Description
White Paper Business Intelligence. Optimized decisions quickly
White Paper
Business Intelligence.
Optimized decisions quickly
and transparently.
Contents at a glance.
3 1. Introduction.
4 2. Sourcing information from company data: from data to knowledge.
4 2.1 Problems of an information glut – BI as the new solution?
5 2.2 BI in practice – a simplified example.
6 2.3 Evaluation of the German BI market.
8 3. Possible pitfalls in the realization of BI projects.
8 3.1 Lack of direction in the corporate strategy and lack of a BI vision.
8 3.2 Lack of BI acceptance within the company due to insufficient communication and transparency.
9 3.3 Insufficient alignment of the BI solution to the business processes.
10 3.4 Complex technology & opaque procedural model.
13 4. Conclusion & outlook.
14 5. Glossary.
16 6. Bibliography.
1. Introduction.
Ever shorter market cycles, faster changing customer requirements and growing cost and competitive pressures
continuously intensify the demands placed on companies. Fast, effective and sound strategic decisions are needed
to achieve long-term competitivity. Comprehensive and up-to-date, but above all precisely tailored, information
helps not only to better understand one’s own business and the underlying contexts, it also creates a solid basis
for decision-making. In this context, Business Intelligence has laid the foundations in recent years, allowing deci-
sion-making that is increasingly well-founded. However, explosively burgeoning data quantities and ever faster deci-
sion-making cycles require even better, more comprehensive, sophisticated and up-to-date reporting avenues.
In the early days of electronic data processing, pure data management was the focus. It did not, however, take
long to see that the data gathered could, when structured and compressed, provide valuable information on the
most varied of questions. The growing support of corporate business processes by IT, the global use of Internet
technology and the new methods and techniques for gathering and processing data have virtually revolutionized
the opportunities for management support. Nowadays we can use the Business Intelligence (BI) approach
derived from this to transform collected company and competitor data quickly and transparently into decision-
relevant knowledge. Very specific questions, such as “What is the best-selling product in a certain region?”,
can be answered in a matter of seconds and there are no longer any time lags between creating reports and
their presentation. The data arrives on the manager’s desk in real-time, i.e. the minute it is created.
However, there is no plug & play version of the perfect BI application. Given the high complexity of the data
structures and the multiple relationships as well as the constantly developing requirements, the introduction
of a BI application must not only be well planned it must be very prudently implemented. This white paper
deals with the pitfalls facing a company when planning, introducing and using BI solutions and offers a brief
look into the future of analyzing and preparing company information.
3
“The decisive factor for a manager is to filter from a huge glut of data
the current, correct and above all relevant information, thus enabling
him to make sound decisions.”
!
2. Sourcing information from company data:
from data to knowledge.
2.1 Problems of an information glut – BI as the new solution?
The growth in data quantities which has been exponential in recent years is posing great challenges for many
companies. From 2000 to 2002, for example, more data was generated than in all of human history before that.
In the period from 2003 to 2005 this data quantity quadrupled again. Naturally, the resultant glut of information
brings challenges as well as just advantages for companies. While they have – in contrast to times past – a great
deal of information at their fingertips, it is becoming increasingly difficult to sift out the information that is truly
relevant to them.
This is where Business Intelligence solutions come in. From an enormous quantity of data, these solutions find
the data that is actually relevant, organize and analyze it and then present it for review in the desired format.
Business Intelligence is generally understood here as a process that converts data into information. The objective
is to provide real-time, transparent, reliable and high quality data to offer the decision-makers within a company
a valid base of information. This is done with the aid of analytical concepts and IT systems, which evaluate the data
about the company itself, the employees or the market development. With the findings gained, companies can
derive the action that is required to optimize their customer and supplier relationships, lower costs or minimize risks
and thus improve their added value on a permanent basis.
The data to be evaluated here is based on the existing business processes in a company. Hence the data that op-
timally supports the fulfillment of the relevant individual tasks is always provided, tailored to the user – whether
a manager or employee.
Historically speaking, BI applications represent a further development of Management Information Systems (MIS),
Decision Support Systems (DSS), Executive Information Systems (EIS) and Data Warehouse (DW). If the first infor-
mation systems were designed for pure administration of data, then the data warehouses were already designed
for real-time provision of information to a variety of decision-makers and employees in different departments (Figure 1).
Figure 1: Historic development towards Business Intelligence.
4
“More and more companies have come to realize that data and informa-
tion quality is a value-added factor they have thus far largely neglected.”
!
MIS
Efficient file
processing
Integrated systems
Vision of the automatic
decision generator
BI
Key performance
indicator systems
(Balanced Scorecards)
Analytical applications
Data mining
DW
Integration of multiple
data sources
Interactive queries
/OLAP
Historical data
EIS
Multimed. modelling
Separation of operative
systems
Restriction to top
management
DSS
Statistical algorithms,
“what-if”
Complex, rigid
structures
Database orientation
1960 2000 1990 1980 1970
MIS - Management Information Systems
DSS - Decision Support Systems
EIS - Executive Information Systems
DW - Data Warehouses
BI - Business Intelligence
By focusing on the information that is relevant for decision-making, Business Intelligence shall provide methods
and tools which will enable company and competitor data to be transformed into decision-relevant knowledge.
This yields not only a comprehensible and targeted representation of this information (consistent, comparable
and cross-area), but naturally also promises a wealth of further advantages:
Shorter access time to relevant information thanks to real-time data for a faster response to the market
Minimization of the risk of a bad decision, through established, detailed and targeted market information
Ongoing evaluation of the delivery terms and conditions with resultant optimization options
A comprehensive customer understanding through cross-selling or preference analyses to maximize
the customer lifetime values
Detailed overview of the individual cash flows in a company and thus naturally also the identification
of cost saving potential
In addition to the competitive advantages that can be generated, BI applications also cater to the increasing
regulatory requirements (e.g. Basel II, KonTraG, Sarbanes-Oxley Act) and international accounting standards
(in particular IFRS and US-GAAP). These force companies to develop cleaner data scenarios, allowing them
to shape the information flows transparently even with complex processes.
2.2 BI in practice – a simplified example.
To demonstrate the BI approach, a simple real-world example of an industrial bakery, i.e. a medium-sized branch,
is very effective. You are responsible for a chain of bakeries with locations in five different German states and now
intend to open a new branch in North Rhine Westphalia. First you must find a suitable location. To this end you will
analyze, for example, in what region the ratio of turnover to labor and operating costs is particularly cost-effec-
tive or where the competitive situation might be advantageous for you. Germany boasts the most bread vari-
eties (approx. 1,200) in the world, so you are very interested in discovering the bread varieties primarily demanded
in the region. Of course, further, very different and more in-depth analyses are also interesting. These may include
the following, e.g.:
Product and range (success) analysis:
How has the sale of individual products (e.g. poppy seed rolls) developed over a defined period in the
region in focus?
Turnover analysis:
What was turnover development like for the other branches initially? What setbacks must be planned
for? What level of turnover can be realistically expected in the first 12 months?
Branch success analysis:
What products are offered in the branches, which achieved the largest turnover in one month in the region?
How large is the branch and how many people work there?
With the aid of a BI tool, these and many other specific questions can easily be answered from a structured data-
base. To this end, the data is first imported from various data sources and stored in a multi-dimensional informa-
tion cube, and if required collated anew into smaller data cubes that are optimized from a user viewpoint (so-called
“data marts”). The axes of a data cube are used to categorize the different facets of the information obtained.
The content of such a data cube can include quantities sold, quantities produced and turnover, but also regions,
branches, sellers etc. Thus, depending on the focus of the user (product, turnover, sales), the relevant data is ex-
tracted from the data cube and presented in a specially generated view (see Figure 2).
5
Figure 2: Multidimensional view: Information from data; Source: Orbit.
From a structured database, a BI tool can be used to extract all the relevant information quickly and easily and to display
this information clearly. Thus, it is possible to use the data stocks of a company more effectively as a production factor
using this data cube.
2.3 Evaluation of the German BI market.
The advantages of a BI solution are obvious. Nonetheless, the Business Intelligence approach spans a multitude
of different tools and concepts. Correspondingly, there are numerous suppliers, who all promise a timely and
cost-effective launch of BI solutions. The market for Business Intelligence software and services is thus currently
very unclear. Although all suppliers market using the label “Business Intelligence”, ultimately the individual prod-
ucts all too frequently address entirely different commercial tasks and target groups. This section therefore deals
with the “status quo” on the German BI market and provides a brief outlook at the direction in which this market
will move in the coming years.
At the moment the BI market is in a strong consolidation phase and in 2007 it was particularly defined by large takeovers.
Three acquisitions have influenced the market, in particular, in this year: the takeovers of Hyperion by Oracle, of Cognos
by IBM and not the least of Business Objects by SAP. The aim of the established software companies is to position
themselves well in a future growth market through acquisitions. As a consequence of this development, BI has become
an integral component of future-oriented ERP solutions.
According to estimates of the Experton Group, the BI market in Germany in 2007 was set to reach the dimensions
of a mass market of approx. € 1.5 billion. In 2007, more and more medium-sized businesses invested in the area
of Business Intelligence. In total, this was split approximately 47% on software and 53% on services, whereby
“the ratio of software to services was clearly tipped towards software amongst the SMEs. For large companies,
the services share is significantly higher due to the great integration complexity and the required adjustment to specif-
ic business processes” explains Frank Schmeiler, Research Director of the Experton Group.
In the coming years, the BI market will continue to show above-average annual growth approximating 20% and will
increase by 2010 to represent a market value of more than € 2.5 billion. And as BI projects are becoming ever
more complex and comprehensive, the services market will grow in step with the BI software market. The rather op-
timistic growth forecast of 20% can largely be traced back to the required investment from medium-sized companies,
Raisins
Sesame seed
Poppy seed
Wholemeal
Normal
B
e
r
l
i
n
L
o
w
e
r
S
a
x
o
n
y
N
o
r
t
h
R
h
in
e
W
e
s
t
p
h
a
lia
R
h
i
n
e
l
a
n
d
-
P
a
l
a
t
i
n
a
t
e
H
e
s
s
e
2006
2005
2004
2003
2002
P
r
o
d
u
c
t
Raisins
Sesame seed
Poppy seed
Wholemeal
Normal
2006
2005
2004
2003
2002
B
e
r
l
i
n
L
o
w
e
r
S
a
x
o
n
y
N
o
r
t
h
R
h
in
e
W
e
s
t
p
h
a
lia
R
h
i
n
e
l
a
n
d
-
P
a
l
a
t
i
n
a
t
e
H
e
s
s
e
P
r
o
d
u
c
t
Region
6
Raisins
Sesame seed
Poppy seed
Wholemeal
Normal
B
e
r
l
i
n
L
o
w
e
r
S
a
x
o
n
y
N
o
r
t
h
R
h
in
e
W
e
s
t
p
h
a
lia
R
h
i
n
e
l
a
n
d
-
P
a
l
a
t
i
n
a
t
e
H
e
s
s
e
2006
2005
2004
2003
2002
Region
P
r
o
d
u
c
t
Region
Focus: Product Focus: Sales Focus: Turnover
Product management Sales management Financial management
T
i
m
e
T
i
m
e
T
i
m
e
which are being forced to invest more heavily in better Business Intelligence applications, not the least because
of the more stringent compliance requirements. Investments in BI solutions are thus not the sole domain of
large companies; indeed the medium-sized companies hold enormous growth potential and it is these that will
rightly be targeted by all BI suppliers in the intervening period. The need becomes crystal clear when one
thinks that previously only one in four companies used BI software. New legal requirements must be fulfilled
and outdated BI software replaced. Furthermore, the increased functionality and interoperability of the new BI
tools must be represented in the processes. For medium-sized companies, software-as-a-service packages or
hosted solutions for Business Intelligence offer a cost-effective alternative to purchasing software, allowing
them to satisfy the heightened requirements governing reporting and evaluations of business administration. The
Experton Group estimates the total volume of the SaaS market in Germany 2007 at approx. € 270 million (expected
to reach € 577 million by 2010).
In addition to the suppliers of comprehensive BI suites (such as e.g. SAS, Cognos/Applix, Microstrategy etc.)
and the database and ERP suppliers (SAP/Business Objects, Oracle/Hyperion, Microsoft, IBM/Cognos etc.),
the market for Business Intelligence is host to a plethora of specialized suppliers with individual BI tools
or BI functions (data mining, analysis, data integration, data quality management, OLAP, reporting or meta data
management). Furthermore, increasing numbers of open source solution providers are entering the BI market,
including such names as Pentaho, Weka, Rapid Miner, Polo, BIRT etc.
At the moment BI tools are predominantly used in companies in the areas of controlling, accounting and corpo-
rate management (reporting, data analysis, planning/budgeting), however they are making strong inroads into
further specialist areas including marketing and sales. The future challenge facing all sectors, but particularly
the production-intensive ones (manufacturing, automotive and chemicals industries) is to dovetail the data ob-
tained from administration and production in a more structured manner. Therefore suppliers will see increased
demand for solutions which are more sector-specific, as customers attempt to implement Business Intelligence
projects customer-specifically.
Recent studies also reveal that the introduction of a certain BI product is no longer any guarantee of success,
i.e. that the specific goals being pursued (boosted turnover, stronger competitive position, increased employee
satisfaction) will actually be achieved (Source: Orbit). Thus both the selection process for the right software and
that for the right service provider become the primary challenges in the implementation of BI projects.
7
“In the future, BI software should be capable of interaction with Office
applications and search engines, enabling key performance indicator
cockpits and allowing better integration in operative systems and cor-
porate processes.”
!
3. Possible pitfalls of implementing BI-projects.
BI projects are not simple IT projects. BI solutions must be closely dovetailed with the business processes within
the company. It is no wonder then that difficulties in the implementation of a BI project are not exclusively due
to the complex technology or an opaque procedural model in the realization (Sect. 3.4), but that often the fault
lies with the fact that there has been insufficient alignment with business processes (Sect. 3.3). Moreover, many
companies fail to create sufficient acceptance for the planned measures by means of targeted communication
activities amongst the employees (Sect. 3.2). In the end, a BI project can also fail because of a very basic issue.
For example, the BI project is often insufficiently incorporated into the overall context of the corporate strategy
(Sect. 3.1). The challenges facing a company during the planning and implementation of a BI project and the
specific pitfalls to be watched out for are dealt with in this section.
3.1 Lack of direction in the corporate strategy and lack of a BI vision.
Many BI projects fail despite having defined goals and using good BI tools. One frequent reason for this is short-
term, actionistic behavior. Most companies are not pursuing a company-wide BI strategy in their planning and in-
troduction. Rather BI projects are all too often initiated in the specialist areas and driven “from the bottom up”.
Such frequently department-related and uncoordinated BI projects generally lead to disparate island solutions.
As a consequence, the users are faced with a mish-mash of isolated BI warehouses, data mining tools and soft-
ware dashboards, through which they end up trawling once again for the information they need. The formulation
and implementation of an integrated BI strategy, i.e. one that is coordinated to the corporate strategy, is therefore
indispensable.
BI projects often have little in common with traditional IT projects structured according to the waterfall model,
instead it is typical that many details and requirements do not arise until the project is running. Therefore, start
with the motto “Think Big, Start Small!”. Think Big stands for the formulation of a BI vision and encompasses the
analysis of the entire BI potential of the firm itself as well as the coordination of requirements with partners, cus-
tomers and suppliers. For a BI strategy, the path and the fundamental goals must be clearly defined, and this
without spending months on end working on the description of comprehensive data models and BI require-
ments, which become obsolete anyway because of new requirements or findings in the course of the project.
This is why it is important at the start of a BI project to concentrate on small but essential tasks (“Start Small”),
in order to be able to realize visible successes as quickly as possible (Quick Wins). With this step-by-step ap-
proach, the user learns, from the very outset of the project, a great deal about his customer or production data,
particularly in relation to its quality, and can incorporate this knowledge into the next project steps. The solution
is extended iteratively in this way and in time takes shape.
3.2 Lack of BI acceptance within the company due to insufficient communication and transparency.
Contradictory expectations, political reasons, disputes about data sovereignty and budgets often prevent the
creation of a comprehensive BI solution. In addition the introduction of new methods and tools can have a sig-
nificant influence on the organization and work processes, which are not always welcomed enthusiastically by all
employees. The most frequent objections to the introduction of BI applications are that they are difficult to use,
have poor response times and would be demotivating for users. Consequently a BI project can only achieve its
goals if the necessary changes are accepted by the employees.
8
“Bring about a common view of the organization and data that
is shared by all involved and ensure the support of business manage-
ment by pointing out the corporate goals that are supported by BI.”
!
9
“According to a survey, 30% of German salespeople see BI analyses
as an unwanted means of performance control.”
!
Accordingly, it is essential that one’s behavior when introducing and implementing BI projects should be transpar-
ent to the employees and be inherently consistent. This is best supported by the early and targeted inclusion of “key
people” and by defining a BI program that is freely accessible to all employees and easily understood by them.
Having some of the key people lead by example in terms of using BI is a promising introduction strategy, particu-
larly in medium-sized businesses. These key people use the BI system in the firm and thus convince the rest of
the employees by, for example, using the data delivered by the BI system in a very targeted fashion to query suc-
cesses and failures of projects, processes, etc. In this way, they demonstrate the opportunities and the benefits
of the key performance indicators generated by the BI system.
It is important that there must be a clear vision (“Think Big”) behind such a project, which should, above all, pre-
clude short-term actionism. Our project experiences show that those companies that have a strong link between
the various operator and user units are precisely those who enjoy most success in their BI activities. The key
people should therefore cooperate actively on projects from the BI program and involve themselves to a great
degree. Moreover it has proved very practical to nominate a concrete process and/or BI coordinator for the per-
formance of complex and comprehensive projects; the role of these persons would also involve assisting in defin-
ing the required communications and/or information processes. You should therefore initiate an ongoing dialog
between the BI team and the specialist departments, not the least to ensure that the mutual expectations in the
project remain realistic and are continuously coordinated. It is also important to ensure a clear commitment on
the part of business management with respect to the implementation of the BI program and the associated provi-
sion of required resources.
If possible, it is also practical to incorporate an additional, external partner, who has experience in both the BI en-
vironment and also in Change Management. This is because a long-term, committed BI program must go hand
in hand with careful and sensitive change management. This facilitates the required organizational changes and
helps the affected employees to join in. Inform the employees regularly about the current status of the project,
to counteract well in advance any doubts they may have and prevent any negative communication. Early training
measures for the users will increase acceptance and thus boost efficiency and the benefits of the new BI solution.
3.3 Insufficient alignment of the BI solution to the business processes.
Once the BI strategy is formulated and all those involved/affected have been at least won over to the idea, you will
come up against a further obstacle in the path to a successful BI solution: the alignment and incorporation of the
BI solution to and in the business processes.
In order to remain competitive, companies must monitor countless mutually independent activities across all busi-
ness processes, analyze individual results and make decisions on the basis of this information. Often however the
data quantity is not at all manageable and is isolated in ERP, CRM, personnel planning and production systems.
The new BI infrastructure must thus be able to permit access to cross-company, consolidated data from all these
downstream systems.
“The employees must feel that this is about providing them with more
task-related information than was previously the case so that they can
respond more easily, more flexibly and faster to indirect processes.”
!
To this end it is indispensable that the previously defined information processes are precisely mapped and that
they interact closely with the data management and security systems. The aim must be to automate the data col-
lection, storage and evaluation processes to as great an extent as possible. The smaller the manual influence on
this process is, the less the likelihood of errors and the greater the cost optimization.
A deep understanding of the business processes and, in particular, the underlying data sources is thus critical!
However, in the implementation of BI projects, it is frequently the technology that is at the forefront. With the as-
sumption that the users already recognize the added value, many IT departments design BI projects with a strong
focus on improving individual applications but with too little attention on the optimization, adjustment and correct
analysis of the business processes.
Although control of the technological facilities is without doubt a significant prerequisite for a successful implemen-
tation of a BI project, it is important not to lose sight of the processes and sector-specific peculiarities. Thus one
should focus less on the data itself and more on its use, i.e. the business processes that are to benefit from it – such
as sales, procurement, financial services or HR.
Valuable information also depends on partners, suppliers and customers, which means that they must be in-
corporated into the implementation of the BI projects. Having many people involved and numerous data islands
inevitably leads to a high level of complexity. Correctly representing these complex information structures and
business processes in the BI tools is without doubt the greatest technical challenge facing companies.
3.4 Complex technology & opaque procedural model.
A fundamental understanding of the sector and processes is elementary for BI projects, but their successful im-
plementation is also strongly influenced by technology and a well-defined procedural model. Numerous techni-
cal challenges must be overcome – challenges where poor planning and implementation can quickly bring any
BI project to its knees.
Requirements of a BI project from a predominantly technological viewpoint:
Creation of transparency despite the great complexity of the processes and the different technologies
High scalability and quick adjustability of the solution
Managing the flood of data and creation of transparency
Clarification of the data sovereignty to ensure consistency
Generation of a high data quality
Safe handling of sensitive data
Technical evaluation of the BI solutions
A lack of transparency quickly leads to errors, high costs and increased rejection. This applies both for the BI proj-
ect and for the day-to-day work with the new BI tool. To avoid this, particular emphasis should be placed on the ar-
chitecture of the BI solution and on the optimization of any existing BI infrastructure. The BI infrastructure should
be able to represent the defined information processes completely, making a detailed analysis of the specific re-
quirements of the perfect BI infrastructure indispensable. In this it is important to note that frequently these can-
not be transferred from pre-existing IT requirements due to the high process orientation. The highly complex na-
ture of the processes and the multitude of applications lead again and again to an intermingling of logic and tech-
nology here.
A good means of combating the complexity – in keeping with the Start Small scenario – is to develop a modular
approach. This can contribute towards reducing the complexity and thus creating more transparency. Now more
than ever, IT organizations must consistently introduce new solutions that are in harmony with the commercial
10
requirements. This requires that IT coordinators react flexibly and quickly build up new IT processes that are
in synch with the business requirements. Service-oriented architectures (SOA) make a valuable contribution here.
SOA is more than just a technology, rather it is a concept that structures the IT of a company according to services,
which are modular and can be combined flexibly to implement IT processes. The IT-specific components of the
concept are realized with an IT architecture, which is based on loosely coupled services which are mutually inde-
pendent from a technical viewpoint and the interoperability of which is based on open standards, thus enabling
a clear separation between logic and technology. The standardized and modular structure of SOA can be directly
incorporated into tailored solution portfolios and offers many advantages:
Efficiency and short time-to-market thanks to a standardized and simple development
of new services and processes.
Reusability of IT assets and costs savings through the use of common resources.
Transparency of the IT architecture through clearly defined services.
Increased flexibility through standardized interfaces and the loose coupling of services.
Data quality encompasses the importance, relevance and correctness of information. Insufficient quality or the lack
of required starting data as well as a possible paucity of data structure can all impair the kick-off phase of a BI proj-
ect. In ongoing operation, poorly maintained data is a frequent problem. In addition to the enforcement of data
maintenance standards (clear policies), data quality should also be improved through automatic quality checks and
the implementation of a data quality assurance process. This is where Master Data Management (MDM) comes in-
to play. A good MDM system that is planned early on produces clearly defined master data and is thus the basis for
high data quality. In addition, by stipulating clear policies it must be possible to clearly regulate who must provide
what data and the quality and format of such data (data governance). Data governance is becoming more and
more important, particularly given the ever increasing data quantities, and forms the basis for defining clear respon-
sibilities. Accordingly a specific area/specific department must be uniquely responsible for clearly defined busi-
ness processes and the resultant data. Only such a regulation will guarantee consistency in the data. A continuous
integration and consolidation of data sources serves not only to create a company-wide view but also improves da-
ta quality, as the different data sources often only hold the individual values in very different formats. The simplest
example is the date format (e.g 25 Feb. 74, 25.02.1974 etc.). In addition, the data structure of operative systems dif-
fers widely from the structure of the multidimensional data cubes presented above.
11
“Service-oriented architectures increase the required adjustability
of a BI solution and strengthen the separation between technology
and logic, through the use of clearly defined services for representing
individual business processes.”
!
12
“Missing or faulty data reduces the benefits of BI solutions, promotes
incorrect decisions and makes it more difficult to adhere to legal frame-
work conditions.”
!
A further important aspect of MDM is the issue of data security. A lack of security will lead not only to rejection
by the majority of users, but also makes unauthorized access to valuable data much easier. This can jeopardize
the competitivity of individual departments and even the entire company. For this reason, data security must
be a significant issue in the definition of a BI program. The central elements of a high security standard are
contained in the definition of clear data security strategies as well as the consideration of the most modern se-
curity tools during implementation. Through a continuous further development of the security processes intro-
duced, the achievement of hopefully high security standards is guaranteed.
MDM plays a vital role in achieving a high level of transparency in the implementation of BI projects, despite the
complexity of the technology. The success of BI is predicated on the consistent definition and implementation
of rules to ensure data quality and security that includes a clear description of responsibilities. The implementa-
tion of MDM policies also requires a detailed procedural model, in order to set up the planned BI solutions and
the relevant business processes in accordance with the specific customer challenges.
4. Conclusion & outlook.
13
“The decisive factor for future business success will be the ability to iden-
tify as quickly as possible from amongst the gargantuan mass of data
the correct and important information, irrespective of its data structure.”
!
“This is merely the dawn of the information age. When one thinks that
information technology is only a very young science, and when one
considers the explosion in the volume of data that we have witnessed
in recent years, it is easy to see that the importance of information will
continue to grow in this fast-paced environment.”
!
At the heart of Business Intelligence is the wish to identify as quickly as possible and to better understand the
cause-and-effect relationships and mechanisms that are of relevance to one’s own business by analyzing exist-
ing data. Business-relevant information must therefore be made available in real time and in a comprehensible
format; this data must be of a high quality and be up-to-date. On the one hand, users should be able to easily ac-
cess the information, while on the other the data must be protected against unauthorized access. If it is imple-
mented consistently, a Business Intelligence system that is integrated into the operative, tactical and strategic
business can make a huge contribution to the success of the company.
The desire for more findings that are available faster and are above all relevant is pitted against the increasing
flood of data and the ever more complex data analyses this entails. This is a continuous development, which
is strengthened not the least by the merger of IT and TK. New, innovative ICT solutions and concepts such as
Medical Card, Location Based Services, Homeland Security, etc. are generating data quantities in the petabyte
range, and in this environment it is not only the number of structured information units that is rising but also the
quantity of unstructured and semi-structured information. This includes e.g. documents, emails, tables etc.,
which are saved in a disparate fashion, but in which the information is not clearly structured.
According to an Ovum study, approx. 80-85% of a company’s information, i.e. by far the greatest share of knowl-
edge, is poorly structured or entirely unstructured even today. Consequently it is necessary not only to link the dif-
ferent concepts and systems more closely together but above all to gain control over the integration and analysis
of structured and unstructured data stocks. This is because the anticipated explosion of the Internet through
Web 2.0 and Web 3.0 is of late leading to a further exponential rise in unstructured and semi-structured information,
which will bring with it a new, even larger and nearly overwhelming flood of information.
In the future, BI concepts must merge with Information Management concepts, the approach of which is to efficiently
enable both access to and analysis of structured and unstructured information. It is critical for the acceptance
of future solutions that data is made available in real time, that it is specific to the users and above all that it is clear.
The users should not, however, come into contact with the complexity of the data sources and structures under-
lying such solutions – to avoid, inter alia, excessive demand.
Only companies that succeed in identifying, analyzing and preparing appropriately the critical data as quickly as
possible and which can thus identify the changes and trends at an early stage, allowing them to react as quickly
as possible, will be able to compete at the highest levels in the long term. Consequently, innovative strategies,
concepts and technologies from Information Management, which allow the right information to be obtained reli-
ably and independent of location, data source and format in real time, and thus form the basis to be able to make
fast and intelligent decisions, are important.
5. Glossary.
Balanced Scorecard Performance measurement technique, which allows corporate management to
measure the various internal functions and its results against the outside world.
Organizations use this to achieve strategic goals which are often divided into four
fields (finances, processes, customers/market, personnel).
Basel II Basel II is the title used to cover all the equity capital regulations proposed by the
Basel Committee on Banking Supervision in recent years. Since January 1, 2007
these must be applied to all credit and financial services institutions in the mem-
ber states of the European Union.
Business Intelligence Business Intelligence describes a collection of methods, which can be used to gain
valuable information from structured data.
Change Management Change Management covers all tasks, measures and activities, the aim of which
is to bring about a comprehensive, cross-group and far-reaching change – to
implement new strategies, structures, etc. – within an organization.
CRM Customer Relationship Management (CRM) describes procedures and techniques
with which the relationship between customers and suppliers can be represented.
Cross-Selling Cross-Selling is the term used in marketing to designate the sale of supplementary
products and services.
Customer Lifetime Value Customer Lifetime Value (CLV) is, generally speaking, the current value of the likely
future income stream generated by a customer over his entire “customer life”.
Dashboard Dashboard designates a visualization form for information in compressed format.
Data Mart In contrast to a data warehouse, a data mart is not a company-wide database.
Data marts are restricted to parts of companies, e.g. departments, groups, prod-
uct divisions and are therefore a good means of implementing BI solutions
in a short space of time, in accordance with a modular Start Small approach.
Data Mining Data mining is a special form of data analysis which reveals hidden trends. In ad-
dition, analysis tools are used to discover invisible patterns and to validate known
factors.
Data consistency Data consistency for databases is, generally speaking, the lack of contradictions
in data.
Data Warehouse (DW) A data warehouse is a database, isolated from the operational DP system, which
serves as a company-wide database for all system versions to support manage-
ment. A data warehouse is, among other things, the basis (enabling technology)
for CRM, Data Mining, E-Business etc.
14
ERP The term Enterprise Resource Planning describes the corporate task of deploying
the resources (capital, equipment or personnel) available in a company efficiently
for the purpose of the operational processes.
Governance Governance is, in general terms, the control and regulation system for the structures
(setup and process organization) of a company, organization or also an operation.
IFRS The International Financial Reporting Standards (IFRS) are international accounting
regulations.
IT Governance IT governance comprises management, organizational structures and processes
that ensure that the IT supports the corporate strategy and goals.
IT Outsourcing Transfer of full responsibility for IT functions with a high IT share to legally inde-
pendent – i.e. external – service providers for a defined period.
KonTraG The Gesetz zur Kontrolle und Transparenz im Unternehmensbereich (law providing
for control and transparency in the Group), KonTraG for short, is a comprehensive
Artikelgesetz, a composite act covering the amendment and enactment of the re-
lated legislation, the aim of which is to improve corporate governance in German
companies.
OLAP Online Analytical Processing (OLAP) is the analysis and evaluation of multidimen-
sional data, to gain information for company decisions.
SaaS Software as a Service (SaaS) is software the use of which is granted to a customer
as an ongoing service for a usage fee (“rent”) that must be paid periodically.
Sarbanes-Oxley Act The Sarbanes-Oxley Act (SOX) is a US capital market law enacted in 2002, accord-
ing to which all companies listed on US stock exchanges must have their internal
control systems overseen, documented and tested by auditors.
SOA Service Oriented Architecture (SOA) is a management concept and requires a system
architecture concept as a secondary concern only. This concept aspires to an infra-
structure which is aligned to the desired business processes and which can react
quickly to changed requirements in the business environment. The system architec-
ture concept provides for the provision of technical services and functionalities
in the form of services that represent nuclear process steps.
US-GAAP US-GAAP is the abbreviation for the United States Generally Accepted Accounting
Principles. It refers to the US accounting regulations that govern the accounting
practices and annual financial statements of companies.
Source: According to the studies, articles, et al documented in the bibliography.
15
6. Bibliography.
Sascha Alexander, “Business Intelligence: Was Unternehmen wirklich brauchen” [Business Intelligence: What
Companies Really Need”], Computerwoche, 2007
Dr. Michael Böhnlein, “Todsünden des Data Warehouse” [Seven Deadly Sins of the Data Warehouse], 2007
Dr. Michael Böhnlein , “Stolpersteine bei BI-Projekten und aktuelle BI-Trends” [Stumbling Blocks of BI Projects
and Current BI Trends], 2007
Barney Finucane, “Vor- und Nachteile neuer Business-Intelligence-Ansätze” [Advantages and Disadvantages
of New Business Intelligence Approaches], CIO, 2007
Dr. Peter Gluchowski, “Quo vadis Business Intelligence?”, BI-Spektrum, 2006
Knut Hildebrand (Hrsg.), “Business Intelligence”, HMD-Praxis der Wirtschaftsinformatik, 2001
Dr. Thoms Jurisch, “Business Intelligence – Schneller und besser als die Wettbewerber” [Business Intelligence –
Faster and Better than the Competition], 2007
Dr. Joachim Philippi, “SOA trifft auf BI” [SOA meets BI], CIO, 2007
Dr. Joachim Philippi, “Die sieben größten BI-Fehler und wie man sie vermeidet” [The Seven Greatest BI Errors
and How to Avoid Them], Computerwoche, 2005
Dr. Joachim Philippi, “Weltweiter Markt für Business Intelligence wächst” [Global Market for Business Intelligence
is Growing], Computerwoche, 2007
Dr. Joachim Philippi, “Deutsche Vertriebsmitarbeiter haben ein gespaltenes Verhältnis zu Business Intelligence”
[German Sales Employees Have an Ambivalent Attitude to Business Intelligence], Computerwoche, 2007
Dr. Joachim Philippi, “Die Lösung heißt Maßanzug” [The Solution is a Custom-Made Suit], Orbit, 2007
Dr. Joachim Philippi, “BI-Strategie – Der Weg zum maximalen Return-On-BI-Invest” [BI Strategy – The Road
to the Maximum Return-On-BI-Investment], TDWI, 2007
Dr. Joachim Philippi “White Paper Dynamic Services – Flexible ICT-Ressourcen” [White Paper Dynamic Services
– Flexible ICT Resources], T-Systems 2007
Dr. Joachim Philippi, “White Paper Real ICT – Flexibilität durch ganzheitlich integrierte IT- und TK-Services”
[White Paper Real ICT – Flexibility through Comprehensively Integrated IT and TK Services], T-Systems, 2007
Dr. Joachim Philippi, “White Paper Identity und Access Management (IAM)” [White Paper Identity and Access
Management (IAM)], T-Systems, 2007
Andreas Schaffry, “Die richtige Strategie bei BI-Projekten” [The Right Strategy for BI Projects], CIO, 2007
Kurt Schlegel, “Business Intelligence Trends and Best Practices”, Gartner, 2007
Markus Trottnow, Markus Bereszewski: “Den Wandel intelligent gestalten” [Making the Change Intelligently],
Information Week, 2007
Rainer Volck, “Business Intelligence in Echtzeit” [Business Intelligence in Real Time], 2007
Nicolas Zeitler, “Mangelnde Ausrichtung an der Unternehmensstrategie – Firmen planen BI-Projekte oft
schlecht” [Lack of Alignment to the Corporate Strategy – Firms Often Plan BI Projects Poorly], CIO, 2007
16
Publisher:
T-Systems Enterprise Services GmbH
Mainzer Landstr. 50
D - 60325 Frankfurt
Responsible for the contents:
Sales Management BI & Performance Management
Contact:
T-Systems Enterprise Services GmbH
Corporate Marketing & Communications
Uli Kunesch
Fasanenweg 5
D- 70771 Leinfelden-Echterdingen
E-mail: [email protected]
T-Systems Business Services GmbH
Sales Management BI & Performance Management
Franz Pfitzner
Dachauer Str. 665
D- 80995 München
E-mail: [email protected]
doc_742088429.pdf
White Paper Business Intelligence. Optimized decisions quickly
White Paper
Business Intelligence.
Optimized decisions quickly
and transparently.
Contents at a glance.
3 1. Introduction.
4 2. Sourcing information from company data: from data to knowledge.
4 2.1 Problems of an information glut – BI as the new solution?
5 2.2 BI in practice – a simplified example.
6 2.3 Evaluation of the German BI market.
8 3. Possible pitfalls in the realization of BI projects.
8 3.1 Lack of direction in the corporate strategy and lack of a BI vision.
8 3.2 Lack of BI acceptance within the company due to insufficient communication and transparency.
9 3.3 Insufficient alignment of the BI solution to the business processes.
10 3.4 Complex technology & opaque procedural model.
13 4. Conclusion & outlook.
14 5. Glossary.
16 6. Bibliography.
1. Introduction.
Ever shorter market cycles, faster changing customer requirements and growing cost and competitive pressures
continuously intensify the demands placed on companies. Fast, effective and sound strategic decisions are needed
to achieve long-term competitivity. Comprehensive and up-to-date, but above all precisely tailored, information
helps not only to better understand one’s own business and the underlying contexts, it also creates a solid basis
for decision-making. In this context, Business Intelligence has laid the foundations in recent years, allowing deci-
sion-making that is increasingly well-founded. However, explosively burgeoning data quantities and ever faster deci-
sion-making cycles require even better, more comprehensive, sophisticated and up-to-date reporting avenues.
In the early days of electronic data processing, pure data management was the focus. It did not, however, take
long to see that the data gathered could, when structured and compressed, provide valuable information on the
most varied of questions. The growing support of corporate business processes by IT, the global use of Internet
technology and the new methods and techniques for gathering and processing data have virtually revolutionized
the opportunities for management support. Nowadays we can use the Business Intelligence (BI) approach
derived from this to transform collected company and competitor data quickly and transparently into decision-
relevant knowledge. Very specific questions, such as “What is the best-selling product in a certain region?”,
can be answered in a matter of seconds and there are no longer any time lags between creating reports and
their presentation. The data arrives on the manager’s desk in real-time, i.e. the minute it is created.
However, there is no plug & play version of the perfect BI application. Given the high complexity of the data
structures and the multiple relationships as well as the constantly developing requirements, the introduction
of a BI application must not only be well planned it must be very prudently implemented. This white paper
deals with the pitfalls facing a company when planning, introducing and using BI solutions and offers a brief
look into the future of analyzing and preparing company information.
3
“The decisive factor for a manager is to filter from a huge glut of data
the current, correct and above all relevant information, thus enabling
him to make sound decisions.”
!
2. Sourcing information from company data:
from data to knowledge.
2.1 Problems of an information glut – BI as the new solution?
The growth in data quantities which has been exponential in recent years is posing great challenges for many
companies. From 2000 to 2002, for example, more data was generated than in all of human history before that.
In the period from 2003 to 2005 this data quantity quadrupled again. Naturally, the resultant glut of information
brings challenges as well as just advantages for companies. While they have – in contrast to times past – a great
deal of information at their fingertips, it is becoming increasingly difficult to sift out the information that is truly
relevant to them.
This is where Business Intelligence solutions come in. From an enormous quantity of data, these solutions find
the data that is actually relevant, organize and analyze it and then present it for review in the desired format.
Business Intelligence is generally understood here as a process that converts data into information. The objective
is to provide real-time, transparent, reliable and high quality data to offer the decision-makers within a company
a valid base of information. This is done with the aid of analytical concepts and IT systems, which evaluate the data
about the company itself, the employees or the market development. With the findings gained, companies can
derive the action that is required to optimize their customer and supplier relationships, lower costs or minimize risks
and thus improve their added value on a permanent basis.
The data to be evaluated here is based on the existing business processes in a company. Hence the data that op-
timally supports the fulfillment of the relevant individual tasks is always provided, tailored to the user – whether
a manager or employee.
Historically speaking, BI applications represent a further development of Management Information Systems (MIS),
Decision Support Systems (DSS), Executive Information Systems (EIS) and Data Warehouse (DW). If the first infor-
mation systems were designed for pure administration of data, then the data warehouses were already designed
for real-time provision of information to a variety of decision-makers and employees in different departments (Figure 1).
Figure 1: Historic development towards Business Intelligence.
4
“More and more companies have come to realize that data and informa-
tion quality is a value-added factor they have thus far largely neglected.”
!
MIS
Efficient file
processing
Integrated systems
Vision of the automatic
decision generator
BI
Key performance
indicator systems
(Balanced Scorecards)
Analytical applications
Data mining
DW
Integration of multiple
data sources
Interactive queries
/OLAP
Historical data
EIS
Multimed. modelling
Separation of operative
systems
Restriction to top
management
DSS
Statistical algorithms,
“what-if”
Complex, rigid
structures
Database orientation
1960 2000 1990 1980 1970
MIS - Management Information Systems
DSS - Decision Support Systems
EIS - Executive Information Systems
DW - Data Warehouses
BI - Business Intelligence
By focusing on the information that is relevant for decision-making, Business Intelligence shall provide methods
and tools which will enable company and competitor data to be transformed into decision-relevant knowledge.
This yields not only a comprehensible and targeted representation of this information (consistent, comparable
and cross-area), but naturally also promises a wealth of further advantages:
Shorter access time to relevant information thanks to real-time data for a faster response to the market
Minimization of the risk of a bad decision, through established, detailed and targeted market information
Ongoing evaluation of the delivery terms and conditions with resultant optimization options
A comprehensive customer understanding through cross-selling or preference analyses to maximize
the customer lifetime values
Detailed overview of the individual cash flows in a company and thus naturally also the identification
of cost saving potential
In addition to the competitive advantages that can be generated, BI applications also cater to the increasing
regulatory requirements (e.g. Basel II, KonTraG, Sarbanes-Oxley Act) and international accounting standards
(in particular IFRS and US-GAAP). These force companies to develop cleaner data scenarios, allowing them
to shape the information flows transparently even with complex processes.
2.2 BI in practice – a simplified example.
To demonstrate the BI approach, a simple real-world example of an industrial bakery, i.e. a medium-sized branch,
is very effective. You are responsible for a chain of bakeries with locations in five different German states and now
intend to open a new branch in North Rhine Westphalia. First you must find a suitable location. To this end you will
analyze, for example, in what region the ratio of turnover to labor and operating costs is particularly cost-effec-
tive or where the competitive situation might be advantageous for you. Germany boasts the most bread vari-
eties (approx. 1,200) in the world, so you are very interested in discovering the bread varieties primarily demanded
in the region. Of course, further, very different and more in-depth analyses are also interesting. These may include
the following, e.g.:
Product and range (success) analysis:
How has the sale of individual products (e.g. poppy seed rolls) developed over a defined period in the
region in focus?
Turnover analysis:
What was turnover development like for the other branches initially? What setbacks must be planned
for? What level of turnover can be realistically expected in the first 12 months?
Branch success analysis:
What products are offered in the branches, which achieved the largest turnover in one month in the region?
How large is the branch and how many people work there?
With the aid of a BI tool, these and many other specific questions can easily be answered from a structured data-
base. To this end, the data is first imported from various data sources and stored in a multi-dimensional informa-
tion cube, and if required collated anew into smaller data cubes that are optimized from a user viewpoint (so-called
“data marts”). The axes of a data cube are used to categorize the different facets of the information obtained.
The content of such a data cube can include quantities sold, quantities produced and turnover, but also regions,
branches, sellers etc. Thus, depending on the focus of the user (product, turnover, sales), the relevant data is ex-
tracted from the data cube and presented in a specially generated view (see Figure 2).
5
Figure 2: Multidimensional view: Information from data; Source: Orbit.
From a structured database, a BI tool can be used to extract all the relevant information quickly and easily and to display
this information clearly. Thus, it is possible to use the data stocks of a company more effectively as a production factor
using this data cube.
2.3 Evaluation of the German BI market.
The advantages of a BI solution are obvious. Nonetheless, the Business Intelligence approach spans a multitude
of different tools and concepts. Correspondingly, there are numerous suppliers, who all promise a timely and
cost-effective launch of BI solutions. The market for Business Intelligence software and services is thus currently
very unclear. Although all suppliers market using the label “Business Intelligence”, ultimately the individual prod-
ucts all too frequently address entirely different commercial tasks and target groups. This section therefore deals
with the “status quo” on the German BI market and provides a brief outlook at the direction in which this market
will move in the coming years.
At the moment the BI market is in a strong consolidation phase and in 2007 it was particularly defined by large takeovers.
Three acquisitions have influenced the market, in particular, in this year: the takeovers of Hyperion by Oracle, of Cognos
by IBM and not the least of Business Objects by SAP. The aim of the established software companies is to position
themselves well in a future growth market through acquisitions. As a consequence of this development, BI has become
an integral component of future-oriented ERP solutions.
According to estimates of the Experton Group, the BI market in Germany in 2007 was set to reach the dimensions
of a mass market of approx. € 1.5 billion. In 2007, more and more medium-sized businesses invested in the area
of Business Intelligence. In total, this was split approximately 47% on software and 53% on services, whereby
“the ratio of software to services was clearly tipped towards software amongst the SMEs. For large companies,
the services share is significantly higher due to the great integration complexity and the required adjustment to specif-
ic business processes” explains Frank Schmeiler, Research Director of the Experton Group.
In the coming years, the BI market will continue to show above-average annual growth approximating 20% and will
increase by 2010 to represent a market value of more than € 2.5 billion. And as BI projects are becoming ever
more complex and comprehensive, the services market will grow in step with the BI software market. The rather op-
timistic growth forecast of 20% can largely be traced back to the required investment from medium-sized companies,
Raisins
Sesame seed
Poppy seed
Wholemeal
Normal
B
e
r
l
i
n
L
o
w
e
r
S
a
x
o
n
y
N
o
r
t
h
R
h
in
e
W
e
s
t
p
h
a
lia
R
h
i
n
e
l
a
n
d
-
P
a
l
a
t
i
n
a
t
e
H
e
s
s
e
2006
2005
2004
2003
2002
P
r
o
d
u
c
t
Raisins
Sesame seed
Poppy seed
Wholemeal
Normal
2006
2005
2004
2003
2002
B
e
r
l
i
n
L
o
w
e
r
S
a
x
o
n
y
N
o
r
t
h
R
h
in
e
W
e
s
t
p
h
a
lia
R
h
i
n
e
l
a
n
d
-
P
a
l
a
t
i
n
a
t
e
H
e
s
s
e
P
r
o
d
u
c
t
Region
6
Raisins
Sesame seed
Poppy seed
Wholemeal
Normal
B
e
r
l
i
n
L
o
w
e
r
S
a
x
o
n
y
N
o
r
t
h
R
h
in
e
W
e
s
t
p
h
a
lia
R
h
i
n
e
l
a
n
d
-
P
a
l
a
t
i
n
a
t
e
H
e
s
s
e
2006
2005
2004
2003
2002
Region
P
r
o
d
u
c
t
Region
Focus: Product Focus: Sales Focus: Turnover
Product management Sales management Financial management
T
i
m
e
T
i
m
e
T
i
m
e
which are being forced to invest more heavily in better Business Intelligence applications, not the least because
of the more stringent compliance requirements. Investments in BI solutions are thus not the sole domain of
large companies; indeed the medium-sized companies hold enormous growth potential and it is these that will
rightly be targeted by all BI suppliers in the intervening period. The need becomes crystal clear when one
thinks that previously only one in four companies used BI software. New legal requirements must be fulfilled
and outdated BI software replaced. Furthermore, the increased functionality and interoperability of the new BI
tools must be represented in the processes. For medium-sized companies, software-as-a-service packages or
hosted solutions for Business Intelligence offer a cost-effective alternative to purchasing software, allowing
them to satisfy the heightened requirements governing reporting and evaluations of business administration. The
Experton Group estimates the total volume of the SaaS market in Germany 2007 at approx. € 270 million (expected
to reach € 577 million by 2010).
In addition to the suppliers of comprehensive BI suites (such as e.g. SAS, Cognos/Applix, Microstrategy etc.)
and the database and ERP suppliers (SAP/Business Objects, Oracle/Hyperion, Microsoft, IBM/Cognos etc.),
the market for Business Intelligence is host to a plethora of specialized suppliers with individual BI tools
or BI functions (data mining, analysis, data integration, data quality management, OLAP, reporting or meta data
management). Furthermore, increasing numbers of open source solution providers are entering the BI market,
including such names as Pentaho, Weka, Rapid Miner, Polo, BIRT etc.
At the moment BI tools are predominantly used in companies in the areas of controlling, accounting and corpo-
rate management (reporting, data analysis, planning/budgeting), however they are making strong inroads into
further specialist areas including marketing and sales. The future challenge facing all sectors, but particularly
the production-intensive ones (manufacturing, automotive and chemicals industries) is to dovetail the data ob-
tained from administration and production in a more structured manner. Therefore suppliers will see increased
demand for solutions which are more sector-specific, as customers attempt to implement Business Intelligence
projects customer-specifically.
Recent studies also reveal that the introduction of a certain BI product is no longer any guarantee of success,
i.e. that the specific goals being pursued (boosted turnover, stronger competitive position, increased employee
satisfaction) will actually be achieved (Source: Orbit). Thus both the selection process for the right software and
that for the right service provider become the primary challenges in the implementation of BI projects.
7
“In the future, BI software should be capable of interaction with Office
applications and search engines, enabling key performance indicator
cockpits and allowing better integration in operative systems and cor-
porate processes.”
!
3. Possible pitfalls of implementing BI-projects.
BI projects are not simple IT projects. BI solutions must be closely dovetailed with the business processes within
the company. It is no wonder then that difficulties in the implementation of a BI project are not exclusively due
to the complex technology or an opaque procedural model in the realization (Sect. 3.4), but that often the fault
lies with the fact that there has been insufficient alignment with business processes (Sect. 3.3). Moreover, many
companies fail to create sufficient acceptance for the planned measures by means of targeted communication
activities amongst the employees (Sect. 3.2). In the end, a BI project can also fail because of a very basic issue.
For example, the BI project is often insufficiently incorporated into the overall context of the corporate strategy
(Sect. 3.1). The challenges facing a company during the planning and implementation of a BI project and the
specific pitfalls to be watched out for are dealt with in this section.
3.1 Lack of direction in the corporate strategy and lack of a BI vision.
Many BI projects fail despite having defined goals and using good BI tools. One frequent reason for this is short-
term, actionistic behavior. Most companies are not pursuing a company-wide BI strategy in their planning and in-
troduction. Rather BI projects are all too often initiated in the specialist areas and driven “from the bottom up”.
Such frequently department-related and uncoordinated BI projects generally lead to disparate island solutions.
As a consequence, the users are faced with a mish-mash of isolated BI warehouses, data mining tools and soft-
ware dashboards, through which they end up trawling once again for the information they need. The formulation
and implementation of an integrated BI strategy, i.e. one that is coordinated to the corporate strategy, is therefore
indispensable.
BI projects often have little in common with traditional IT projects structured according to the waterfall model,
instead it is typical that many details and requirements do not arise until the project is running. Therefore, start
with the motto “Think Big, Start Small!”. Think Big stands for the formulation of a BI vision and encompasses the
analysis of the entire BI potential of the firm itself as well as the coordination of requirements with partners, cus-
tomers and suppliers. For a BI strategy, the path and the fundamental goals must be clearly defined, and this
without spending months on end working on the description of comprehensive data models and BI require-
ments, which become obsolete anyway because of new requirements or findings in the course of the project.
This is why it is important at the start of a BI project to concentrate on small but essential tasks (“Start Small”),
in order to be able to realize visible successes as quickly as possible (Quick Wins). With this step-by-step ap-
proach, the user learns, from the very outset of the project, a great deal about his customer or production data,
particularly in relation to its quality, and can incorporate this knowledge into the next project steps. The solution
is extended iteratively in this way and in time takes shape.
3.2 Lack of BI acceptance within the company due to insufficient communication and transparency.
Contradictory expectations, political reasons, disputes about data sovereignty and budgets often prevent the
creation of a comprehensive BI solution. In addition the introduction of new methods and tools can have a sig-
nificant influence on the organization and work processes, which are not always welcomed enthusiastically by all
employees. The most frequent objections to the introduction of BI applications are that they are difficult to use,
have poor response times and would be demotivating for users. Consequently a BI project can only achieve its
goals if the necessary changes are accepted by the employees.
8
“Bring about a common view of the organization and data that
is shared by all involved and ensure the support of business manage-
ment by pointing out the corporate goals that are supported by BI.”
!
9
“According to a survey, 30% of German salespeople see BI analyses
as an unwanted means of performance control.”
!
Accordingly, it is essential that one’s behavior when introducing and implementing BI projects should be transpar-
ent to the employees and be inherently consistent. This is best supported by the early and targeted inclusion of “key
people” and by defining a BI program that is freely accessible to all employees and easily understood by them.
Having some of the key people lead by example in terms of using BI is a promising introduction strategy, particu-
larly in medium-sized businesses. These key people use the BI system in the firm and thus convince the rest of
the employees by, for example, using the data delivered by the BI system in a very targeted fashion to query suc-
cesses and failures of projects, processes, etc. In this way, they demonstrate the opportunities and the benefits
of the key performance indicators generated by the BI system.
It is important that there must be a clear vision (“Think Big”) behind such a project, which should, above all, pre-
clude short-term actionism. Our project experiences show that those companies that have a strong link between
the various operator and user units are precisely those who enjoy most success in their BI activities. The key
people should therefore cooperate actively on projects from the BI program and involve themselves to a great
degree. Moreover it has proved very practical to nominate a concrete process and/or BI coordinator for the per-
formance of complex and comprehensive projects; the role of these persons would also involve assisting in defin-
ing the required communications and/or information processes. You should therefore initiate an ongoing dialog
between the BI team and the specialist departments, not the least to ensure that the mutual expectations in the
project remain realistic and are continuously coordinated. It is also important to ensure a clear commitment on
the part of business management with respect to the implementation of the BI program and the associated provi-
sion of required resources.
If possible, it is also practical to incorporate an additional, external partner, who has experience in both the BI en-
vironment and also in Change Management. This is because a long-term, committed BI program must go hand
in hand with careful and sensitive change management. This facilitates the required organizational changes and
helps the affected employees to join in. Inform the employees regularly about the current status of the project,
to counteract well in advance any doubts they may have and prevent any negative communication. Early training
measures for the users will increase acceptance and thus boost efficiency and the benefits of the new BI solution.
3.3 Insufficient alignment of the BI solution to the business processes.
Once the BI strategy is formulated and all those involved/affected have been at least won over to the idea, you will
come up against a further obstacle in the path to a successful BI solution: the alignment and incorporation of the
BI solution to and in the business processes.
In order to remain competitive, companies must monitor countless mutually independent activities across all busi-
ness processes, analyze individual results and make decisions on the basis of this information. Often however the
data quantity is not at all manageable and is isolated in ERP, CRM, personnel planning and production systems.
The new BI infrastructure must thus be able to permit access to cross-company, consolidated data from all these
downstream systems.
“The employees must feel that this is about providing them with more
task-related information than was previously the case so that they can
respond more easily, more flexibly and faster to indirect processes.”
!
To this end it is indispensable that the previously defined information processes are precisely mapped and that
they interact closely with the data management and security systems. The aim must be to automate the data col-
lection, storage and evaluation processes to as great an extent as possible. The smaller the manual influence on
this process is, the less the likelihood of errors and the greater the cost optimization.
A deep understanding of the business processes and, in particular, the underlying data sources is thus critical!
However, in the implementation of BI projects, it is frequently the technology that is at the forefront. With the as-
sumption that the users already recognize the added value, many IT departments design BI projects with a strong
focus on improving individual applications but with too little attention on the optimization, adjustment and correct
analysis of the business processes.
Although control of the technological facilities is without doubt a significant prerequisite for a successful implemen-
tation of a BI project, it is important not to lose sight of the processes and sector-specific peculiarities. Thus one
should focus less on the data itself and more on its use, i.e. the business processes that are to benefit from it – such
as sales, procurement, financial services or HR.
Valuable information also depends on partners, suppliers and customers, which means that they must be in-
corporated into the implementation of the BI projects. Having many people involved and numerous data islands
inevitably leads to a high level of complexity. Correctly representing these complex information structures and
business processes in the BI tools is without doubt the greatest technical challenge facing companies.
3.4 Complex technology & opaque procedural model.
A fundamental understanding of the sector and processes is elementary for BI projects, but their successful im-
plementation is also strongly influenced by technology and a well-defined procedural model. Numerous techni-
cal challenges must be overcome – challenges where poor planning and implementation can quickly bring any
BI project to its knees.
Requirements of a BI project from a predominantly technological viewpoint:
Creation of transparency despite the great complexity of the processes and the different technologies
High scalability and quick adjustability of the solution
Managing the flood of data and creation of transparency
Clarification of the data sovereignty to ensure consistency
Generation of a high data quality
Safe handling of sensitive data
Technical evaluation of the BI solutions
A lack of transparency quickly leads to errors, high costs and increased rejection. This applies both for the BI proj-
ect and for the day-to-day work with the new BI tool. To avoid this, particular emphasis should be placed on the ar-
chitecture of the BI solution and on the optimization of any existing BI infrastructure. The BI infrastructure should
be able to represent the defined information processes completely, making a detailed analysis of the specific re-
quirements of the perfect BI infrastructure indispensable. In this it is important to note that frequently these can-
not be transferred from pre-existing IT requirements due to the high process orientation. The highly complex na-
ture of the processes and the multitude of applications lead again and again to an intermingling of logic and tech-
nology here.
A good means of combating the complexity – in keeping with the Start Small scenario – is to develop a modular
approach. This can contribute towards reducing the complexity and thus creating more transparency. Now more
than ever, IT organizations must consistently introduce new solutions that are in harmony with the commercial
10
requirements. This requires that IT coordinators react flexibly and quickly build up new IT processes that are
in synch with the business requirements. Service-oriented architectures (SOA) make a valuable contribution here.
SOA is more than just a technology, rather it is a concept that structures the IT of a company according to services,
which are modular and can be combined flexibly to implement IT processes. The IT-specific components of the
concept are realized with an IT architecture, which is based on loosely coupled services which are mutually inde-
pendent from a technical viewpoint and the interoperability of which is based on open standards, thus enabling
a clear separation between logic and technology. The standardized and modular structure of SOA can be directly
incorporated into tailored solution portfolios and offers many advantages:
Efficiency and short time-to-market thanks to a standardized and simple development
of new services and processes.
Reusability of IT assets and costs savings through the use of common resources.
Transparency of the IT architecture through clearly defined services.
Increased flexibility through standardized interfaces and the loose coupling of services.
Data quality encompasses the importance, relevance and correctness of information. Insufficient quality or the lack
of required starting data as well as a possible paucity of data structure can all impair the kick-off phase of a BI proj-
ect. In ongoing operation, poorly maintained data is a frequent problem. In addition to the enforcement of data
maintenance standards (clear policies), data quality should also be improved through automatic quality checks and
the implementation of a data quality assurance process. This is where Master Data Management (MDM) comes in-
to play. A good MDM system that is planned early on produces clearly defined master data and is thus the basis for
high data quality. In addition, by stipulating clear policies it must be possible to clearly regulate who must provide
what data and the quality and format of such data (data governance). Data governance is becoming more and
more important, particularly given the ever increasing data quantities, and forms the basis for defining clear respon-
sibilities. Accordingly a specific area/specific department must be uniquely responsible for clearly defined busi-
ness processes and the resultant data. Only such a regulation will guarantee consistency in the data. A continuous
integration and consolidation of data sources serves not only to create a company-wide view but also improves da-
ta quality, as the different data sources often only hold the individual values in very different formats. The simplest
example is the date format (e.g 25 Feb. 74, 25.02.1974 etc.). In addition, the data structure of operative systems dif-
fers widely from the structure of the multidimensional data cubes presented above.
11
“Service-oriented architectures increase the required adjustability
of a BI solution and strengthen the separation between technology
and logic, through the use of clearly defined services for representing
individual business processes.”
!
12
“Missing or faulty data reduces the benefits of BI solutions, promotes
incorrect decisions and makes it more difficult to adhere to legal frame-
work conditions.”
!
A further important aspect of MDM is the issue of data security. A lack of security will lead not only to rejection
by the majority of users, but also makes unauthorized access to valuable data much easier. This can jeopardize
the competitivity of individual departments and even the entire company. For this reason, data security must
be a significant issue in the definition of a BI program. The central elements of a high security standard are
contained in the definition of clear data security strategies as well as the consideration of the most modern se-
curity tools during implementation. Through a continuous further development of the security processes intro-
duced, the achievement of hopefully high security standards is guaranteed.
MDM plays a vital role in achieving a high level of transparency in the implementation of BI projects, despite the
complexity of the technology. The success of BI is predicated on the consistent definition and implementation
of rules to ensure data quality and security that includes a clear description of responsibilities. The implementa-
tion of MDM policies also requires a detailed procedural model, in order to set up the planned BI solutions and
the relevant business processes in accordance with the specific customer challenges.
4. Conclusion & outlook.
13
“The decisive factor for future business success will be the ability to iden-
tify as quickly as possible from amongst the gargantuan mass of data
the correct and important information, irrespective of its data structure.”
!
“This is merely the dawn of the information age. When one thinks that
information technology is only a very young science, and when one
considers the explosion in the volume of data that we have witnessed
in recent years, it is easy to see that the importance of information will
continue to grow in this fast-paced environment.”
!
At the heart of Business Intelligence is the wish to identify as quickly as possible and to better understand the
cause-and-effect relationships and mechanisms that are of relevance to one’s own business by analyzing exist-
ing data. Business-relevant information must therefore be made available in real time and in a comprehensible
format; this data must be of a high quality and be up-to-date. On the one hand, users should be able to easily ac-
cess the information, while on the other the data must be protected against unauthorized access. If it is imple-
mented consistently, a Business Intelligence system that is integrated into the operative, tactical and strategic
business can make a huge contribution to the success of the company.
The desire for more findings that are available faster and are above all relevant is pitted against the increasing
flood of data and the ever more complex data analyses this entails. This is a continuous development, which
is strengthened not the least by the merger of IT and TK. New, innovative ICT solutions and concepts such as
Medical Card, Location Based Services, Homeland Security, etc. are generating data quantities in the petabyte
range, and in this environment it is not only the number of structured information units that is rising but also the
quantity of unstructured and semi-structured information. This includes e.g. documents, emails, tables etc.,
which are saved in a disparate fashion, but in which the information is not clearly structured.
According to an Ovum study, approx. 80-85% of a company’s information, i.e. by far the greatest share of knowl-
edge, is poorly structured or entirely unstructured even today. Consequently it is necessary not only to link the dif-
ferent concepts and systems more closely together but above all to gain control over the integration and analysis
of structured and unstructured data stocks. This is because the anticipated explosion of the Internet through
Web 2.0 and Web 3.0 is of late leading to a further exponential rise in unstructured and semi-structured information,
which will bring with it a new, even larger and nearly overwhelming flood of information.
In the future, BI concepts must merge with Information Management concepts, the approach of which is to efficiently
enable both access to and analysis of structured and unstructured information. It is critical for the acceptance
of future solutions that data is made available in real time, that it is specific to the users and above all that it is clear.
The users should not, however, come into contact with the complexity of the data sources and structures under-
lying such solutions – to avoid, inter alia, excessive demand.
Only companies that succeed in identifying, analyzing and preparing appropriately the critical data as quickly as
possible and which can thus identify the changes and trends at an early stage, allowing them to react as quickly
as possible, will be able to compete at the highest levels in the long term. Consequently, innovative strategies,
concepts and technologies from Information Management, which allow the right information to be obtained reli-
ably and independent of location, data source and format in real time, and thus form the basis to be able to make
fast and intelligent decisions, are important.
5. Glossary.
Balanced Scorecard Performance measurement technique, which allows corporate management to
measure the various internal functions and its results against the outside world.
Organizations use this to achieve strategic goals which are often divided into four
fields (finances, processes, customers/market, personnel).
Basel II Basel II is the title used to cover all the equity capital regulations proposed by the
Basel Committee on Banking Supervision in recent years. Since January 1, 2007
these must be applied to all credit and financial services institutions in the mem-
ber states of the European Union.
Business Intelligence Business Intelligence describes a collection of methods, which can be used to gain
valuable information from structured data.
Change Management Change Management covers all tasks, measures and activities, the aim of which
is to bring about a comprehensive, cross-group and far-reaching change – to
implement new strategies, structures, etc. – within an organization.
CRM Customer Relationship Management (CRM) describes procedures and techniques
with which the relationship between customers and suppliers can be represented.
Cross-Selling Cross-Selling is the term used in marketing to designate the sale of supplementary
products and services.
Customer Lifetime Value Customer Lifetime Value (CLV) is, generally speaking, the current value of the likely
future income stream generated by a customer over his entire “customer life”.
Dashboard Dashboard designates a visualization form for information in compressed format.
Data Mart In contrast to a data warehouse, a data mart is not a company-wide database.
Data marts are restricted to parts of companies, e.g. departments, groups, prod-
uct divisions and are therefore a good means of implementing BI solutions
in a short space of time, in accordance with a modular Start Small approach.
Data Mining Data mining is a special form of data analysis which reveals hidden trends. In ad-
dition, analysis tools are used to discover invisible patterns and to validate known
factors.
Data consistency Data consistency for databases is, generally speaking, the lack of contradictions
in data.
Data Warehouse (DW) A data warehouse is a database, isolated from the operational DP system, which
serves as a company-wide database for all system versions to support manage-
ment. A data warehouse is, among other things, the basis (enabling technology)
for CRM, Data Mining, E-Business etc.
14
ERP The term Enterprise Resource Planning describes the corporate task of deploying
the resources (capital, equipment or personnel) available in a company efficiently
for the purpose of the operational processes.
Governance Governance is, in general terms, the control and regulation system for the structures
(setup and process organization) of a company, organization or also an operation.
IFRS The International Financial Reporting Standards (IFRS) are international accounting
regulations.
IT Governance IT governance comprises management, organizational structures and processes
that ensure that the IT supports the corporate strategy and goals.
IT Outsourcing Transfer of full responsibility for IT functions with a high IT share to legally inde-
pendent – i.e. external – service providers for a defined period.
KonTraG The Gesetz zur Kontrolle und Transparenz im Unternehmensbereich (law providing
for control and transparency in the Group), KonTraG for short, is a comprehensive
Artikelgesetz, a composite act covering the amendment and enactment of the re-
lated legislation, the aim of which is to improve corporate governance in German
companies.
OLAP Online Analytical Processing (OLAP) is the analysis and evaluation of multidimen-
sional data, to gain information for company decisions.
SaaS Software as a Service (SaaS) is software the use of which is granted to a customer
as an ongoing service for a usage fee (“rent”) that must be paid periodically.
Sarbanes-Oxley Act The Sarbanes-Oxley Act (SOX) is a US capital market law enacted in 2002, accord-
ing to which all companies listed on US stock exchanges must have their internal
control systems overseen, documented and tested by auditors.
SOA Service Oriented Architecture (SOA) is a management concept and requires a system
architecture concept as a secondary concern only. This concept aspires to an infra-
structure which is aligned to the desired business processes and which can react
quickly to changed requirements in the business environment. The system architec-
ture concept provides for the provision of technical services and functionalities
in the form of services that represent nuclear process steps.
US-GAAP US-GAAP is the abbreviation for the United States Generally Accepted Accounting
Principles. It refers to the US accounting regulations that govern the accounting
practices and annual financial statements of companies.
Source: According to the studies, articles, et al documented in the bibliography.
15
6. Bibliography.
Sascha Alexander, “Business Intelligence: Was Unternehmen wirklich brauchen” [Business Intelligence: What
Companies Really Need”], Computerwoche, 2007
Dr. Michael Böhnlein, “Todsünden des Data Warehouse” [Seven Deadly Sins of the Data Warehouse], 2007
Dr. Michael Böhnlein , “Stolpersteine bei BI-Projekten und aktuelle BI-Trends” [Stumbling Blocks of BI Projects
and Current BI Trends], 2007
Barney Finucane, “Vor- und Nachteile neuer Business-Intelligence-Ansätze” [Advantages and Disadvantages
of New Business Intelligence Approaches], CIO, 2007
Dr. Peter Gluchowski, “Quo vadis Business Intelligence?”, BI-Spektrum, 2006
Knut Hildebrand (Hrsg.), “Business Intelligence”, HMD-Praxis der Wirtschaftsinformatik, 2001
Dr. Thoms Jurisch, “Business Intelligence – Schneller und besser als die Wettbewerber” [Business Intelligence –
Faster and Better than the Competition], 2007
Dr. Joachim Philippi, “SOA trifft auf BI” [SOA meets BI], CIO, 2007
Dr. Joachim Philippi, “Die sieben größten BI-Fehler und wie man sie vermeidet” [The Seven Greatest BI Errors
and How to Avoid Them], Computerwoche, 2005
Dr. Joachim Philippi, “Weltweiter Markt für Business Intelligence wächst” [Global Market for Business Intelligence
is Growing], Computerwoche, 2007
Dr. Joachim Philippi, “Deutsche Vertriebsmitarbeiter haben ein gespaltenes Verhältnis zu Business Intelligence”
[German Sales Employees Have an Ambivalent Attitude to Business Intelligence], Computerwoche, 2007
Dr. Joachim Philippi, “Die Lösung heißt Maßanzug” [The Solution is a Custom-Made Suit], Orbit, 2007
Dr. Joachim Philippi, “BI-Strategie – Der Weg zum maximalen Return-On-BI-Invest” [BI Strategy – The Road
to the Maximum Return-On-BI-Investment], TDWI, 2007
Dr. Joachim Philippi “White Paper Dynamic Services – Flexible ICT-Ressourcen” [White Paper Dynamic Services
– Flexible ICT Resources], T-Systems 2007
Dr. Joachim Philippi, “White Paper Real ICT – Flexibilität durch ganzheitlich integrierte IT- und TK-Services”
[White Paper Real ICT – Flexibility through Comprehensively Integrated IT and TK Services], T-Systems, 2007
Dr. Joachim Philippi, “White Paper Identity und Access Management (IAM)” [White Paper Identity and Access
Management (IAM)], T-Systems, 2007
Andreas Schaffry, “Die richtige Strategie bei BI-Projekten” [The Right Strategy for BI Projects], CIO, 2007
Kurt Schlegel, “Business Intelligence Trends and Best Practices”, Gartner, 2007
Markus Trottnow, Markus Bereszewski: “Den Wandel intelligent gestalten” [Making the Change Intelligently],
Information Week, 2007
Rainer Volck, “Business Intelligence in Echtzeit” [Business Intelligence in Real Time], 2007
Nicolas Zeitler, “Mangelnde Ausrichtung an der Unternehmensstrategie – Firmen planen BI-Projekte oft
schlecht” [Lack of Alignment to the Corporate Strategy – Firms Often Plan BI Projects Poorly], CIO, 2007
16
Publisher:
T-Systems Enterprise Services GmbH
Mainzer Landstr. 50
D - 60325 Frankfurt
Responsible for the contents:
Sales Management BI & Performance Management
Contact:
T-Systems Enterprise Services GmbH
Corporate Marketing & Communications
Uli Kunesch
Fasanenweg 5
D- 70771 Leinfelden-Echterdingen
E-mail: [email protected]
T-Systems Business Services GmbH
Sales Management BI & Performance Management
Franz Pfitzner
Dachauer Str. 665
D- 80995 München
E-mail: [email protected]
doc_742088429.pdf