Anda di halaman 1dari 15

SAP White Paper

Business Intelligence

Managing the Total Cost of Ownership

of Business Intelligence
A 360-Degree Perspective

3 About the Author ^11 Understanding the Cost Drivers of

Business Intelligence
^ 4 Executive Summary 11 Business Needs
11 Analytics Tools
^ 5 What Is My Business Intelligence 12 People
Total Cost of Ownership? 12 Hardware
5 Business Need Drives IT Spend 12 Data
12 Governance
^ 6 How Should Enterprise
Business Intelligence Run? ^13 Best Practices for
Managing TCO
^ 8 Data Explosion and Strategic
Information Management ^14 Summary
9 Capitalizing on the Data Explosion 14 For More Information
with Top-Down Development

^10 Controlling the Cost of

Business Intelligence
10 Providing Appropriate
Analytical Tools

About the Author

Dr. Wayne Applebaum is a principal business consultant with over 25 years of

experience in business intelligence and business analytics strategy. He develops
analytic frameworks that allow clients to take action by merging the key elements
of business analytics, data design, and technology into a cohesive whole.

Dr. Applebaum takes a systemic view of enterprise information management and

works closely with each client to ensure a complete solution, combining the
talents of both the client and SAP to create a unique perspective.

He holds a doctorate in research methodology from the University of Pittsburgh.

Executive Summary
If You Don’t Meet Your Needs,
Cost Doesn’t Matter

A major question that companies are User expectations are growing. BI users How does an organization position itself
asking business intelligence (BI) pro­ expect all information to be available all to deliver a cost-effective BI solution
viders is, “What is your total cost of the time and in a format that can be that can meet its user community’s
ownership?” easily understood and easily accessed. thirst for information and offer the
In addition, the data must be of high lowest TCO? How does it manage the
For many companies, total cost of quality, transparent, and relevant to TCO of BI? For IT professionals, this is
ownership (TCO) is out of control. their jobs. a problem that keeps us awake at night.
And the problem is growing, fueled by
ever-increasing demands from the user These demands are not that unreason­ The purpose of this paper is to focus
community, massive new sources for able considering the culture we live in on developing a framework for monitor­
data, new capabilities, shadow IT land­ today, with information coming at us ing and managing the TCO of BI.
scapes, and the cost of keeping people from the time we wake in the morning
abreast of all the changes. to until the end of the day.

For many companies, total cost of ownership is out

of control. And the problem is growing, fueled by
ever-increasing demands from the user community,
massive new sources for data, new capabilities,
shadow IT landscapes, and the cost of keeping
people abreast of all the changes.

4 SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence
What Is My Business Intelligence Total
Cost of Ownership?
Answering the Right Question

A number of metrics have been pro­ that the cost of business intelligence, as archiving data, that can increase
posed for determining total cost of like any other part of business opera­ system performance signif­i­cantly while
owner­s­hip of business intelligence, such tions, be controlled by assessing and saving hundreds of gigabytes of storage
as cost per user and cost per report. understanding the business need and at the data ware­house level, to say
While calculating the TCO of BI – or for letting that drive the priority of spend. nothing of the back­ups, change logs,
that matter any IT solution – may be and duplicate copies.
interesting, the real issue is more about Business need changes and shifts over
the management of TCO. To manage time and so does the need for BI within Another example is the use of the right
BI TCO, just like managing any organi­ an enterprise. New needs arise, while tool for the right job. Most business
za­tional expenditure, we have to look others disappear as does the need intelligence enterprises put the burden
at business need. for certain types of data. One of the of developing analytics on the IT depart­
quickest ways to decrease the cost of ment or power users within the busi­
Business Need Drives IT Spend BI within an enterprise is to thoroughly
examine each area and determine what
The TCO of BI is tightly coupled to portions of it are no longer required to A key driver for managing
business need. It is business need that support business needs. These could
drives IT spend in this area. In general, include obsolete reports, tools, and data.
the total cost of business
the maturity of a company’s BI capa­ In some cases, there may even be the intelligence is a company’s
bilities has a large impact on TCO. opportunity to simplify solution sets,
Business intelligence tools are scalable standardize the data model, and rethink
ability to effectively incor­
with respect to capacity, so the cost as the user community delivery model. porate business needs into
load or users increase is fairly propor­
tional. Once a BI effort reaches critical For example, if a data warehouse is five a flexible yet sustainable
mass in a company, the increase in years old, and if we assume that most solution.
cost per headcount, per query, or per reports are run only over this year’s
report remains relatively constant. and last year’s data, then only 40% of
Therefore, a key driver for managing the data in the data warehouse may ness units. A tool that uses in-memory
the total cost of business intelligence is actually be required within the produc­ analytics and an intuitive interface
a company’s ability to effectively incor­ tion system. Some of the data, like that lets users explore and analyze large
porate business need into a flexible yet used in reports on shipped items or data sets and save the view for future
sustainable solution. This is not to say closed sales orders, might have a shorter refer­ence. This could eliminate the
we are advocating limiting business useful life. By examining the useful life need for hundreds of reports to be
need to control costs, but we advocate of the data, we can take actions, such generated. 

SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence 5
How Should Enterprise Business
Intelligence Run?
Managing TCO for BI Requires
a Systemic View

In an ideal world, we would have a target data store under the guidance of A spreadsheet culture has
struc­­ture for enterprise business intelli­ extraction, transformation, and loading
gence that is similar to the structure in (ETL) activities; governance; and docu­ evolved over time to bridge
Figure 1. mented data processes. The focus is
on providing quality data and a single
the gap between the busi­
Enterprise business intelligence is version of the truth. This data store ness need and what BI
framed by business processes and feeds carefully constructed measure­
business decisions, which dictate the ment frameworks and analytics, which pro­vides. The result is an
requirements for information technol­ are then combined to provide accurate, increase in the actual TCO
ogy. Information flows from transaction business-relevant information that
systems and other data sources to the helps support business decisions. of BI that is rarely measured.

Business decisions

Measurement Analytics Data

Target data store

Extraction, transformation, Governance Processes


Transactional systems and other data sources

Business processes

Figure 1: The Structure of Enterprise Business Intelligence

6 SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence
Unfortunately, BI has historically started Even if we normalize this landscape, often require a substantial amount of
as a departmental effort. In this case, a deeper problem may not be solved staff time to maintain. A spreadsheet
there was an over-emphasis on achiev­ as easily. BI often is used to provide culture has evolved over time to bridge
ing quick wins but little emphasis on information that feeds a spreadsheet. the gap between the business need
how to organize the efforts to form a Spreadsheet-driven processes have a and what BI provides. The result is an
single structure. This has resulted in high cost, inhibit information sharing, increase in the actual TCO of BI that is
enterprise BI structures more similar are subject to higher error rates, and rarely measured.
to Figure 2.
Business decision Business decision

When viewed from the enterprise level,

the business intelligence structure
Measurement Analytics Data Measurement Analytics Data

appears as a series of disjointed over­ target data store target data store
Business decision

lapping silos. Often different software extraction,

transformation, loading governance Processes extraction,
transformation, loading governance Processes
Measurement Analytics Data
tools support this type of BI structure. Transactional systems and other data sources Transactional systems and other data sources

This is neither a simple nor a sustain­ target data store

Business processes Business processes

able landscape. This type of structure Business decision
transformation, loading governance Processes

leads to multiple versions of the truth, Transactional systems and other data sources
Measurement Analytics Data
requiring reconciliation of reports and Business decision Business processes
duplicate data sets. The cost of main­ target data store

taining this kind of structure is higher

Measurement Analytics Data
extraction, governance Processes
transformation, loading

than necessary. To make matters worse, Transactional systems and other data sources
target data store

Business decision
as more business needs arise and extraction,
transformation, loading governance Processes
Business processes
are met, complexity and duplication Transactional systems and other data sources
Measurement Analytics Data

con­tinue to grow. target data store

Business decision Business processes

extraction, governance Processes
transformation, loading
Measurement Analytics Data
Business decision
Transactional systems and other data sources
target data store
Measurement Analytics Data
Business processes
extraction, governance Processes
transformation, loading target data store

Transactional systems and other data sources extraction,

transformation, loading governance Processes

Business processes
Transactional systems and other data sources

Business processes

Figure 2: Typical Business Intelligence Structure

SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence 7
Data Explosion and Strategic
Information Management
Taking Advantage of Data to Become an
Analytic Competitor

Creating additional urgency to managing Not long ago, data was scarce. Data hoard data, and we still rely on experts,
TCO is the increased burden created was coveted because it was difficult to whether IT or power users, to convert
by the data explosion depicted in obtain and difficult to capture in digital data to information. This creates IT cost
Figure 3. form, and tools for analysis were in the as well as multilayered and siloed struc­
hands of a few experts. While the cir­ tures, increasing the cost of BI and
The digital universe (defined as all con­ cumstances have changed drastically, preventing organizations from realizing
tent stored in digital form) will expand the sociology of data has not. We still their true potential.
by a factor of 44 between 2009 and
2020.1 This growth creates both an
opportun­ity and a challenge. The oppor­
tunity is to glean value out of this data
and improve business processes. The
chal­lenge is that this explosion taxes
current business intelligence practices

The digital universe (defined

as all content stored in digital
form) will expand by a factor
of 44 between 2009 and

by creat­ing more demand for already

scarce resources. With this greater
demand comes increased cost. The
alternative, not leveraging this newly
available data, leaves an organization at
risk of losing a competitive advantage. Figure 3: The Data Explosion

1. IDC white paper sponsored by EMC, The Digital Universe Decade – Are You Ready?, May 2010. The most recent version of the study can be found at

8 SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence
Capitalizing on the Data Explosion One of the success factors critical to One of the first activities of a CoE is
with Top-Down Development managing BI TCO is establishing a BI conducting an inventory of what exists
center of excellence (CoE) to provide in the BI landscape for the entire com­
Understanding business needs is critical a top-down focus. This requires com­ pany. Focusing on and understanding
to the entire process. But how should mitted senior executive sponsorship, the cost drivers, described in a follow­
business needs be constructed and ing section, provides a solid framework
normalized? Many BI efforts start with for conducting such an inventory.
and perpetuate a bottom-up approach. Simplifying the tools in the
A bottom-up approach has the adverse IT landscape is a power- Another activity is understanding and
effect of creating disjointed, overlapp­ing aligning the inventoried items to the busi­
siloed structures and multiple data ­ful cost driver, not only in ness. How does the entire BI infra­struc­
sets. It fails to capitalize on the cross- ture compare to best-practice principles?
functional use of data. This results in
license and maintenance While this is a time-consuming process,
duplicated efforts to collect, store, and costs but in training, usabil­ it is necessary if the organization is going
analyze data while increasing TCO. to normalize practices and streamline
ity, and staff­ing. It avoids costs. For cross-functional areas, large
Implementing a top-down approach the expense of training savings can be made by leveraging
may be easier for an organization just information and analysis, eliminating
embarking on its BI processes than for
people on different analysis the need for the duplication of data
an organization that has a substantial tools when they change stores, metadata, and analytics.
embedded base. For example, consider  
a company that has over 20 installations organizations.
of different BI tools. There are multiple
definitions of items such as revenue
and income. There are also multiple especi­ally in a mature BI organization.
data stores that support thousands of In an organization with multiple data defi­
reports. In short, without appropriate nitions, data stores, and tools, senior
coordination the company’s BI efforts leadership is necessary to resolve differ­
evolved into a very expensive and ences and prevent a stalemate between
inefficient system. different constituencies.

SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence 9
Controlling the Cost of Business
Optimized Data Model and Appropriate
Analytical Tools

One of the keys to controlling the cost busi­ness value. Part of being respon­ Additionally, you need to be careful not
of BI involves examining the data archi­ sive to business needs may be accom­ to confuse understanding of tools with
tecture to determine what is necessary plished quickly by providing the appro­ understanding of data. Both are neces­
and what is not. Tracing how many times priate analytical tools to the right people. sary. It can take users longer to under­
a piece of information is duplicated and This can trigger a process of discovery stand the definitions of data they are
stored is a powerful method to under­ that is highly relevant to bridging the gap analyzing than to learn about the tools
stand the complexity of the data model between IT and the rest of the business. that they use to analyze it. Training
and provides the opportunity to opti­ people on both the data and analytics
mize the model without destroying the Dashboards have gained great popular­ tools is necessary. It’s not just about
ability to provide effective information ity over the last few years. They are a the availability of data; it’s about under­
to users. powerful tool that allows an individual standing business relevancy as well.
to easily explore a limited information
Another area of focus is understanding domain. Dashboards are particularly Gaining a competitive advantage in
data’s intrinsic value in answering busi­ use­ful when you have a good handle on today’s accelerated data explosion is a
ness questions. Not all data has the the guided analysis that the user requires. huge opportunity. Companies must effi­
same value, and a thoughtful business Guided analysis relies on anticipating ciently manage the influx of data while
understanding allows organizations to users’ questions and providing them with also controlling costs. This requires
effectively prioritize their burgeoning an easy way of accessing the infor­ma­ taking a holistic view of bringing data
list of requests for information. tion, whether through the use of drill- and an analytic architecture together
down techniques, dynamic links, or with business needs and training. In this
Providing Appropriate Analytical parameterized reports. way you can manage and reduce the
Tools TCO of BI while still fulfilling increas­ingly
Creating a dashboard assumes you have complex business needs. There is even
A typical complaint about using data enough knowledge of the analytical a possibility of being able to reduce
analysis to solve business problems is process to anticipate the next question. TCO while expanding functionality.
that by the time the data is collected, Sometimes the questions are not easily
cleaned, and analyzed, the deadline for anticipated, or the number of possible
making the decision has passed. This views is extremely large. In these cases,
issue can be overcome by creating an different tools, such as in-memory
analytical architecture that is respon­sive analytics, might provide the user with
to business needs and is driven by the ability to explore data more freely
and easily.

10 SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence
Understanding the Cost Drivers of
Business Intelligence
A Complete View of the BI Cost Structure

Figure 4 provides an overview of the Business Needs ness. This type of CoE acts more like
cost drivers for business intelligence. an information exchange rather than
A key function of a BI center of excel­ like a governing body. It provides trans­
Each of the cost drivers in Figure 4 lence is the alignment of projects to parency into and accessibility of the BI
provides a different view of the BI cost business needs. There are a number of activities of different groups so that
structure. While each of these views ways to structure a BI CoE to coordi­ efforts can be leveraged throughout
has a different focus, it is important to nate and prioritize business needs. the company. Different combinations of
keep in mind that they are interdepen­ At one extreme, the CoE acts as the these structures should be examined
dent and must be balanced. Policies coordinator of all BI projects. This type against the backdrop of your own orga­
and efforts in one area affect how of organization has a mandate that nizational and change management
things work in another. Business needs nothing happens without the CoE’s seal culture to effectively control the
permeate all of the areas. of approval. At the other extreme is a business-need element of the cost
CoE that seeks only to foster aware­ framework.

Analytics Tools

Managing the strategy for using analyt­

Analytic tools
ics tools is also critical to controlling
cost. As a rule, an organization should
strive to maintain the minimum number
People of tools that will satisfy business needs.
Business needs Developers and Simplifying the tools in the IT landscape
is a powerful cost driver, not only in
license and maintenance costs but in
Total cost of training, usability, and staffing. It avoids
ownership having to train people on different analy­­
sis tools when they change organiza­
tions. It also provides for a large core
Data process
of people who understand the tools
Storage and
portfolio processing and are available to mentor newcomers.
The key is to understand and match
tools and techniques with business
Processes, data quality,
needs. What are the consequences of
and quality tools eliminating or adding a tool? Does the
current tool set allow for the flexibility
and hands-on analysis demanded by
Figure 4: The Cost Drivers of Business Intelligence today’s users?

SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence 11
People of providing users with the necessary As the investment in hardware grows,
information to take action. the cost to execute on BI projects also
People costs fall into two categories: grows because of additional hardware
developers of BI reports and analytics If the work product is complete, a user and an increasingly complex environ­
and consumers of information. In some should be able to obtain the information ment. By taking a more thoughtful
cases, developers and users are the without any additional processing of approach to data rationalization and
same people, but in many cases they the data. However, in many cases, it is near-line storage options, you can free
are not. necessary to employ an additional tool up system resources to increase per­
such as Microsoft Excel. This leads to for­mance and give more focus to data
According to a survey conducted by questions about the effectiveness of and information quality.
SAP and the Americas’ SAP Users’ the work product itself, since the end
Group of more than 230 companies, user must perform additional steps or Data
users of BI and reports outnumber use a second product. The calculation
developers of reports by 5 to 1 (per of BI TCO should take into account this For the purpose of this paper, we define
1000 employees in a company):2 additional work time. the cost of data as the cost of the pro­
• Percentage of people that create BI   cesses it takes to manage data and
reports: 6% Hardware make it available in a usable form. This
• Percentage of people that consume includes the cost of the tools, pro­cess­es,
the information in BI reports: 29% The cost of hardware includes storage, and people that go into creating a data
processing, and networking. While warehouse and into developing and
This finding raises the following ques­ hardware cost has decreased per unit, maintaining tools for ETL and data quality.
tions. Is this ratio too high or is it not the amount of hardware required keeps
high enough? What should be the cor­ increasing. Many companies find them­ Governance
rect target for the ratio of creators to selves buying additional hardware on a
consumers of BI reports and analytics? quarterly or semiannual basis to keep Governance completes the circle.
The answer to these questions depends up with data growth. While some of this These are the rules of operation for
upon the nature of the analytics that added hardware is leveraged to the enterprise business intelligence. They
are being performed, the extent of user organization’s bottom line, there is no are very much a part of the BI center
self-services that are involved, and how doubt some waste. Storing unneeded of excellence and align with an orga­ni­
effective the work products are in deliv­ data in production systems, running zation’s business needs.
ering analytics. One way to measure and maintaining analytics that are not
effectiveness is how well the work used, and maintaining underutilized  
products match users’ needs in terms data sources add to the cost.

2. The BI survey was launched in January 2007 and continues to be actively administered through the ASUG and SAP benchmarking program Web sites.
Respondents to this survey are from many different industries and geographies and range from small businesses to large enterprises.

12 SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence
Best Practices for Managing TCO
Learning from Industry Leaders

In 2005 a study conducted by SAP in • To manage critical processes and nition of business value to prevent inter­
collaboration with ASUG studied best limit dependency on external resources departmental arguments over resources.
practices for managing TCO.3 It included after an IT initiative, make it a priority
30 companies with incomes ranging to leverage, build, and retain in-house The study points out the importance of
from US$100 million to more than talent. maintaining an internal core of people
$100 billion and the number of employ­ • To minimize TCO, develop centers and limiting the dependency on external
ees ranging from 400 to 150,000. The of excellence (shared services resources after the initial initiative. The
companies represented a variety of operations). key here is to avoid expensive external
industries. While this study looked at • Simplify the business intelligence resources to do continuing jobs beyond
TCO in general, its conclusions appear land­scape and standardize on the initial bubble. It also points out the
to apply to TCO for BI. Although the common tools. importance of maintaining and retaining
study was conducted almost five years a group of well-trained individuals.
ago, these findings are still extremely The study states that business align­ment
relevant today. is critical to success. It becomes even We have already spoken about the
more important when companies have importance of establishing a center of
The study cites four primary best prac­ either geographically or legally disparate excellence. The study concludes that
tices that industry leaders in TCO have entities. In these cases a strong central­ companies with a center of excellence
in common: ized governance structure is a critical had a 47% lower TCO per active employ­
• To achieve the optimal balance factor to manage TCO. ee than companies that lacked one.
between IT cost and business benefit,
focus on the business impact of every The balance between IT cost and busi­ Lastly, simplifying the landscape creates
IT solution and ensure that C-level ness benefit means continually involving benefits at all levels in the organization.
decision makers understand the the business in determining the port­ These are summarized in the table
issues that arise throughout the folio of projects. This allows IT to better below, which shows that benefits of
entire lifecycle. allocate scarce resources. Care must be landscape simplification are present in
taken on setting standards for the defi­ many areas across the C-level.

Principal Process Increased Faster Risk IT Portfolio

Benefit Standardization Efficiency Growth Reduction Simplification


Benefits • Global design, • Optimal cost • Global and timely • No disruption of • Simpler IT
local flexibility structure visibility production and architecture
• Acquired back- • Collaboration • Acceleration of delivery • More responsive IT
office operations • Speed of acquisition change initiatives • Compliance with • More funds for
• Replication of • Ability to take Sarbanes-Oxley Act innovation
business practices advantage of • International • Reduction of IT
opportunities compliance cost base

3. Hercules Bothma, Best Practices in Managing the Total Cost of Ownership, ASUG and SAP, June 2005.

SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence 13
Road Map for Continual TCO Improvement

The key to managing TCO is inter­ For More Information

depen­dently managing its six drivers: To achieve the optimal
• Business needs To learn more about the software and
• Analytics tools that can help you manage TCO bal­ance between IT
• People for BI while enhancing the availability cost and business
• Hardware and quality of enterprise BI for better-
• Tools informed decision making, please benefit, focus on the
• Governance contact your SAP Services represen­ business impact of
tative today or go to
For companies that have a mature /usa/services/consulting/bts. every IT solution and
embedded base of business intelli­ ensure that C-level
gence, managing TCO requires a care­
ful examination of how they stand in decision makers
relation to managing these six factors. understand the issues
The gaps between the current state
and best practices can be assessed. that arise throughout
A road map of continual TCO improve­ the entire lifecycle.
ment can be put in place to migrate the
organization to an efficient and effec­
tive environment.

14 SAP White Paper – Managing the Total Cost of Ownership of Business Intelligence
50 101 513 (10/08) Printed in USA.
©2010 SAP AG. All rights reserved.

SAP, R/3, SAP NetWeaver, Duet, PartnerEdge, ByDesign,

SAP BusinessObjects Explorer, and other SAP products and services
mentioned herein as well as their respective logos are trademarks or
registered trademarks of SAP AG in Germany and other countries.

Business Objects and the Business Objects logo, BusinessObjects,

Crystal Reports, Crystal Decisions, Web Intelligence, Xcelsius, and
other Business Objects products and services mentioned herein
as well as their respective logos are trademarks or registered trade­-
marks of Business Objects Software Ltd. in the United States and
in other countries.

All other product and service names mentioned are the trademarks
of their respective companies. Data contained in this document serves
informational purposes only. National product specifications may vary.

These materials are subject to change without notice. These materials

are provided by SAP AG and its affiliated companies (“SAP Group”)
for informational purposes only, without representation or warranty of
any kind, and SAP Group shall not be liable for errors or omissions with
respect to the materials. The only warranties for SAP Group products and
services are those that are set forth in the express warranty statements
accompanying such products and services, if any. Nothing herein should
be construed as constituting an additional warranty. /contactsap