How To Build A Cloud 720787 NDX
How To Build A Cloud 720787 NDX
Cloud migration projects are often complex and involve multiple engagements with
system integrators and/or managed service providers. I&O leaders can use this
guidance to build a realistic budget and timeline for their cloud migration projects.
Overview
Key Findings
■ Different types of cloud migration require different types of cloud migration activities, making it
difficult for I&O leaders to identify the costs that apply to their situation.
■ The scope and timing of many migration costs become clear only after the selection of specific
migration approaches, so I&O leaders often struggle to build reliable estimates before the analysis
phase of their project is complete.
■ Common cloud migration activities lack publicly available reference costs and consulting prices vary
widely, creating a challenge for I&O leaders looking to benchmark their estimates and negotiate cost-
effective contracts.
Recommendations
I&O leaders tasked with building cloud migration cost estimates should:
■ Identify potentially relevant costs by assembling a complete range of migration activities necessary
for the organization’s specific case, including those commonly missed by migration consultants.
■ Align expected costs to project phases by selecting one or more of Gartner’s five migration
approaches (rehost, revise, rearchitect, rebuild or replace).
■ Improve budget accuracy by using Gartner’s cost guidance to build an initial pro forma estimate and
iteratively revising it as the project progresses.
Introduction
More than 70% of companies have now migrated at least some workloads into the public cloud, and
momentum is increasing: more than two-thirds of companies surveyed in 2020 plan to increase their
cloud adoption in the wake of the disruption caused by COVID-19. 1 But among the many challenges
posed by the cloud, companies currently rank cost control as the most frustrating. 2 Without realistic
project estimates and cost forecasts, I&O leaders will struggle to assess the progress and performance
of cloud migration initiatives.
How can I&O leaders develop reliable cloud migration cost estimates? This research will help I&O leaders
scope their cloud migration projects successfully and ensure a more complete tracking of true costs over
time.
Analysis
Assemble a Complete Range of Potential Costs
I&O leaders under pressure to migrate quickly to the public cloud are susceptible to several common cost
mistakes (see Figure 1).
■ The Wrong Team. Selecting a migration partner based on familiarity or low pricing rather than
experience, or giving a migration project to an internal team not ready for it, leading to mistakes and
rework.
■ The Wrong Emphasis. Defaulting too quickly to “lift and shift” migration approaches and deferring the
cost to modernize or replace critical applications, resulting in higher cloud operating costs after
migration.
■ Rushed Application Assessments. Failing to fully assess the data center workloads to be migrated,
resulting in incomplete specification of migration requirements and downstream scope creep.
■ Poor Landing Zone Design. Failing to properly architect and implement the underlying cloud “landing
zone” environments into which workloads are migrated, increasing the costs of security and
compliance.
■ Dependency Bottlenecks. Failing to discover and account for the interdependencies between on-
premises systems being moved, resulting in incorrect grouping and ordering of application migrations,
network performance issues and cascading delays.
■ Hidden Indirect Costs. Omitting the indirect costs of cloud migration, such as organizational
transformation costs and the residual (sunk) costs of vacated data center capacity, that must be
absorbed in other organizational budgets.
I&O leaders planning a cloud migration must determine how much of the total effort associated with
cloud adoption belongs in the migration estimate as opposed to a larger “digital transformation”
business case. Many cloud-related activities — such as strategic planning, architectures for business
resilience in the cloud such as high availability (HA) or disaster recovery (DR), or the costs of
organizational transformation — are necessary to meet the organization’s cloud objectives, but are not a
direct cost of cloud migration.
It is often best to constrain a cloud migration estimate to costs that must be incurred to transform a
target set of workloads from a prior (for example, on-premises-based) state to a “stable operational
state” in the public cloud. This avoids conflating the business case for specific workload
transformations with the business case for other broader organizational goals, or with the ongoing costs
of cloud operations. It is also easier to manage vendor contracts for workload migration, application
modernization and managed cloud services separately from “strategic digital transformation” contracts,
as the ROI for these broader initiatives is measured differently. For more information on transforming the
I&O organization for digital business, see Design the I&O Organization for Digital Transformation Using a
Cloud Service Brokerage.
After constraining their focus to workload transformation only, I&O leaders must ensure that their
estimate reflects a complete set of transformation costs. Table 1 summarizes the range of different
costs that are commonly incurred over the course of a large-scale, multiyear cloud migration initiative.
Source: Gartner
Select the Right Migration Approach and Align Costs to Project Phases
After identifying the range of costs that belong in a complete cloud migration estimate, I&O leaders must
place those costs in the right phases of work during the project. A large-scale cloud migration project can
be a multiyear initiative. I&O leaders must work closely with the finance department to forecast and align
migration and operational funding for cloud applications across budget cycles (see Manage the Impact
of Cloud Applications on Opex and Capex Budgets).
The most critical early phase in a cloud migration project is application assessment. This phase will
determine which cloud migration approach should be used for each application. Under pressure to move
quickly, many I&O leaders prioritize the “lift and shift” approach of moving workloads into the cloud
without modifying them. But moving too quickly through application assessment often means failing to
consider that, for many on-premises applications, the best “move” for the business may not be a move at
all. Rather, it may be to rewrite and rerelease an application in a cloud-native way, or even to replace it
entirely with a SaaS-based cloud alternative.
Broadly, Gartner defines five ways to migrate a workload to the public cloud:
1. Rehost: “Lift and shift” your application from its current physical or virtual environment onto a cloud
platform, making as few changes to the application and its runtime environment as possible.
2. Revise: “Lift, shift and adjust” your application just enough to make it safer, easier and less costly to
manage in the public cloud.
3. Rearchitect: Materially alter or refactor the application toward a cloud-optimized architecture, making
some use of cloud-native capabilities.
4. Rebuild: Optimize for the cloud by rewriting the app from scratch, preserving core business logic and
algorithms, but letting go of legacy code and rebuilding on cloud platforms and services.
5. Replace: Replace an application with a third-party SaaS alternative, configuring or extending the SaaS
environment to meet requirements, and (if necessary) migrating legacy data into the new
environment.
Each migration approach has a different profile in terms of the types of costs and when they occur (see
Figure 2). I&O leaders can increase the accuracy of their migration cost estimates by categorizing the
applications they will migrate according to the migration approach they will use and adapting their
estimates to fit the specific cost profile of each approach.
The rehost and revise approaches often have the lowest migration and organizational transformation
costs because they require the fewest changes to how applications are already deployed and managed
on-premises. However, because these workloads are not adapted to the cloud, they tend to run less
efficiently and require more manual support. This can often result in higher long-term total cost of
ownership (TCO) than other cloud migration options.
In contrast, the rearchitect and rebuild approaches require more upfront investment, usually in activities
such as database replacement, refactoring code into container-based microservices, and the adoption of
agile DevOps methodologies and twelve-factor app design. 3 Organizations willing to make this
investment typically are rewarded with lower operating costs, greater resilience and better long-term ROI.
Many organizations choose to replace some legacy applications with SaaS alternatives. Because
moving to a SaaS application does not require the physical migration of workloads, the immediate
migration costs for SaaS can be lower than other alternatives. I&O leaders should keep in mind, however,
that the cost to migrate user accounts and data from previous systems, along with the cost to integrate
the SaaS application with other enterprise systems, can be significant and must be properly accounted
for. Moreover, moving to SaaS does not eliminate ongoing operational costs: These costs are simply
transferred into the vendor’s software licensing price.
Once all potential costs have been itemized and a rough understanding of how those costs map to
project phases has been laid out, I&O leaders must build an initial pro forma migration budget. This pro
forma budget will not be accurate in its first iteration, but it can form the basis for subsequent analyses
and collaboration with vendors that result in a complete and accurate forecast. If feasible, I&O leaders
should allow their migration vendor to propose an initial cost for all effort, but submit a revised
statement of work and final price (subject to approval) after the application assessment phase is
complete.
As a migration cost estimate may be developed collaboratively by several teams, the estimation
approach and the cost units being used must be well understood by all. Until actual project data is
available, pro forma estimates should be grounded in “generally acceptable” cost ranges and
approaches. Table 2 includes approaches and ranges Gartner commonly sees, based on which
migration approach is being used.
Source: Gartner
I&O leaders should also call out in their estimate any key “indirect” costs tied to cloud migration that are
not otherwise accounted for in other budgets or plans. The most important of these are the costs
associated with transforming their organization to operate effectively in the public cloud, and the
residual (sunk) costs of existing data center resources that become less valuable as workloads
transition into the cloud. These costs are frequently unavoidable, but are just as frequently overlooked or
underestimated.
Transformation costs to watch for include the cost to reskill existing teams, the cost to raise salaries to
match market levels for cloud roles, changes to organizational structure and operating procedures, and
the cost to adopt agile DevOps practices across the IT organization. For organizations intending to
“outsource” their cloud operations, the costs to onboard and establish a managed service provider
(MSP) can be considered a transformation cost.
Common residual (sunk) costs from a cloud migration include losses in productivity due to vacated
facilities and hardware, unused software licenses or unproductive staff. It can also involve the cost to
run duplicate versions of the same system during a migration cutover period. A complete accounting for
residual costs may also include the cost to cancel or restructure outsourcing and colocation contracts,
and the time and effort to decommission source systems and archive or destroy data.
For more on how to estimate the larger, strategic costs of cloud transformation, see Replace
Assumptions With Accurate Estimates Before Migrating to Public Cloud.
■ Creation and refinement of operational runbooks, including automation templates and scripts
■ Expanded system administration duties “above the IaaS level” to maintain databases, application
PaaS environments or the migrated applications themselves
I&O leaders who intend to contract for their cloud migration or cloud operations should be alert to
potential costs that their vendor is placing out of scope. Work with the vendor to ensure all critical
requirements are within the scope of the contract and within a given budget range, even if there is a
mutual agreement to establish final pricing on certain activities at a later point. Requirements that the
vendor cannot or will not address should be faced head on in project planning and not deferred.
Preliminary estimates for these costs should be incorporated into the complete cloud migration cost
estimate, even if workarounds or alternate vendors have not yet been identified.
I&O leaders will be more able to hone and improve their migration cost estimates over time if they
structure their cloud migration initiative in iterative phases. Work with the migration vendor to group
applications into different clusters of workloads that will be migrated or modernized together. The first
clusters to be migrated should be treated as “learning migrations.” They should include a selection of
workloads diverse enough to determine the actual cost of different migration approaches, but limited
enough that mistakes and rework will not derail the full project schedule or costs.
Over time, as each subsequent migration cluster is moved or modernized, costs should be tracked and
the overall estimate should be refined and improved. There is no magic spreadsheet for calculating the
“right” total cost of a cloud migration project in advance. Accurate results require continuous effort
anchored in real data and honest collaboration between parties.
Evidence
1
2020 Gartner Cloud End-User Buying Behavior Survey was conducted to understand how technology
leaders approach buying, renewing and using cloud technology.
The research was conducted online from July through August 2020 among 850 respondents from
midsize and larger (more than $100 million in revenue) organizations in the U.S., Canada, U.K., Germany,
Australia and India. Industries surveyed include energy, financial services, government, healthcare,
insurance, manufacturing, retail and utilities. All organizations were required to currently have cloud
deployed.
Respondents are involved, either as a decision maker or decision advisor, in new purchases, contract
renewals, or contract reviews for one of the following cloud types in the past three years: public cloud
infrastructure (IaaS), public cloud platform (PaaS), public cloud software (SaaS), private cloud
infrastructure, hybrid cloud infrastructure or multicloud infrastructure. Respondents were also required to
work in IT-focused roles, with a small subset of procurement respondents.
The study was developed collaboratively by Gartner analysts and the Primary Research Team.
Disclaimer: Results of this study do not represent global findings or the market as a whole, but reflect
sentiment of the respondents and companies surveyed.
2
2020 Gartner Cloud End-User Buying Behavior Survey.
3
The Twelve-Factor App.
■ AWS DirectConnect rates $2.25 per port hour for 10 Gbps, $0.30 per port hour for 1 Gbps. For Azure,
$3,400/month for the port, and $1.87/hour ($1,350/month) for the internal Azure private network
gateway to route traffic to the customer’s private Azure network. For Google, $2.38/hour for 10 Gbps
plus $0.10/hour for a VLAN attachment — and between $0.14 and $0.04/hour for network address
translation (NAT) gateway (total ~$1,860/month).
■ Data transfer management software examples vary widely in capabilities and cost. Examples range
from Google Storage Transfer Service ( $40/TB), to Cloud FastPath ( $1/GB), to Signiant ( $35/year
for 200TB or $0.20/GB).
■ Personnel effort and cost will also vary depending on whether and which data transfer management
software is being used and how large the migration project is. Typically data transfer is handled by
the same team migrating virtual machines. Gartner often sees migration vendor proposals that do not
break out storage data transfer from other migration costs, but for those that do, vendor pricing
between $10/TB to $30/TB transferred is not uncommon. However, rates will be highly dependent on
the amount of data cleaning, reformatting or restructuring that is required.
Adopt a New I&O Operating Model and Organization Design for Digital Business
2020 Planning Guide Overview: Building Skills for Digital Transformation
© 2021 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. and its
affiliates. This publication may not be reproduced or distributed in any form without Gartner's prior written permission.
It consists of the opinions of Gartner's research organization, which should not be construed as statements of fact.
While the information contained in this publication has been obtained from sources believed to be reliable, Gartner
disclaims all warranties as to the accuracy, completeness or adequacy of such information. Although Gartner research
may address legal and financial issues, Gartner does not provide legal or investment advice and its research should not
be construed or used as such. Your access and use of this publication are governed by Gartner’s Usage Policy. Gartner
prides itself on its reputation for independence and objectivity. Its research is produced independently by its research
organization without input or influence from any third party. For further information, see "Guiding Principles on
Independence and Objectivity."