showing the
inputs, outputs, and activities
at a subprocess and process,
including intended personas.

Gaps solution Design solution, update process

design flow, and confirm approach.
Create development plan.

Implementation Guide: Success by Design: Appendix: Initiate phase 685

Topic Deliverables Activities/Tasks People

Business process Low-level For all gaps identified in Functional consultant

strategy technical design the fit-gap assessment, it Test consultants
continued (CE) is recommended to have
low-level technical design
documentation (TDD) in place.

Focus on how the

requirements are achieved
using the Dynamics 365
platform constructs. This must
be detailed to the core.
▪ Define customizations,
such as business rules,
plug-ins, JavaScript,
workflows, and Power
Automate workflows.
▪ Depict a technical flow
on how a functional
requirement is fulfilled
using the combination
of client-side scripts and
server-side rules.

Application User experience Design the visualizations, such Functional consultant

strategy (UX) design as UI form definitions, UI views, UX consultant
UX (e.g., forms, and custom controls.
views) design

Solution design: Consider the design and Technical lead(s)

API limits/service- development impacts to API Solution architect
protection limits limits/service protection limits.
Customer IT architect
planning
Technical consultant(s)
Technical lead
Functional consultant
Technical consultant
continued on next page

Implementation Guide: Success by Design: Appendix: Initiate phase 686

Topic Deliverables Activities/Tasks People

Application ISV/add-ons Include the installation of Functional consultant

strategy the ISV/add-ons early in the Technical lead(s)
continued solution design so that they
Data migration lead
are incorporated into the
processes, data, security, and Technical consultant(s)
integration design. Customer IT team

Solution design: Note that based on expected Customer data

storage capacity data migration activities and migration consultant(s)
capacity planning of the
solution, the design should
include storage capacity needs.

Build solution Develop the solution per the

development design documents.
code

Develop UX Build UX constructs (e.g.,

forms, views, controls, web
resources)

Data strategy Data model Confirm data model is

precise and complete. A clear
definition of the following
aspects of the data model
should be available as part of
defining the data model:
▪ Data types
▪ Data format
▪ Data length
▪ Rules (e.g., auditable,
change tracking)

Data migration Often, it is possible to write

design custom tools or leverage
ISV tools to execute data
migration needs. Whatever the
approach taken, it is important
to design: continued on next page

Implementation Guide: Success by Design: Appendix: Initiate phase 687

Topic Deliverables Activities/Tasks People

Data strategy Data migration ▪ ETL (extract-transform- Functional consultant

continued design (continued) load) processes. Technical lead(s)
▪ Data dictionary: data
Data migration lead
migration mapping
between incumbent and Technical consultant(s)
Dynamics 365 system. Customer IT team
▪ Data migration scripts Customer data
design. migration consultant(s)
▪ Execution report.
For projects with heavy data
migration requirements, it is
recommended to maintain a
separate design document for
Data Migration activities.

Configuration Create data packages and/

management or golden configuration
(F&O) environment.

Configuration Store approved configuration

management and establish:
(F&O) ▪ Change management
procedures.
▪ Transfer process to and
from repository to the
various development and
test environments.
▪ Plan to move to
production.

Historical data Confirm requirements, scope,

migration and approach.

Evaluate migration in
Dynamics vs. data warehouse
storage.

Implementation Guide: Success by Design: Appendix: Initiate phase 688

Topic Deliverables Activities/Tasks People

Integration Integration design It is recommended to maintain Integration lead

strategy a separate integration design Technical consultants
document if the integration
Azure architect/SMEs
needs are heavy. This should
(if required)
cover:
▪ Integration patterns to be Solution architect
considered
▪ Detailed integration
interfaces and mapping of
data attributes between
the systems.
▪ Transformation rules
▪ Integration response
processing
▪ Integration type (e.g., sync,
async, batch).
▪ Integration technologies
(e.g., logic apps, Power
Automate, third-party
ESBs)

Dual-write If dual-write is used, harmonize

infrastructure concepts such as company,
business unit, security,
etc., between CE and F&O.
Determine the plan to architect
the data properly between
apps.

Intelligence BI design If the project has a high BI lead

strategy number of reporting Technical consultants
requirements and the project
Functional consultant
needs custom reporting
beyond what Dynamics 365
offers out of the box, it is
recommended to build a
detailed design document
covering the following:
Implementation Guide: Success by Design: Appendix: Initiate phase 689
Topic Deliverables Activities/Tasks People

Intelligence ▪ Sources for data beyond

strategy Dynamics 365 apps.
continued ▪ Visualizations using
reporting tools such as
Power BI or SSRS.
▪ Reporting queries.
▪ Logic to make up the
reporting interfaces.

Security strategy Security design Security implementation Technical lead

is one of the critical Solution architect
pillars in Dynamics 365
Functional consultant
implementations. Design
should be thorough and Customer IT architect
requires extensive analysis Customer information
of the current requirements security group (ISG)
before deciding on the
security constructs suiting the
Dynamics 365 apps. Example:
For Dynamics 365 CE apps
or Power Platform, design
of the following is of utmost
importance:
▪ Business units
▪ Security roles and access
levels
▪ Ownership, access, and
AAD teams
▪ Field-level security
▪ Custom security options,
if any

Implementation Guide: Success by Design: Appendix: Initiate phase 690

Topic Deliverables Activities/Tasks People

Security strategy Security roles When a custom role needs to Technical lead
continued (F&O) be created, a good practice Solution architect
is to not modify the standard
Functional consultant
roles because it will impact
the continuous updates from Customer IT architect
Microsoft. Customer information
security group (ISG)
Apply strict change
management control.

Include security roles in the

golden configuration.

Segregation of Once the security roles are

duties (SoD) assigned to process activities,
identify SoD rules and baseline
the security testing.

Compliance Track, assign, and verify your

management organization’s regulatory
compliance activities related to
Microsoft Professional Services
and Microsoft cloud services,
such as Office 365, Dynamics
365, and Azure.

ALM strategy Develop CI/CD For continuous integration DevOps consultant

pipelines and continuous development Technical consultant(s)
needs, it is recommended to
build the routines for DevOps.

Environment and Solution design: Design considerations for Technical lead(s)

capacity strategy performance high-level customizations Technical consultant(s)
from the viewpoint of system
performance, such as UI form
loads, error handling, query
performance, solution health
checks, and more.

Implementation Guide: Success by Design: Appendix: Initiate phase 691

Topic Deliverables Activities/Tasks People

Environment and Regression Build regression automation Technical lead(s)

capacity strategy automation test tool and test scripts to be able Technical consultant(s)
continued to frequently uptake Microsoft
service releases.

Implementation Guide: Success by Design: Appendix: Initiate phase 692

Phase
Prepare Regardless of which implementation methodology
you use, here are the typical implementation
deliverables, activities, and recommended roles that
need to be performed by the project team during
the Prepare phase of the project.

This table is for reference only; a missing deliverable,

activity/task, or role might not be an indication that
something is missing in your project.

Note that the names of deliverables, details

regarding activities and tasks, and roles and job titles
might change depending on your project.

When a deliverable or activity is application-specific,

it is specified as either:
▪ F&O for Finance and Operations apps; or
▪ CE for Customer Engagements apps

Implementation Guide: Success by Design: Appendix 693

Topic Deliverables Activities/Tasks People

Program strategy Signoff The customer/partner has Customer information

documents gone through all internal security group/team
approvals, information security Customer IT architect
reviews, penetration testing,
Solution architect
and other steps to ensure
production readiness of the Functional consultant
system. PMO

Go-live readiness The cutover plan, go-live date, Customer project

document/ and go/no-go criteria are sponsor(s)
cutover checklists agreed upon with customer
business stakeholders and
the production deployment
runbook is created with all
tasks, owners, start time,
duration, and rollback plan
determined.

Train the trainer Complete training of the Functional consultant

customer trainers for Technical writers
functional and administration
Customer training
activities.
admins
End-user training Prepare and deliver the end- Customer business
manuals user training for effective use users
of Dynamics 365 Business
Apps.

Online help Revise online help guides

guides based on feedback from UAT,
if any.

Testing strategy UAT test cases Refine UAT test cases used for Customer business
validating the solution. users (UAT test users)

UAT test report Prepare and present UAT test Test lead
execution report that baselines Functional consultant
the UAT signoffs. Test consultant(s)

Implementation Guide: Success by Design: Appendix: Initiate phase 694

Topic Deliverables Activities/Tasks People

Testing strategy Performance These results help establish the Performance test lead
continued benchmark confidence to go live for the Solution architect
testing results anticipated concurrent loads,
Customer IT architect
and transactions.

UAT signoff Once the tests are completed, Program director

business signs off on user Test lead
acceptance testing.

Business Mock cutover Conduct mock cutover and Tech lead

processes strategy approval validate and approve outputs. Technical consultant(s)
Sign off.
Solution architect
Customer IT architect

Data strategy Data migration Team completes the last Tech lead
execution results/ cutover migration of the Technical consultant(s)
signoff data required following the
Solution architect
shutdown of the incumbent
systems before opening Customer IT architect
Dynamics 365 for go-live.

ALM strategy Escrow Escrow all the project artifacts Tech lead
collected during the Initiate, Technical consultant(s)
Implement, and Prepare
Solution architect
phases.
Customer IT architect
Knowledge Register the documents and
transition maintain all pre- and post-
documents deployment activities that the
support team should consider.

Maintenance Maintainability workshops are

register/ set of trainings/documents
maintainability that help the support
workshops team learn the operational
procedures of the production
solution.

Implementation Guide: Success by Design: Appendix: Initiate phase 695

Topic Deliverables Activities/Tasks People

Environment and Service-level Confirm and understand the Tech lead

capacity strategy agreement (SLA) SLA. Technical consultant(s)
Production Release and deploy Solution architect
environment production. Customer IT architect

Implementation Guide: Success by Design: Appendix: Initiate phase 696

Phase
Operate Regardless of which implementation methodology
you use, here are the typical implementation
deliverables, activities, and recommended roles that
need to be performed by the project team during
the Operate phase of the project.

This table is for reference only; a missing deliverable,

activity/task, or role might not be an indication that
something is missing in your project.

Note that the names of deliverables, details

regarding activities or tasks, and the roles or role
titles in the People column might change depending
on your project.

When a deliverable or activity is application-specific,

it is specified as either:
▪ F&O for Finance and Operations apps; or
▪ CE for Customer Engagements apps

Implementation Guide: Success by Design: Appendix 697

Topic Deliverables Activities/Tasks People

Project Post-go-live review Customer support

postmortem team
Customer IT architect
Production Operations Refine the operations guides/ PMO
operations guides/checklists checklists that help maintain
Solution architect
(maintenance) the production solution for
updates, changes, and solution Functional consultant
health.

Platform release Operations Update the cloud solution. Customer Support

process and guides/checklists Test customizations supporting Team
updates the new updates. Customer IT architect
PMO
Solution architect
Functional consultant
Technical consultant(s)
Test consultant(s)

Maintenance Operations This is a very important Customer support

guides/checklists practice to ensure necessary team
maintenance schedules are Customer IT architect
defined to undertake activities
PMO
such as:
▪ Platform updates.
▪ Custom solution updates.
▪ Batch data updates.
▪ Archival activities.

Periodical execution of archival

and data retention.

Implementation Guide: Success by Design: Appendix: Initiate phase 698

Topic Deliverables Activities/Tasks People

Change Change request Documents that explain the Customer support

management Change impact change request (e.g., hotfix, team
assessment feature enhancement) to be Customer IT architect
Change control brought into the production
PMO
solution and the impact
assessment of the existing
components, as well as the
plan to implement and deploy
the change.

User adoption/ User adoption Document the solution usage Customer business
usage document (user patterns and user interviews, users
interviews, along with strategies to Customer IT architect
feedback loops) improve the adoption of the
PMO
system.

Implementation Guide: Success by Design: Appendix: Initiate phase 699