G00261728

Create a Centralized and Decentralized

Organizational Model for Business Intelligence
Published: 4 November 2014

Analyst(s): Kurt Schlegel, Owen Milbury, Frank Buytendijk, Dan Sommer

Self-service business intelligence and analytics requires a centralized team

working in collaboration with a finite number of decentralized teams. IT
leaders should create a two-tier organizational model where the business
intelligence competency center collaborates with decentralized teams.

Key Challenges
■ Overly centralized business intelligence (BI) and analytics teams can't deliver the domain
expertise and responsiveness most organizations require. While the centralized team does a
good job in creating consistency and governance across certain key subject areas (such as
sales and finance) it creates a bottleneck where most users are waiting too long to get their
requirements met.
■ Overly decentralized BI and analytics teams have the opposite problem. This model delivers
plenty of domain expertise, agility and responsiveness, but struggles to deliver consistency
across its information sources and analytic models. In addition, they struggle to share best
practices.

Recommendations
■ Create a two-tiered organizational model with a single centralized team working collaboratively
with a collection of decentralized teams distributed throughout the enterprise.
■ Create a finite list of the decentralized teams throughout your organization requiring some
autonomy for how they integrate, analyze, and report data. These teams can be organized by
horizontal function such as sales, HR, finance, line of business function or geographic region.
■ BI and analytic leaders should strike the balance of power between centralized and
decentralized teams that is suitably based on the top-down versus bottom-up culture of the
organization.

Table of Contents
 Introduction............................................................................................................................................ 2
Analysis.................................................................................................................................................. 3
Create a Two-Tiered Organizational Model....................................................................................... 3
Create a Finite List of the Decentralized Teams Throughout Your Organization in Need of Autonomy
.........................................................................................................................................................5
BI and Analytic Leaders Should Strike the Balance of Power Between Centralized and Decentralized
Teams.............................................................................................................................................. 5
Top Down.................................................................................................................................. 5
Divide and Conquer.................................................................................................................... 6
Bottom Up..................................................................................................................................6
Gartner Recommended Reading............................................................................................................ 7

List of Figures

Figure 1. The Future of Analytics Is a Balanced Centralized and Decentralized Organizational Model for
BI........................................................................................................................................................... 4

Introduction
IT leaders often work as part of a centralized BI and analytics unit that integrates, reports and
analyzes data across the enterprise. However, this centralized team does not always represent the
complete picture.

While there may be a strong centralized framework and architecture, decentralized analysis is also
occurring throughout the company at various levels. Business units often have more specific and
more in-depth needs for analytics and reporting. They use their own analytical tools, and if this is
unsupported by the BI center of excellence, they do it themselves as part of a so-called "shadow
IT" (or unofficial) process.

Successful companies have to navigate the complications that arise when data analysis is occurring
both through a centralized function and within decentralized silos throughout the organization. The
future of BI and analytics is about enabling both a centralized BI function as well as the
decentralized analysis occurring within the company. Create a two-tier organizational model in
which the business intelligence competency center (BICC) collaborates with decentralized teams.
Allow for these two layers to form an agile equilibrium between bottom-up and top-down
approaches.

Gartner's research over the last decade has promoted the concept of a competency center or
center of excellence for BI and analytics. While this concept is still very much a recommended
approach, we no longer believe it is sufficient. The BI/analytics center of excellence or competency
center needs to collaborate with a network of decentralized BI and analytic teams.

Page 2 of 8 Gartner, Inc. | G00261728

 Gartner recommends that the central tier of this two-tiered organizational model maps directly to
the BICC or center of excellence. Gartner also provides three strategies — top down, divide and
conquer, and bottom up — for organizations to use as the foundation of their BI and analytics
strategies — which are not mutually exclusive. Most companies will need to blend all three
approaches.

Analysis
Create a Two-Tiered Organizational Model
Many organizations have traditionally based their BI and analytics models around a centralized, IT-
dominated BI team. These companies are emphasizing governance and consistency across the
enterprise.

However, the reality is that in most organizations, data analysis is also occurring in various
departments throughout the company at different levels, for example, analysis within horizontal
units such as sales, services, marketing, HR, financial or other departments. Line-of-business units
or geographies may also be engaged in their own analysis. This results in a lot of information being
analyzed, integrated and applied to decisions within organizations that is not a part of a centralized
BI and analytics team.

Create a single, centralized team working collaboratively with a collection of decentralized teams
distributed throughout the enterprise. The first step is to understand this framework and recognizing
how to craft a successful BI and analytics strategy; see "Gartner's Business Analytics Framework"
for a detailed description.

For most organizations, the reality resembles Figure 1.

Gartner, Inc. | G00261728 Page 3 of 8

 Figure 1. The Future of Analytics Is a Balanced Centralized and Decentralized Organizational Model for BI

Source: Gartner (November 2014)

Page 4 of 8 Gartner, Inc. | G00261728

 Create a Finite List of the Decentralized Teams Throughout Your Organization in
Need of Autonomy
Draw your company's analytics map to discover:

■ Where analytics is occurring within your organization.

■ How many decentralized analytics units are in operation, and what they look like.
■ Whether individuals are undertaking a lot of work on their own with packaged analytical apps.
■ Whether your decentralized teams have many skill sets, with the know-how to integrate and
model data well.
■ The number of data scientists working in your teams.
■ The number of individuals actively involved in analytics.

Cloud-based applications such as Workday for HR or Omniture for clickstream analysis, offer easy-
to-consume analytical capabilities directly to decentralized teams. This work is often carried out
exclusively in Excel, but in reality, individuals in decentralized teams have created effective,
mission-critical applications that fulfill the equivalent function of many more officially defined IT
roles, such as data integration and modeling.

To be clear, Gartner is not suggesting that the numerous Excel-based applications in your company
are eliminated, rather, that an inventory is created to acknowledge that decentralized BI and
analytics teams participate as part of your company's overall BI and analytic efforts. The key point
is to foster collaboration between the centralized and decentralized teams that previously worked as
independent silos. IT leaders need to move from "gatekeepers" to "air-traffic controllers."

BI and Analytic Leaders Should Strike the Balance of Power Between Centralized
and Decentralized Teams
Each organization should develop a BI and analytics strategy that strikes a balance between a
centralized and decentralized model, as there is no "one size fits all" model. There is virtually no
limit to the different permutations for this model, and companies should craft a strategy specific to
their organizational structure and culture. We suggest a blend of three strategies: top-down, divide
and conquer, and bottom-up. The focus may be slightly different for each strategy, but all three
represent a blend of centralization and decentralization — and are not mutually exclusive.
Organizations should blend all three ideas into their strategies.

Top Down
This approach is more toward the centralized end of the spectrum. The centralized BI and analytics
team decides how to report, integrate and analyze data across the entire enterprise. The data is
integrated and modeling occurs centrally, but individuals are embedded locally throughout the
organization to bring report writing, training, support, Q&A, and other capabilities closer to the
decentralized workgroups. This approach is architected from the top down.

Gartner, Inc. | G00261728 Page 5 of 8

 One variant of this top-down approach is to use a franchise model, which is a globally consistent
way to bring a brand to a local market. For example, the centralized team could set up analytic
franchises in various departments, the equivalent of "local markets."

The central BI team creates a global process to integrate, report and analyze data, implementing it
consistently at a local level (in a similar way to a restaurant or retail franchise), with variations for
different regional locations. For example, one department or geography might calculate a local
measure slightly differently to the global measure. BI and analytic leaders need to find just the right
permutation of these ideas for a top-down franchise and apply it to the company.

Divide and Conquer

In the divide-and-conquer approach, the organization identifies which subject areas need to be
globally consistent across the enterprise. A centralized team usually assumes responsibility for
these areas. Information may come from ERP, CRM or payroll systems. Regardless of the source of
the information, the centralized team ensures global consistency.

However, outside of the domain of the centralized team, other subject areas exist that are the
responsibility of the decentralized teams. Think of a government model when envisioning how a
divide-and-conquer approach may work for your organization. Decide the specific work of the
global and local teams. This strategy provides decentralized teams with autonomy over the choice
of tools purchased and how solutions are architected within their domains.

Ultimately, the global team retains authority when it has identified a particular subject area as
important for the whole enterprise. Jurisdiction is the key to this divide and conquer idea. The
organizational model needs to contain clear guidance on which groups (centralized or decentralized)
have the authority to define measures.

Bottom Up
In the bottom-up approach, everything starts locally. In every local department or workgroup, there
is the empowerment to buy whatever tools are necessary, architect solutions to fit local needs, and
the authority for data modeling, reporting and analysis.

There is still a centralized team in overall charge that finds and then promotes interesting analysis
across the entire organization. Data discovery is about empowering individuals and small
workgroups to integrate, store and present data in a single proprietary tool.

The market continues to head in the direction of the bottom-up model. Local teams are being
empowered to create and innovate. The centralized team identifies the most successful work being
done at a local level, and provides a platform to share and promote this work globally. Not only will
decentralized teams following the bottom-up model have more autonomy to pick their own tools,
but they will also employ more data integration and modeling skills.

Page 6 of 8 Gartner, Inc. | G00261728

 Gartner Recommended Reading
Some documents may not be available as part of your current Gartner subscription.

"Gartner's Business Analytics Framework"

"How to Deliver Self-Service Business Intelligence"

"Deliver Business Intelligence With a 'Think Global Act Local' Organizational Model"

Gartner, Inc. | G00261728 Page 7 of 8

 GARTNER HEADQUARTERS

Corporate Headquarters
56 Top Gallant Road
Stamford, CT 06902-7700
USA
+1 203 964 0096

Regional Headquarters
AUSTRALIA
BRAZIL
JAPAN
UNITED KINGDOM

For a complete list of worldwide locations,

visit http://www.gartner.com/technology/about.jsp

© 2014 Gartner, Inc. and/or its affiliates. All rights reserved. Gartner is a registered trademark of Gartner, Inc. or its affiliates. This
publication may not be reproduced or distributed in any form without Gartner’s prior written permission. If you are authorized to access
this publication, your use of it is subject to the Usage Guidelines for Gartner Services posted on gartner.com. 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 and shall have no liability for errors, omissions or inadequacies in such information. This
publication consists of the opinions of Gartner’s research organization and should not be construed as statements of fact. The opinions
expressed herein are subject to change without notice. Although Gartner research may include a discussion of related legal issues,
Gartner does not provide legal advice or services and its research should not be construed or used as such. Gartner is a public company,
and its shareholders may include firms and funds that have financial interests in entities covered in Gartner research. Gartner’s Board of
Directors may include senior managers of these firms or funds. Gartner research is produced independently by its research organization
without input or influence from these firms, funds or their managers. For further information on the independence and integrity of Gartner
research, see “Guiding Principles on Independence and Objectivity.”

Page 8 of 8 Gartner, Inc. | G00261728