Lila Rao-Graham (Author)_ Maurice L. McNaughton (Author)_ Gunjan Mansingh (Author) - Business Intelligence for Small and Medium-Sized Enterprises_ an Agile Roadmap Toward Business Sustainability-Auerb

Business Intelligence for
Small and Medium-Sized

Enterprises
Business Intelligence for
Small and Medium-Sized
Enterprises
An Agile Roadmap toward Business
Sustainability
Lila Rao-Graham, Maurice L. McNaughton,

and Gunjan Mansingh
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2020 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
Printed on acid-free paper
International Standard Book Number-13: 978-1-138-58421-1 (Hardback)
This book contains information obtained from authentic and highly regarded sources.
Reasonable efforts have been made to publish reliable data and information, but the
author and publisher cannot assume responsibility for the validity of all materials or the
consequences of their use. The authors and publishers have attempted to trace the copyright
holders of all material reproduced in this publication and apologize to copyright holders if
permission to publish in this form has not been obtained. If any copyright material has not
been acknowledged please write and let us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted,
r eproduced, transmitted, or utilized in any form by any electronic, mechanical, or other
means, now known or hereafter invented, including photocopying, microfilming, and
recording, or in any information storage or retrieval system, without written permission
from the publishers.
For permission to photocopy or use material electronically from this work, please access
www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance
Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-
for-profit organization that provides licenses and registration for a variety of users. For
organizations that have been granted a photocopy license by the CCC, a separate system of
payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks,

and are used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com
and the CRC Press Web site at

http://www.crcpress.com
The measure of intelligence is the ability to change
Albert Einstein
Contents
F o r e w o r d xi
A u t h o r s xv
I n t r o d u c t i o n C o n c e p t i o n s of Agilit y and Business
Intelligence for S ME s 1
Digital Business Trends 1
Conceptions of Agility 2
Business Intelligence as an Enabler of Agility 3
SMEs, Business Intelligence and Agility 4
About this Book 5
Pa r t I B
I L a n d s c a p e – O p p o r t u n i t i e s for SME s
C h a p t e r 1 B a r r i e r s a n d S t r at e g i e s f o r E n t e r p r i s e
ICT A d o p t i o n i n S ME s 9
Introduction (Digital Economy – Implications for Business) 9
SMEs and ICT in Developing Economies 11
SMEs and ICT Adoption – Barriers & Challenges 11
BI – Value Opportunities for SMEs 13
Distribution and Retail 13
Credit and Micro-Finance Services 14
C h a p t e r 2 A n A g i l e I n t e g r at e d M e t h o d o l o gy f o r
S t r at e g i c B u s i n e s s I n t e l l i g e n c e ( AIMS - BI) 17
Introduction 17
Why the Need for New Methodologies? 19
vii
viii C o n t en t s
The Need for a Strategic, yet Agile Perspective 19

Know the Current State of your Information Maturity 21
End User and Key Stakeholder Engagement 22
Design of AIMS-BI 22
Description of AIMS-BI 23
Step 1: Information Maturity Assessment 24
Step 2: BI Opportunity Discovery 26
Step 3: BI Portfolio Evaluation 27
Step 4: Proof of Concept Prototypes 28
Strategic Business Intelligence Roadmap 29
Findings from Applying AIMS-BI 30
Conclusion 31
Pa r t II N av i g at i n g the A g i l e BI P r o c e s s
C h a p t e r 3 I n f o r mat i o n M a n a g e m e n t (IM) M at u r i t y
A s s e s s m e n t : E va l uat i n g B u s i n e s s
C a pabi l i t i e s a n d G a p s 37
Introduction 37
Maturity Models in Organizations 39
Implementing the IM Maturity Model 40
Applying the AIMS-BI IM Maturity Assessment 43
Example of the Application of IM Maturity Assessment 43
Conclusion 48
C h a p t e r 4 C r e at i n g BI P o r t f o l i o s 49
Introduction 49
The Discovery of BI Opportunities 49
BI Portfolio Evaluation 51
Example: BI Opportunity Discovery and Portfolio Evaluation 53
Conclusion 56
C h a p t e r 5 Th e P r o c e s s a n d Va l u e o f B u i l d i n g
P r o o f - o f - C o n c e p t P r o t o t y p e s 59
Introduction 59
Why Develop PoCs? 60
BI – The Different Layers of Analytics 61
Monitoring Layer 62
What Is Data Visualization? 62
Data Visualization Process Models (VPM) 62
Understanding the Business Context 63
Get Data 64
Visualization Design Considerations 65
How Will the User Respond to the Visualization? 65
How Will the Visualization Be Used? 65
How Should the Data Be Encoded? 66
Building Visualizations 67
C o n t en t s ix
Business Insights 67
Decisions and Actions 68
Prediction Layer 68
What Is Data Mining? 68
IKDDM Process Model for Data Mining 70
Business Understanding Phase 70
Data Understanding 70
Data Preparation 70
Modeling (Data Mining) 71
Evaluation 71
Deployment 72
Lessons Learned from Applying the Process Models to the
Development of the PoCs 72
Conclusion 73
C h a p t e r 6 D ata G o v e r n a n c e a n d D ata Q ua l i t y
M a n a g e m e n t 75
Introduction 75
Data Governance and Data Quality 76
Data Quality Management Process 79
Stage 1: Defining Data Standards Quality Metrics 79
Stage 2: Data Quality Assessment 81
Stage 3: Data Maintenance/Stewardship 85
Stage 4: Data Life-Cycle Process Review 88
Conclusion 88
C h a p t e r 7 D ata I n t e g r at i o n : M a n a g i n g D ata S o u r c e s
f o r A g i l i t y 91
Introduction 91
Opportunities through Data Integration 92
Key Deliverables from Realizing Integration Opportunities 93
Common Approach to Data Integration 94
Possible Integration Approaches 94
Examples of Data Integration Techniques 95
Physical Integration 95
Virtual Integration 96
Physical vs. Virtual Integration 96
Selecting a Data Integration Approach 98
Recommendations for SMEs 100
Conclusion 101
Pa r t III A B l u e p r i n t for Action

C h a p t e r 8 D e v e l o p i n g a R o a d ma p f o r S t r at e g i c
B u s i n e s s I n t e l l i g e n c e 105
Introduction 105
x C o n t en t s
Information Management (IM) Maturity Assessment

(Capability Gaps) 106
Proof of Concept (PoC) Prototypes 107
Data Quality 107
Data Integration 108
Architecture of a Business Intelligence (BI) Roadmap 109
Performance 110
People 110
Process 112
Platform 112
Example Roadmap – Portfolio of BI Initiatives 113
Navigating the BI Roadmap 115
Conclusion 115
C h a p t e r 9 C r e at i n g B u s i n e s s Va l u e f r o m
D ata A s s e t s 117
Introduction 117
Measuring Value of your Data Assets 118
Method of Information Value Assessment 119
Managing the Value of your Data Assets 121
Benefits Realization Management (BRM) 122
Conclusion/Recommendations 124
E p i lo g u e: L e s s o n s L e a rn e d 125
A p p e n d i x : C a s e S t u d i e s 129
G l o s s a r y 135
B ib l i o g r a p h y 139
I n d e x 145
Foreword
Drs Lila Rao-Graham, Gunjan Mansingh and Maurice L.

McNaughton, my illustrious coconspirators in the quest for excellence
in the p rovision of high quality and effective Information Systems
applications, particularly for small and medium enterprises, have
undertaken an enormously important assignment: the writing of a
book on Business Intelligence (BI) with particular emphasis on its
utilization in Small and Medium–sized Enterprises (SME’s), target-
ing owners, d ecision-makers, and workers as well as academics and
students. It is a project that I am extraordinarily proud to be asso-
ciated with for several reasons, but particularly because of my long
association with all three authors, as former protégés who are now
full-fledged colleagues, confidantes and collaborators in the quest to
advance knowledge in the often-discussed, but not well-understood
subject area of the application of Business Intelligence (BI) in SME’s.
The application of BI in SMEs requires a solid understanding of the
nuances as well as the peculiarities of these enterprises in comparison
to their larger counterparts. It is an understanding that the authors
have amply demonstrated in this exceptional publication.
But I can think of no better group to address the peculiar issues
that are associated with this demanding undertaking than this super
experienced team of Drs Rao, Mansingh and McNaughton. They
have invested heavily in a deep understanding of these issues and
xi
x ii F o re w o rd
have spent a large part of their scholarly pursuits studying, assimilat-

ing and disseminating knowledge in the areas of Business Analytics
and SMEs. This is not a declaration I make lightly; I have observed
their impressive work in this area close up, with great admiration.
Dr Rao has undertaken extensive work in BI, Data Quality and
Decision Support Systems with particular focus on assisting orga-
nizations leverage the value from data through effective decision-
making. Dr McNaughton runs UWI Mona’s Centre of Excellence for
IT-enabled Business Innovation with particular emphasis on Small
and Medium Enterprises and Dr. Mansingh has focused her research
and interest largely on Business Intelligence, Data Mining, Decision
Support Systems and Knowledge Management.
The notion of BI has evolved over several years as organizations
have extended their online transaction processing (OLTP) capa-
bilities and applications to support their routine operations and
established the capability to extract internal and external data from
a variety of sources to specifically obtain intelligence about non-
routine and often less-structured arrangements, through Online
analytical processing (OLAP). BI therefore refers to applications and
technologies that are used to gather, provide access to, and analyze
data and information about the operations of an organization, with the
capability of providing comprehensive insights into the more volatile
factors affecting the business and its operations, thereby facilitating
enhanced decision-making quality and contributing to the creation
of business value. Larger and more sophisticated organizations have
long been e xploiting these capabilities. This work seeks to provide
guidance to SMEs to replicate this experience and is the unique con-
tribution of Business Intelligence for SME’s: An Agile Roadmap toward
Business Sustainability.
The authors have done an excellent job in explicating this evolution
and how the various terms we have used along the way to describe the
migration toward the exceedingly important organizational notions
and use of the data and information. They possess the skills and expe-
rience to not merely seek out operational efficiencies but, on a hugely
different magnitude, to extract intelligence from the information
and data available to organizations. They are quick to point out that
successful BI implementations have generated significant increases
in revenue and cost savings, however, the failure rates are also very
F o re w o rd x iii
high. But more importantly the emphasis of this p ublication (which is

the huge selling point) is that these capabilities are not the exclusive
purview of large organizations, but may yet become the saving grace
of SMEs. They have begun to make extensive use of these tech-
niques to ensure the kind of agility that will endow them with the
organizational capability to sense and respond to opportunities and
threats in an increasing dynamic business environment – an envi-
ronment in which smaller organizations could have an increasingly
larger role by establishing, among other capabilities, the agility to
leverage internal and external data and information assets to enhance
competitiveness by having a comprehensive understanding of a
process that will ultimately be the “key” to an agile roadmap for busi-
ness sustainability – the authors have emphasized this beautifully
throughout this publication.
Drs Rao, Mansingh and McNaughton underscore the point that,
in the not too distant past, there was the misapprehension that BI is
for large, resource-intensive enterprises, however, this is no longer the
case. SMEs have a natural latent agility that can only be enhanced by
the appropriate application of BI. In addition, they have introduced
a process model born out of their experiences and insights with real-
world examples that is a must read for SMEs. This model, and the
critical success factors for their implementation, can be adapted by
both practitioners and researchers, in their quest to acquire greater
agility through strategic BI. The authors, recognizing the fact that
several organizations have invested in the BI tools and technologies
without the appropriate focus on the prerequisites and processes that
are required to maximize the benefits, present a process to assist orga-
nizations in harnessing the returns from their BI investments. They
also provide recommendations to assist SMEs in identifying and
implementing their BI projects; not in a generic manner but condi-
tioned by the peculiarities of their focused needs and capabilities.
Traditional BI approaches have often been considered unsuitable
for SMEs and were once positioned as the strategic apparatus of larger
and more resourced organizations. The authors effectively contradict
this fiction and identify the several opportunities and approaches that
are available to SMEs in this domain and provide real and excellent
examples of how these organizations can identify, understand and
apply appropriate BI approaches to their peculiar benefit.
xiv F o re w o rd
The book is tremendously well organized. It effectively introduces

the subjects of BI and the concepts of agility and their importance and
traces the evolution of the subject area. Commendably, it provides an
extremely useful discourse on strategies for enhancing, as well as bar-
riers to, the adoption of ICT in SMEs. It provides an excellent (must
read) overview of Agile Process Model for Strategic BI, and then
tackles the important issue of how to assess Information Maturity.
The discussion of how to evaluate business capabilities and gaps in
strategic BI is simply excellent. There is also a must-read explication of
important aspects of BI, including the exceedingly important notions
of data governance, data integration and data services and quality
management. This unusual volume closes with two extraordinarily
important discussions: how to develop a roadmap for strategic BI and
how to measure the success of BI initiatives.
This book is a gem! It is a conclusive treatise that BI and indeed
data analytics are extraordinarily significant artilleries for SMEs
and certainly not the exclusive apparatus of the corporate world.
The authors have convincingly presented a superb case for the use of
these powerful techniques that have the capability to assist smaller
organizations to reap the enormous BI benefits that have long accrued
to their corporate counterparts; allowing them to take targeted action
to improve decision-making about their operational performance,
reduce risk and identify opportunities for improvement, and plan for
the future.
Evan W. Duggan, Ph.D.
Former Visiting Professor at the University
of Alabama in Birmingham
Former Dean of the Faculty of Social Sciences and
Professor of MIS at UWI Mona, Jamaica
Authors
Lila Rao-Graham is the deputy executive director and a senior

lecturer in Information Systems at the Mona School of Business and
Management, The University of the West Indies (UWI), Jamaica.
She holds a PhD in Information Systems from The University of the
West Indies. She lectures courses in Database Management Systems,
Enterprise Data Management and Business Analytics.
er research interests include Business Intelligence, Data Quality.

H
Knowledge Management and Decision Support Systems.
Maurice L. McNaughton is the director of the Centre of Excellence

for IT-enabled Innovation at the Mona School of Business and
Management, The University of the West Indies (UWI), Jamaica.
He holds a PhD in Decision Sciences from Georgia State University.
He lectures courses in Modelling and Decision Support Systems; IT
Economics; IT Governance and Strategic use of ICT.
is research interest spans the domain of emerging Open ICT ecosys-

H
tems, and includes Open Source Software, Open/Big Data, Mobile
and Cloud Computing.
xv
xvi Au t h o rs
Gunjan Mansingh is head of the Department and a Senior Lecturer

at the Department of Computing, The University of the West Indies
(UWI), Mona, Jamaica. She holds a PhD in Information Systems
from The University of the West Indies. She lectures courses in
Business Intelligence, Programming, Artificial Intelligence and
Knowledge Discovery and Analytics.
er research interests include Business Intelligence, Data Mining,

H
Machine Leaning, Decision Support Systems, Knowledge
Management and Expert Systems.
I ntroducti on
Conceptions of Agility and Business
Intelligence for SMEs
In the new world, it is not the big fish which eats the small fish,
it’s the fast fish which eats the slow fish.
Klaus Schwab
Digital Business Trends
Digital is rapidly becoming the dominant mode of interaction for

commercial, social and economic activity, resulting in the emergence
of what is often referred to as the Digital economy. Data has been
described as the oil that fuels the Digital economy and an organiza-
tion’s ability to collect, manage, use and generate insights from Data for
a variety of business scenarios and activities has become an essential
capability. With the increasingly digital nature of business interac-
tions, the ability of organizations to exploit distinctive data assets and
analytics capabilities presents enormous opportunities for increased
operational efficiency, greater customer intimacy and improved ser-
vice delivery and will be critical to their ability to utilize information
as a competitive weapon.
Davenport and Harris (2007) describe many exemplars of orga-
nizations that make widespread use of business analytics and data
assets to create significant value for them, allow them to leapfrog over
competitors who possess more traditional advantages of size, finan-
cial assets and physical infrastructure. The well-known example of
Netflix, which began as a start-up founded on the premise of using
analytics as the basis of competition and was able to dominate and
eventually clobber the much larger, more established Blockbuster,
should inspire many small businesses to embrace Klaus Schwab’s oft-
cited mantra: “In the new world, it is not the big fish which eats the small
1
2 BUSINE S S IN T EL LI G EN C E F O R SM E s
fish, it's the fast fish which eats the slow fish”. In other words, size doesn’t
matter, agility does!
Conceptions of Agility
Being Agile has become a recurrent theme in the modern business dis-
course, and many executives now see this as an increasingly important
organizational capability necessary to be able to compete effectively
in the Digital economy. Business agility is generally defined as “the
capacity of firms to sense and respond readily to opportunities and threats in
a dynamic business environment”. However, there are several identifi-
able dimensions to Agility within the organizational context:
Firstly, Agility has its roots in the Software Development
discipline, where the Agile Manifesto prescribes the following key
principles to improve success rates in project implementation:
• Incremental delivery: Deliver functionality in increments with
the involved customer validating and specifying the require-
ments to be included in each increment.
• People not process: Empower the project team and allow team
members to develop their own ways of working without
prescriptive processes.
• Embrace change: Expect the business requirements to evolve
and design the solution to accommodate changes.
• Maintain simplicity: Focus on simplicity in both the product
being developed and in the development process. Wherever
possible, actively work to eliminate complexity from the system.
These practices are not exclusive to software development, but are
generally applicable to any kind of business activity.
Secondly, modern Human Resource practitioners increasingly
embrace Agile organizational behaviors such as:
• Encouraging a fail-fast, learn-fast culture.
• Talent recruitment through social media platforms such as
LinkedIn and Facebook.
• Flexi-work policies that allow employees to “work when, how,
where you want”.
• Building a workplace environment and culture to attract
millennials and digital natives.
AGILIT Y A N D BI F O R SM E s 3
• Adopting a self-service paradigm for human resource services

delivery.
Thirdly, an increasing number of digital transformational initiatives
are beginning to manifest themselves in organizations that are seek-
ing to increase Operational Agility by employing modern technology
platforms to digitize core enterprise work infrastructure and processes
such as customer-facing channels, order fulfillment, supply chain
relationship management and product and/or service innovation.
Ultimately, business agility is seen as a leadership imperative and
requires executives to:
• Embrace dynamic, adaptive business models;
• Exhibit entrepreneurial alertness that enables the firm to
explore its marketplace, detect areas of marketplace ignorance
and seize on opportunities for action;
• Adopt a visionary perspective through strategic foresighting
and analytic-driven insight;
• Lead and inspire the organizational culture change required
for digital transformation.
Organizations that aspire to this kind of transformational culture,
aptitude and agility will need to develop the analytic orientation
and business intelligence capabilities similar to those of the Analytic
Competitors described by Davenport and Harris (2007).
Business Intelligence as an Enabler of Agility
Business Intelligence, Data Analytics, Business Analytics and Data

Mining are often used interchangeably in the academic and trade lit-
erature as well as by practitioners. For the purposes of this book, we
treat Business Intelligence (BI) as an umbrella term for a wide range
of processes, tools, techniques and capabilities whereby organizations
create business value from their data assets.
This definition of Business Intelligence offered by Forrester (2014) is:
…an approach that combines processes, methodologies, organizational
structure, tools, and technologies that enable strategic, tactical, and
operational decision-makers to be more flexible and more responsive to
the fast pace of changes to business and regulatory requirements.
This definition situates BI as an organizational capability that enables

business agility. BI has become a strategic initiative and is now rec-
ognized by business and technology leaders as instrumental in driv-
ing business effectiveness and innovation. Companies that have been
successful in BI implementations have seen their investments in BI
generate increases in revenue and produce cost savings that realize
significant returns on their investment (ROI). However, many other
organizational BI initiatives struggle through a myriad of technol-
ogy platforms, reporting tools and localized data sources that fail to
deliver the consistent information required to drive value-added and
timely executive decision-making.
The design and implementation of the classical data warehouse
architecture that underpins many BI Systems can be technically and
financially daunting for organizations without the requisite technical
resources. Smaller organizations in particular are often challenged by
the complexity and resource requirements of traditional BI solutions.
SMEs, Business Intelligence and Agility
Does this mean that the potential operational efficiency improve-

ments, customer intimacy and competitive benefits associated with BI
and business analytics have become even more inaccessible to small
businesses that are limited by resource and expertise constraints?
Definitely not! Indeed, we argue that perhaps more so than any
other enterprise information system, BI gives SMEs the opportunity
to combine business agility with their natural flexibility, to create a
distinctive competitive advantage that typically eludes larger organi-
zations confined by more formal structures and bureaucracy.
The importance of the small and medium enterprise (SME) sector
as a principal contributor to the economic well-being of most societies
is acknowledged for both developed and developing economies. SMEs
are often both the product and the producers of entrepreneurial cre-
ativity and, in most countries, have become progressively important
pillars of economic stability. Their actual and potential contributions
to innovation, economic growth, employment and the diversification
of productive endeavors are well documented.
Notwithstanding the acknowledged importance of the SME sec-
tor, many such organizations face a variety of challenges contingent
AGILIT Y A N D BI F O R SM E s 5
on their size, when contemplating the implementation of advanced

ICTs such as BI. These inherent characteristics often include:
• Limited technology competencies and awareness of strategic
ICT value opportunities.
• Lack of the implementation skills and absorptive capacity to
effectively deploy ICT-enabled solutions.
• Typically, a single owner-manager who is the principal
decision-maker, burdened with multiple responsibilities.
• Heavy dependency on external technology advice through
intermediaries or vendors.
As a result, traditionally, SMEs have not done well in the adoption
of Enterprise ICTs as enablers of business innovation often limiting
their efforts to basic operational and transactional ICT systems.
Fortunately, the availability of modern cloud computing
infrastructure and previously prohibitive (due to cost and complex-
ity) enterprise applications through software-as-a-service have made
a range of modern ICT capabilities accessible and affordable to even
the smallest organizations. Small business innovators can now access
and use the most advanced BI solutions for a modest monthly expense
and thus give themselves a fighting chance in the globally competitive
business landscape.
While access to the technology addresses an important barrier, it
alone is insufficient to harness the benefits of BI. SMEs need a sys-
tematic, yet practical approach that allows them to discover value-
adding opportunities for charting their own customized roadmap
toward realizing the strategic benefits of BI. That is the underlying
premise for this book.
About this Book
This book was conceptualized out of the authors’ combined years of

experience in research, education and consulting with large and small
organizations in the application of ICTs to enable business innova-
tion. These engagements have forged a conviction that small busi-
nesses today have an unprecedented opportunity to embrace and
harness modern ICTs, in particular, BI, to unleash and amplify the
natural entrepreneurial spirit and nimbleness of their businesses.
The book is based on the Agile Integrated Methodology for Strategic

Business Intelligence (AIMS-BI), a highly adaptable process model for
implementing BI that was formulated based on the authors’ collective
experiences. It provides SME business and technology managers with
a practical guide and supporting toolkit for the effective implementa-
tion of enterprise business intelligence systems.
The book is structured in three sections:
Part I: BI Landscape – Opportunities for SMEs, examines the
evolving digital economy, implications for business and the
role that BI and analytics can play in helping SMEs to navi-
gate the emerging opportunities and challenges. It introduces
the AIMS-BI methodology as a practical guide to navigate
this landscape.
Part II: Navigating the Agile BI Process, describes the stages and
activities in the AIMS-BI methodology and details not just
what should be done but also how to do it. The procedures are
supported by examples and recommendations for less resource
intensive options that make BI generally accessible for SMEs
with limited financial and technical capacity.
Part III: A Blueprint for Action, guides the formulation of a strate-
gic BI roadmap that provides business decision-makers with a
clear, coherent implementation path that can be executed and
managed within the constraints of the organization’s available
resources. It includes recommendations and provides advice
on the critical success factors for BI implementations.
For SMEs to survive and compete globally with larger, more well-
resourced organizations, business agility is an essential capability.
SMEs have an opportunity to combine agility with their natural
flexibility as a distinctive competitive advantage that typically eludes
larger organizations confined by more formal structures and bureau-
cracy. Recent history is replete with evidence of smaller, more agile
organizations that have leveraged business analytics to outcompete
and overtake larger incumbents. This book will provide an essential
tool in the arsenal of SME business and IT managers to enhance their
chances of competing effectively and sustaining their existence in the
Digital economy.
Part I
BI L andscape –
O pport u nities
for SME s
Business Intelligence (BI) has become a strategic initiative for many

organizations. Successful BI implementations have generated signifi-
cant increases in revenue and cost savings. However, the failure rates
are also very high. One of the reasons for these high failure rates is
that the traditional BI methods are not suited to all organizations;
they are seen by some as being too rigid and resource intensive.
Additionally, the focus of the analytics is often at the localized level
rather than on developing enterprise-wide analytical capabilities. Due
to the resource intensity and rigidity of the traditional methods Small
and Medium-Sized Enterprises (SMEs) have not embraced BI as an
opportunity for them to compete with some of the larger players in
their field using agility as their key advantage. Part I provides the
context of the book by describing the BI landscape and identifying
the opportunities that BI provides for SMEs especially as it relates to
their ability to be more agile in their execution of BI.
Chapter 1 focuses on SMEs, what is unique about them and why
there is a need to consider ICT adoption specifically for them. A gen-
eral overview of SMEs is provided and then the particular barriers
they face in terms of ICT adoption (e.g. lack of technology, expertise
and knowledge) are discussed in more detail. Whatever the barriers, it
must be recognized that there are also several opportunities for SMEs
to use these technologies to their advantage and, by so doing, build
a competitive advantage. However, these SMEs must r ecognize that
7
traditional ICT technologies and methods may not be suited to their

needs/characteristics. BI solutions are one such area where it has been
felt that the technologies are just too costly and resource intensive for
SMEs, and thus they have been slow to adopt them. This SME chap-
ter provides the context for the discussions throughout the rest of the
book as it relates to strategic BI specific to SMEs.
Chapter 2 describes a proposed methodology, an Agile Integrated
Methodology for Strategic Business Intelligence (AIMS-BI), which supports
the agility and strategic perspective needed to achieve optimal business
value from BI initiatives. Although this methodology can be applied to
any organization looking at d eveloping enterprise-wide analytic capa-
bilities, it is particularly applicable to organizations that may not be
suited to traditional BI methods due to limited resources and the need
to demonstrate value quickly, such as SMEs.
AIMS-BI provides a systematic and structured approach that
is enterprise-wide in reach yet agile in execution and focused on
organizational needs and capabilities. The final output, a strategic
BI roadmap, identifies the key BI opportunities, sets out a visible
program of implementation and positions the organization to secure
the executive commitment and resources required to maximize the BI
business value.
1
B arriers S tr ateg ies
and
fo r E nterprise ICT
A d o p ti on in SME s
A point of view can be a dangerous luxury when [used in place of]

insight and understanding
Marshall McLuhan
Introduction (Digital Economy – Implications for Business)
For some time now, business analysts and practitioners have con-
templated the onset of the Knowledge Economy in which economic
growth and competitiveness become increasingly dependent on the
capacity to create, process, accumulate and disseminate knowledge.
This includes knowledge about customers, products, processes and
competitors, or even knowledge itself as a competitive, tradable asset.
Businesses that acquired knowledge by way of closed innovation pro-
cesses or through the accumulation of knowledge assets (e.g. patents)
were able to maintain sustained competitive differentiation.
The more recent emergence of the digital economy has accelerated
the onset of the knowledge economy and changed its c ompetitive
dynamics by significantly lowering the cost of acquiring and
processing of data, information and knowledge. This change has cre-
ated opportunities for greater participation by organizations with less
resources.
Among the disruptive digital technologies that are driving this
evolution are the following:
Cloud Computing: It allows businesses to outsource their com-
puting infrastructure and access high quality servers, net-
work and application resources on demand and without the
9
traditional attendant capital and operating costs or manage-

ment responsibility. This provides a rapid deployment capa-
bility for new products and services and enhanced business
agility.
Mobile Computing: The ubiquitous mobile phone provides an
on-demand channel to customers and consumers in order to
engage, inform, educate, analyze and source/deliver digital
transactions. This enables businesses to embed mobility in the
way they design and deliver products and services.
Big Data Analytics: Incredibly large volumes of data are being
generated each day, hour, minute from a variety of sources,
including mobile data, e-Commerce transactions, social
media traffic, web clickstreams and the Internet of Things
(IoT). Access, retrieval and analysis of these powerful real-
time information flows at the point of creation is becoming
more valuable than any legacy knowledge, and provides the
basis for enhanced business intelligence.
Advances in Machine Learning and Artificial Intelligence:
Provide enhanced algorithms that decrease information
processing costs through the digitization and automation
of knowledge work. Well-designed algorithms can provide
businesses with sophisticated capabilities in areas such as
marketing campaigns, credit risk management, logistics and
e-Commerce recommender systems.
Social Media: Provides digital channels of engagement, infor-
mation and interaction with customers and consumers.
Businesses of all sizes can no longer afford to ignore social
media which increasingly shapes and reflects public senti-
ment, opinions and consensus.
Companies that are slow to adopt these advanced digital ICTs as an
integral part of their business operations and capabilities, or adapt to
the increased competitive dynamics and complexity of the way mod-
ern business is being conducted in this evolving landscape, will find
it difficult to survive. By extension, countries that fail to encourage
and facilitate broad-based ICT adoption within the local business
community will find it difficult to compete in the emerging digital
economy.
E n t e r p r i s e I C T A d o p t i o n in SM E s 11
SMEs and ICT in Developing Economies
While AIMS-BI is highly adaptable and designed to be broadly

applicable to both small and large enterprises in developed and
developing economies, the primary focus of this book is on SMEs in
developing countries.
Micro, small and medium-sized enterprises (mSMEs) are considered
to be one of the main forces in economic growth and job creation, not
only in developed economies, but also in emerging economies or econo-
mies in transition.1 Globally, close to 140 million SMEs in 130 countries
account for 65% of the total labor force (Kotelnikov, 2007). Furthermore,
the impact of SMEs in the domestic economy go well beyond employ-
ment and GDP to contribute substantially to the production and
distribution of goods and services, and the enhancement of innovation,
productivity and competitiveness (Witter and Kirton, 1990).
The drive for the increased adoption and use of ICTs in develop-
ing countries is theorized to be a catalyst that could help these coun-
tries leapfrog the stages of development undergone by more developed
nations (Steinmueller, 2001). Given the importance of SMEs to the
domestic economy of these countries, there is considerable interest
in facilitating the progressive adoption and use of ICTs by SMEs.
These ICTs can play a significant role in helping SMEs to improve
their operating performance and business competitiveness, and hence
their contribution to economic growth. Notwithstanding the obvi-
ous benefits and competitive imperative of adopting ICTs, SMEs are
often overwhelmed and unsure about the path they need to take to
realize the benefits that are being touted by experts, vendors and other
companies in the global landscape and often miss the opportunity
that these technologies can afford. As a result, the consistent adoption
and effective diffusion of advanced ICT solutions among the small
business sector in developing economies have continued to be elusive.
SMEs and ICT Adoption – Barriers & Challenges
A number of factors have been identified as barriers to ICT adoption

by SMEs and they are based on the premise that the characteristics
1 https://oecd.org/industry/smes/31919278.pdf.
of small firms are different from those of their larger counterparts.

Similarly, the factors that affect the adoption and use of ICT in large
organizations are not necessarily the same as those that affect SMEs.
Consequently, theories that explain adoption in large organizations
may not be applicable to small businesses. Small firms, in comparison
to larger firms, require different managerial approaches since there
are fundamental differences between the two. The low level of ICT
literacy and awareness of SME owners and employees, the cost of
acquiring technology, the lack of suitable financing options, the lack
of the business analytic skills, data quality issues, the lack of skills
to identify the business questions that need to be answered and the
absorptive capacity to effectively deploy ICT-enabled solutions have
all been identified as issues specific to SMEs. Generally, it is not a
trivial decision for any organization to adopt ICT; there are numerous
factors that need to be considered when the utilization of ICT is being
contemplated. This is especially so in the case of SMEs as they often
lack financial and human resources.
Several other factors, including the characteristics of the firm, the
orientation of the CEO and the type of industry have been deemed
to be influential in the ICT adoption decision of small businesses.
Constraints on financial resources and the lack of in-house IT exper-
tise increase the level of risk that an investment in IT will represent
to smaller companies. Additionally, the individual characteristics
of the CEOs as well as their attitudes to ICT and innovation and
their level of ICT knowledge are strong determinants of adoption as
they are the main decision-makers in the organization. SMEs led by
CEOs with positive attitudes toward ICTs are more likely to adopt,
particularly if they perceive that ICTs that are compatible with their
operations and comparatively easy to use will be beneficial to their
organization.
One important finding from empirical investigations into ICT adop-
tion by SMEs, is that intermediaries play a crucial role in the adoption
of complex applications by SMEs (Brown and Lockett, 2004). They
also found that the aggregation of advanced ICT solutions, such as
e-Business applications, to provide services to SMEs as a group rather
than to individuals will encourage their a doption. These insights sug-
gest that Cloud Computing and Software-as-a-Service (SaaS) which
offer fully hosted ICT application solutions, could provide a feasible
alternative for SMEs seeking sophisticated ICT services at relatively

low costs, rather than having to purchase and manage these systems
and services for themselves. Thompson and Brown (2008) further
investigated the key roles that intermediaries perform in enabling the
ICT adoption process by SMEs in developing countries. These roles
include knowledge dissemination, financial partnership, technical
advising and solution provision. The AIMS-BI methodology and the
guidelines presented throughout this book fulfill key aspects of this
facilitating role.
BI – Value Opportunities for SMEs
In our introductory chapter, we discuss our use of the term business

intelligence (BI) to represent a wide range of data-driven decision sup-
port applications including data warehousing, data mining, business
analytics, online analytic processing, performance management and
knowledge management. Collectively, BI is generally considered to be
a means of “combining data gathering, data storage, and knowledge
management with analytical tools to present complex and competi-
tive information to planners and decision makers” (Negash and Gray,
2008). Embracing the notion of “Data as an Asset” and establish-
ing an explicit Data Governance mechanism to manage, secure and
exploit the value of data have become key success factors for deriving
strategic value from Information Technology.
The stories of globally successful startups, such as Netflix,
increase awareness and provide motivation for SMEs to adopt BI,
with many senior executives clearly recognizing the strategic oppor-
tunity that BI represents to their businesses. While such celebrated
cases might be considered exceptional, the BI value opportunities
for SMEs are nevertheless considerable. Typical examples include
the following:
Distribution and Retail
Market basket analysis: It increases understanding of customers’

buying behaviors and identifies those items they tend to pur-
chase together. This knowledge can improve stocking, store
layout strategies and marketing promotions.
Sales forecasting: Examining time-based and location-based sales

patterns helps retailers to understand buying seasonality in
order to make optimal supply and stocking decisions.
Merchandise planning and allocation: When retailers contemplate
new stores or market expansion, they can improve merchan-
dise planning and allocation by examining patterns in stores
or regions with similar demographic characteristics.
Distribution logistics: Distribution companies can use low-cost
GPS data collectors with operational dashboards that enable
managers to visualize route performance and facilitate deci-
sion-making in real time. Based on continual analysis of each
driver’s data, route optimization can save significant fuel costs
and enhance delivery services for customers.
Credit and Micro-Finance Services
Credit Decision Analysis: Default and bad debt can be enormously

costly in micro-finance operations and are among the factors
that drives high lending rates. By analyzing past transactions
that were known to be fraudulent or to default, operators
can optimize and reduce the transaction costs of their credit
decisions.
Predictive life-cycle value: Repeat clients are a common feature of
micro-finance operations. Data mining can help operators to
predict each customer’s lifetime value and behavior patterns
to service segments and individuals appropriately (e.g., time-
based offers of special deals and discounts).
These are just a few examples of the kinds of data-driven analyt-
ics solutions that are no longer available exclusively to large scale
enterprises but are now accessible to small and medium-sized
companies.
The world of the artificial (man-made institutions, artifacts and
their interactions) is not getting any smaller. The rate at which data
is being generated from human social and commercial interactions is
growing exponentially. Long before Big Data Analytics and Machine
Learning Algorithms became fashionable business terms, Edward
Wilson made this perennial assertion:
We are drowning in information, while starving for wisdom. The world

henceforth will be run by synthesizers, people able to put together the
right information at the right time, think critically about it, and make
important choices wisely.
Wilson (1999)
Wilson’s “synthesizers” forecast the ascendancy of today’s Analytics

Competitors, organizations such as Google, Amazon, Capital One
and Netflix, that make widespread use of BI and Data Assets to create
significant business value.
With currently available digital technology options, the capabili-
ties of BI are much more accessible to SMEs than before and there
is increased recognition and acceptance by SME leaders of the value
opportunities that strategic BI creates. However, the availability
of mechanisms to support such organizations in the implementa-
tion of BI at a strategic level are limited in literature and practice.
In the Appendix we provide examples of the use of the AIMS-BI
Methodology to enable various kinds of BI applications. The remain-
der of this book provides business and technology executives in SMEs
with clear, step-by-step directions for applying AIMS-BI in the
development of a customized BI roadmap to value-creation that is
tailored to their own businesses.
2
A n A g ile I nteg r ated
M e tho d olo gy for
S tr ateg i c B usines s
I ntellig en ce (AIMS -BI)
Art and science have their meeting point in method.

Edward G. Bulwer-Lytto
Introduction
Organizations that aspire to become data-driven competitors by

developing a data-driven decision-making culture need to develop an
analytic orientation and BI capabilities enterprise wide. Within the
strategy domain, BI is considered to be an enabler of organizational
transformation that can be employed at the enterprise level to intro-
duce new business models and business strategies, and fundamentally
change the way a company competes in the marketplace. Therefore,
“Business Intelligence (BI) has become a strategic initiative and is
now recognized by chief information officers and business leaders as
instrumental in driving business effectiveness and innovation”.1 This
view is reinforced by a growing body of evidence pointing to the stra-
tegic benefits of BI. However, there are also high failure rates and
cases where companies have spent more resources than their com-
petitors on BI solutions with a smaller Return on Investment (ROI)
and at the same time have seen their market share and customer base
shrink. Many organizational BI initiatives struggle through a myriad
of technology platforms, reporting tools and localized data sources
that fail to deliver consistent information that can drive value-added
and timely executive decision-making. These organizations may not
1 http://gartner.com/newsroom/id/500680.
17
even recognize that it is their own organizational weaknesses (e.g.

data security, data access, metadata management and data quality)
that hinder their BI efforts.
Small and Medium-Sized Enterprises (SMEs), in particular, can
be challenged by the technical complexity and resource requirements
of traditional BI solutions. The need to provide agility and strategic
level BI for such organizations is the motivation behind the design
of this new methodology and is guided by the question: What hybrid
methodology will allow for a more nimble and agile approach to enterprise
business intelligence, especially for those organizations with resource and
expertise constraints?
Organizations seeking to deploy agile yet enterprise-level BI need
an effective means of assessing the current state and maturity of their
enterprise information management (IM) practices and of identifying
and addressing critical gaps that could inhibit the desired returns on
successful BI investments. The focus of the analytics is often at the local-
ized level rather than on developing enterprise-wide analytical capabili-
ties. A methodology that can resolve the dichotomy between strategic
enterprise-scale BI and the more agile, lower startup cost profile of BI
initiatives focused in a specific functional area is important for SMEs
with resource and expertise constraints.
Analytics Competitors have been described by Davenport and
Harris (2007) as organizations that improve performance through the
application of data analytics to their business processes. An important
feature of these organizations is that they apply analytics in a struc-
tured way, as part of an organizational strategy, championed by top
leadership and delivered through an integrated enterprise approach,
to empower decision-makers at every level. A Gartner report empha-
sizes the need for structured frameworks that help business and IT
leaders “architect a more complete plan for business intelligence (BI),
analytics and performance management (PM) that will align people,
processes and applications, and ultimately yield better returns”. 2 A key
insight from the report is that each organization needs to derive the
appropriate configuration (e.g. roadmap) based on its own strategic
objectives and business circumstances. The methodology outlined
2 https://gartner.com/doc/1209327/gartners-business-intelligence-analytics-
performance.
A IM S - BI 19
in this chapter provides a sequence of steps that will produce, as its

ultimate output, this configuration.
Davenport and Harris (2007) describe many exemplars of orga-
nizations that, by employing an enterprise approach with strong
senior management advocacy and leadership, make widespread use of
Business Analytics and Data Assets to create significant value for their
business. They suggest a path that organizations need to traverse in
order to arrive at a state of being an “Analytics Competitor” in which
the organization consistently realizes a stream of rich insights from
BI that it is able to translate into actionable value. This path typically
requires navigating through a critical stage (stage 2) in which func-
tional analytics initiatives are used to build momentum and secure
executive interest. The next stage, stage 3, requires an executive com-
mitment and resource allocation to a strategic BI vision. While the
functional analytical initiatives are supported by a number of existing
knowledge discovery process models, methodologies for moving from
functional to enterprise-wide BI are lacking, thereby preventing orga-
nizations from reaching stage 3. AIMS-BI fills this void.
This chapter describes a methodology that addresses the need for
implementing strategic yet agile BI in organizations; an AIMS-BI.
The methodology supports the agility and strategic perspective
needed to achieve optimal business value from BI initiatives. It will
enable organizations to derive appropriate configurations (i.e. a BI
roadmap) based on its own strategic objectives and business circum-
stances. AIMS-BI integrates a number of existing techniques, includ-
ing information maturity assessment, multi-criteria decision-making
and proof of concepts through prototyping, to support the agility and
strategic perspective needed to achieve optimal business value from
BI initiatives. The methodology is a multiphase solution that helps
the organization understand not just what needs to be done but also
when, how and why. Thus, it describes the activities, methods, tools
and techniques that should be used throughout the process.
Why the Need for New Methodologies?

The Need for a Strategic, yet Agile Perspective
The value of adopting a defined systematic and repeatable pro-

cess in planning and carrying out data-driven initiatives has been
widely accepted. There are a number of existing process models which,

collectively, have been termed Knowledge Discovery and Data Mining
(KDDM) process models. Currently the most commonly used of these
is the CRoss-Industry Standard Process for data mining (CRISP-DM)
which consists of six steps: (1) Business Understanding; (2) Data
Understanding; (3) Data Preparation; (4) Modeling; (5) Evaluation;
and (6) Deployment. These models assume that the organization is in
a ready position to adopt BI tools and techniques. However, in reality
this may not be the case and if significant investments are made in the
BI tools and techniques without the prerequisite factors in place, the
BI initiatives will fail. AIMS-BI provides a methodology for assessing
this state of readiness, identifying the capability gaps and developing a
strategic path forward in a very agile way.
Previously, there were two main approaches to BI that are often
viewed as dichotomous – Strategic vs Tactical. The strategic approach
is the top-down, integrated enterprise approach prescribed by both
Davenport and Harris (2007) and Gartner’s Business Intelligence
and Performance Management Framework,3 and is consistent with
the “Bill Inmon” view of the Corporate DW (Inmon, 2005) as the
“single organizational repository of enterprise wide data across many lines
of business and subject areas that: Contains massive and integrated data;
Represents the complete organizational view of information needed to
run and understand the business”. The alternative contending tactical
approach derives from the contemporary view of Ralph Kimball, who
prescribes a bottom-up, dimensional approach to BI based on func-
tional data marts which are ultimately aggregated into the corporate
DW. The Inmon vs Kimball debate is well represented in the litera-
ture. Breslin (2004) summarizes the essential differences between
both approaches (see Table 2.1).
The Kimball approach allows for a more agile, lower start-up costs
and quicker delivery of a BI solution that is tactical and focused in a
specific functional area of the business. However, critics are of the
view that this approach tends to lack integration between disparate
initiatives and enterprise scalability, and, due to its lack of a s trategic
focus, is unlikely to achieve transformational data-driven impacts.
3 http://umsl.edu/~sauterv/DSS4BI/links/pdf/BI/gartners_business_intelligen_
142827.pdf.
A IM S - BI 21
Table 2.1 Characteristics Favoring Inmon’s or Kimball’s Model

CHARACTERISTIC FAVORS KIMBALL FAVORS INMON
Nature of the organization’s Tactical Strategic
decision support
requirements
Data integration Individual business areas Enterprise-wide integration
requirements
Structure of data Business metrics, performance Non-metric data and for data
measures and scorecards that will be applied to meet
multiple and varied
information needs
Scalability Need to adapt to highly volatile Growing scope and changing
needs within a limited scope requirements are critical
Persistency of data Source systems are relatively High rate of change from
stable source systems
Staffing and skills Small teams of generalists Larger team(s) of specialists
requirements
Time to delivery Need for the first data warehouse Organization’s requirements
application is urgent allow for longer start-up time
Cost to deploy Lower start-up costs, with each Higher start-up costs, with
subsequent project costing lower subsequent project
about the same development costs
On the other hand, the Inmon approach is strategic in nature, and

architected for enterprise integration, but demands higher start-up
costs and longer delivery times, can be rigid and may be difficult for
some organizations to undertake, especially those constrained by the
requisite skills, resources and time requirements of this approach.
AIMS-BI can bridge the gap between the two approaches as it sup-
ports organizations in becoming analytical aspirants using a strategic
yet agile approach.
Know the Current State of your Information Maturity
The previous approaches to BI, whether tactical or strategic, consider

the adoption of BI techniques and technologies without necessar-
ily ensuring that the organization is in a state of readiness for BI.
One important aspect that needs to be considered is the state of its
IM maturity. A number of models are available for assessing differ-
ent aspects of maturity within the organization but for AIMS-BI we
have adapted an IM maturity model which is used by organizations to
assess their current state in terms of the maturity of their IM practices

and to identify gaps in these areas. These IM maturity models provide
organizations with a formal means of identifying c apability gaps in
enterprise IM disciplines, critical for strategic BI success. Performing
these periodic maturity assessments can help organizations to moni-
tor and measure how improvements have helped to close the identi-
fied gaps.
End User and Key Stakeholder Engagement
A key component of ensuring the agility of the methodology is the

engagement of the end users and key stakeholders early in the process
and their inclusion in key decision-making related to the BI initia-
tives that should be pursued. This inclusionary approach is important
as the stakeholders must feel that they are a part of key decision-
making. Since the decision about the choice of BI initiatives is subjec-
tive, with different groups having contending views, group subjective
decision techniques become very useful in this process. Therefore the
Analytical Hierarchical Processing (AHP) technique was integrated
into AIMS-BI.
Design of AIMS-BI
AIMS-BI was developed using the Design Science research meth-

odology which provides a set of guidelines for the development and
evaluation of ICT artifacts to address real-world problems. AIMS-BI
is the artifact that has been developed to provide organizations, such as
SMEs, with a prescriptive methodology for guiding the o rganization
on a path to strategic BI.
Design science research requires the rigorous evaluation of the
artifact using one of a number of proposed evaluation techniques.
AIMS-BI has been evaluated using primary cases studies; where its
application in the business environment was analyzed in depth.
AIMS-BI starts with an assessment of the organization’s infor-
mation maturity (Step 1) and progresses to the development of BI
prototypes (Step 4). This methodology bridges the gap between
knowing and doing. The early stages emphasize “knowing” about
the organization, identifying strategic priorities and discovering BI
A IM S - BI 23
Table 2.2 Relevant Techniques

# STEP OF AIMS-BI EXISTING TECHNIQUES USED KEY OUTPUTS/OUTCOMES
1 Information Survey design & administration; IM assessment report; information
maturity capability assessment maturity capability baseline; gap
assessment analysis
2 BI opportunity Semi-structured interviews; case Business understanding, strategic
discovery study analysis priorities; discovery of BI
opportunities; portfolio of
Proof-of-Concept (PoC) initiatives
3 BI portfolio Stakeholder engagement/priority Select priority PoC initiatives; key
evaluation assessment; PoC value executive/participant awareness &
assessment/strategic alignment; engagement
Multi-Criteria Decision Making
(MCDM) a
4 Proof of Technology deployment; agile Data integration/data as a service
concept software prototype development; technology sandbox; proof-of-
prototypes Knowledge Discovery/Data Mining concept analytics prototypes;
(KDDM); data profiling/quality systematic process for data quality
assessment management/metrics; valuation of
data assets
5 Strategic BI BI portfolio planning & strategic Strategic BI roadmap
roadmap alignment
opportunities. The latter stages focus on the “doing” aspect of BI,

including experimentation and the demonstration of business value,
and culminate with the strategic BI roadmap which provides the
organization with a pathway for strategic BI.
To ensure the rigor of the development of AIMS-BI, existing lit-
erature was analyzed to identify relevant methods and/or techniques
that should be integrated into the methodology. These methods and
techniques, which include capability assessment surveys, interviews,
technology deployments, multi-criteria decision-making techniques,
KDDM, modeling and data analytics, are summarized in Table 2.2.
Description of AIMS-BI
The adoption of the readily accessible BI tools, technologies and data

assets alone is not enough to give organizations the distinctive com-
petitive advantage they are looking for through BI. Rather it is the
complex organizational systems and structures, processes, employee
expertise and executive commitment in combination with these tools,
technologies and data assets that provide this advantage. Therefore,

organizations seeking to effectively deploy strategic BI need an
effective means of assessing the current state and maturity of their
enterprise IM practices and identifying critical gaps that could pres-
ent barriers and inhibit the desired returns on BI investments.
Furthermore, there is a need for methodologies that can resolve the
dichotomy between strategic enterprise-scale BI and the more agile,
lower startup cost profile of BI initiatives focused in a specific func-
tional area of the business. This ability is especially important for those
organizations, such as SMEs, that have shied away from investing
in strategic BI solutions due to the perceived high cost and intensive
resource requirements often associated with these strategic solutions.
Therefore, the AIMS-BI methodology proposed is enterprise in scope
yet agile in execution and considers, not just the BI tools and technolo-
gies and data assets, but also other factors of importance to BI at the
strategic level. The final output of this methodology – the strategic BI
roadmap - consists of a dynamic portfolio of initiatives that are strate-
gic in nature yet modular, thus allowing for agile execution.
The relevant techniques described in Table 2.2 were synthesized
into a repeatable methodology referred to as an “Agile Methodology
for Strategic Business Intelligence – AIMS-BI” (Figure 2.1) which
consists of four steps culminating in the generation of a comprehen-
sive strategic BI roadmap. One of the aims of this methodology is to
guide organizations in knowing not just what needs to be done, but
also how it can be done. Each of the steps is described below.
Step 1: Information Maturity Assessment
The key foundational step of AIMS-BI is the IM maturity assess-

ment that will enable the organization to assess and benchmark their
Enterprise IM capabilities. This provides the setting within which
BI initiatives are conceptualized, developed and deployed in order to
deliver the maximum strategic benefits to the organization. The IM
assessment recognizes that enterprise-level BI capabilities are com-
plex, hierarchical and multifaceted in nature, comprising not only
information resources and data assets, but also organizational struc-
tures, people skills and processes. The results of this assessment pro-
vide the organization with a good understanding of its baseline BI
A IM S - BI
Figure 2.1 AIMS-BI.

25
readiness and identify areas of weaknesses that must be addressed

if it is to compete on analytics. Periodic IM maturity assessments,
conducted through self- or third-party assessment, and benchmark-
ing, help to establish an informed basis for formulating effective IM
development programs and tracking continuous improvement.
The outputs of this step form an important input to the BI road-
map as they identify key areas that the organization should focus on
improving if they are to achieve the benefits of strategic BI. This IM
assessment provides a number of benefits: (i) It allows the organiza-
tion to quickly determine the state of its IM capabilities as a pre-
cursor to formulating an enterprise-wide analytics strategy; (ii) It is
done by the stakeholders which ensures stakeholder/user participation
early in the process; (iii) It does not require a significant investment of
expertise or financial resources so is suited to SMEs.
The IM assessment process is described in more depth in Chapter 3.
Step 2: B
I Opportunity Discovery
This step is concerned with arriving at a thorough understanding of

stakeholders and their business needs and identifying strategic BI
opportunities for the organization. This is achieved by using semi-
structured interviews to better understand the business processes, to
assess any current BI initiatives that the organization has in place or
is considering, and discovering BI opportunities. Many organizations
have some idea of the areas which they think would benefit from the
application of BI (e.g. Customer and Product Analytics) and these
are explored further. In parallel with gaining an understanding what
already exists within the given organization, an understanding of
what similar organizations have achieved though BI initiatives and
their critical success factors, is obtained through a review of case stud-
ies. This allows the organization to be clear, not just about what they
are thinking of investing in, but, importantly, identify areas in which
others, globally, have had strategic BI success.
This step may also include other useful opportunity discovery
activities such as:
i. A BI introspection session – a facilitated session, that should
involve the executive leadership and management team,
in which the organization’s vision and aspirations and the
A IM S - BI 27
opportunities and barriers to becoming a BI competitor are

examined.
ii. An Opportunity Discovery Workshop – a facilitated work-
shop with key operations and management team members in
which the IM assessment findings from Step 1 are shared and
opportunities for BI initiatives and organizational priorities
are “discovered” through key stakeholder engagement. This
active involvement of key stakeholders is one of the critical
success factors for BI projects.
The BI opportunities that are identified from this step form the basis
for the development of a portfolio of BI initiatives.
Step 3: B
I Portfolio Evaluation
The aim of this step is to develop and prioritize a set of Proof of

Concepts (PoCs) from the portfolio of possible BI initiatives. It is
unlikely that organizations, especially SMEs, will have the expertise,
skills and other resources to tackle all the possible BI opportunities.
Executives and key decision-makers will have their own preferences
for which BI initiatives are to be implemented; however, there are a
number of criteria to consider in this prioritization process. Therefore
a subjective, multi-criteria, group decision-making technique can be
used to make the process unbiased yet participatory. One such tech-
nique is Analytical Hierarchical Processing (AHP) which provides
a framework for solving different types of multi-criterion decision
problems based on the relative priorities assigned to each criterion’s
role in achieving the stated objective.
In order to more accurately estimate the potential value to the
organization of each initiative, further information is elicited from
the stakeholders through online forms and semi-structured inter-
views. A combination of these and other inputs, including the results
from the IM Assessment (from Step 1), are used to define a set of
possible PoCs tailored to the organization’s context and strategic
priorities which can be used to demonstrate the value of BI to key
stakeholders for organizational buy-in. Further details about the BI
opportunity discovery and portfolio evaluation steps are provided in
Chapter 4.
Step 4: P
roof of Concept Prototypes
The top PoCs to be developed into prototypes, with the available

resources, are selected from the portfolio of prioritized initiatives.
The ability to rapidly mobilize BI prototypes requires a flexible
platform in order to connect to heterogeneous, non-conventional
internal and external sources of data to evaluate and better con-
trol the scope, cost and timeline of their implementation. The tra-
ditional Extract, Transform and Load (ETL)/Data Warehousing
applications for Data Integration are considered to be too rigid and
resource and time intensive to support the agility being sought in
AIMS-BI generally and, more specifically, for the development
of the prototypes. However, the emerging disciplines of Data
Virtualization and Data as a Service present alternative approaches
to the traditional ETL/Data Warehousing applications for Data
Integration and provide a range of functions and Data Delivery
Services that enable Agile BI. Furthermore there are a number of
Open Source solutions for these services which allow for the rapid
development of the prototypes without significant investments in
tools and technologies. The use of open-source Data Integration
software provides for a level of deployment agility and experi-
mentation that yields significant reductions in lead-time, cost and
time-to-value, and will ultimately help to inform the BI investment
decisions.
The prototypes, by demonstrating the strategic value of BI, allow
the members of the executive team to better understand the potential
of these investments. In turn, this will help to get the “buy-in” from
senior management that is so critical to the success of strategic BI.
The development of the prototypes can also help in the early identi-
fication of problems that may have to be addressed before investing
significant resources in a full-blown BI strategy. They provide the
organization with a good sense of how much time and resources will
have to be invested in order to get the data to a form that is suited
for the analytics required. The use of Open Source data integra-
tion software to support the prototypes provides a level of deploy-
ment agility and experimentation that yields significant reductions
in lead-time, cost and time-to-value. Additionally, given that the
prototypes are deployed using iterative, prototyping techniques that
A IM S - BI 29
allow for active user engagement and experimentation, they support

agility.
The details of the development of PoC prototypes are p
rovided in
Chapter 5.
Strategic Business Intelligence Roadmap
The final output of AIMS-BI is the strategic BI roadmap which pro-

vides the organization with a balanced portfolio of BI initiatives and
a pathway for establishing a strategic BI capability. Both the maturity
assessment and the PoCs are integral to the development of the BI
roadmap as they identify key areas on which the organization needs
to focus and develop capabilities if it is to maximize the benefits of its
investment in strategic BI. The roadmap will typically include tech-
nology and applications projects as well as capacity building/training,
governance structures/roles and policy initiatives – all c ritical factors
that are required to be in place to support the implementation of these
initiatives.
The roadmap is designed in such a way that, even though it is
Enterprise-wide in scope, the organization can plan its execution
incrementally, based on its own absorptive capacity for change, with-
out overcommitting financial or human resources. The completed
prototypes provide key bootstrap initiatives that can lead to important
early success and demonstrable value in the execution of the road-
map. The mixture of developmental and applied initiatives will allow
the critical gaps identified from the IM Assessment to be addressed,
while concurrently generating early returns from delivering new BI
capabilities.
By highlighting key decisions, integration points, gaps and depen-
dencies, the roadmap is amenable to formal program management
execution to balance resources and investments across the multiple
BI, analytics and infrastructure initiatives and projects. The built-in
annual IM maturity assessment provides a systematic means by which
the organization can measure improvements in its IM and BI capa-
bilities and will help executives track the realized business impact and
effectiveness of the BI program as it is rolled out.
Further details for the development of the strategic BI roadmap are
provided in Chapter 8.
Findings from Applying AIMS-BI
AIMS-BI was evaluated using the case studies of both a financial

organization and an academic institution. Both organizations were
grappling with the issue of how to get more value from their data and
were in the early stages of investing in BI. Both, therefore, were open
to allowing AIMS-BI to be applied to their organizations.
After the application of the methodology in these organizations,
key stakeholders were interviewed to ascertain the value they saw in
applying the methodology. Some of the findings were:
1. The stakeholders felt that the IM assessment proved to be
extremely useful in identifying the state of readiness of the
organization for strategic BI. It helped them to identify a
number of areas that would have to be addressed in order to
maximize the benefits of its BI investment. Further, the find-
ings from the PoC prototypes supported the findings of the
maturity assessment. For example, in both case studies, the
assessment identified issues with data quality and, in develop-
ing the PoC, specific issues of data quality were identified (e.g.
the inconsistency in data values generally and, more specifi-
cally and concerning, in those fields used to join tables). The
interviewees felt the IM assessment was an extremely valu-
able first step. The results of the assessment process provided
empirical evidence in support of their perceived weaknesses
and shortcomings. This provides a stronger case for investing
resources to address these weaknesses. They were also of the
view that the assessment should not be administered solely
by the internal team but rather in the partnered approach
(i.e. include an independent third-party) as it was felt that a
purely internal approach could be perceived as biased.
2. There was also affirmation for the inclusion of the key
stakeholders and decision-makers in the development and
evaluation of the BI portfolio. It was felt that their inclusion,
especially in the form of a workshop, allowed these diverse
groups to weigh the options in terms of what should be pri-
oritized and why in a collaborative way. The view was that the
stakeholders might have felt excluded if they were not a part
of the process and their buy-in is important.
A IM S - BI 31
3. The PoCs and, more specifically, the development of the pro-

totypes demonstrated to the executive management team and
all key stakeholders the business value of BI, the systematic
approaches to better data quality management, the impor-
tance of data governance and the agility of data as a service
(DaaS). This secured/reinforced executive commitment to the
process. In fact, one of the organizations felt that even closer
collaboration and engagement would have been beneficial for
further transfer of knowledge from the external team to the
data analysts etc.
The findings from the case study also support the literature in terms
of some of the critical success factors, these included:
1. The need for a champion within the organization. The finan-
cial organization’s project manager turned out to be this
champion and the success of the study can be attributed, in
large part, to this person as she ensured that the sponsors and
executive team were available to meet with the researchers in
a timely manner.
2. Data quality is critical to BI success. In both case studies, data
quality received a low score from the IM assessment and the
development of a formal data quality process was included as
an early, mandatory initiative in both the BI roadmaps.
Conclusion
It is being recognized that the traditional methods for BI are too com-
plex and rigid, costly, time consuming and resource intensive for many
organizations and there is, thus, a need for lower cost, less resource
intensive, agile approaches to strategic BI. AIMS-BI provides an agile
methodology for strategic BI. The method integrates various existing
techniques to generate a strategic BI roadmap that reflects the key BI
opportunities, a visible program of implementation, and secures the
executive commitment and resources required to put the organization
on the path to becoming a serious “Analytics Competitor”.
The first step of AIMS-BI, IM assessment is critical as it identifies
a number of key areas in which the organization score may be low but
that are essential for BI success (e.g. data quality). Without addressing
the issues in these areas, the business value that could be achieved
through BI investments would not be realized and senior manage-
ment would consider the project a failure. The second step provided
valuable insight into the current practices and opportunities within
the organization. In the third step, the use of the subjective group
decision-making technique is extremely useful as it engages senior
management and ensures their commitment to the BI initiatives. The
final output, a strategic BI Roadmap, identifies the key BI oppor-
tunities, sets out a visible program of implementation, and positions
the organization to secure the executive commitment and resources
required to maximize the BI business value. AIMS-BI provides a
systematic and structured approach that is enterprise-wide in reach
yet agile in execution and focused on organizational needs and capa-
bilities. Although the primary output of AIMS-BI is the strategic
BI roadmap, an important output of Step 4 would be the working
prototypes that are at an advanced stage of development and will
not require a great deal more effort for full deployment. AIMS-BI is
enterprise in scope yet agile in execution and, if followed, can ensure
that an organization maximizes its investments in strategic BI.
Although this methodology can be applied in any organization
looking at developing enterprise-wide analytic capabilities, it is partic-
ularly applicable to organizations that may not be suited to traditional
BI methods due to limited resources and the need to demonstrate
value quickly, such as SMEs.
Part II
N av ig ating
the A g ile BI
P ro ces s
This section describes the steps of the AIMS-BI methodology which

details not just what should be done but also how. Thus, if followed
carefully, these steps will navigate the organization to a position of
being an analytics competitor. Each chapter highlights the importance
of agility throughout the methodology and identifies ways in which
this can be achieved. Recommendations for less resource-intensive
options are also provided where possible. All these considerations will
allow SMEs to be an important target group for AIMS-BI.
Chapter 3 describes the first key foundational step of AIMS-BI, the
Information Management (IM) maturity assessment, which enables
organizations to assess and benchmark their Enterprise IM capabili-
ties as a precursor to formulating an Agile Business Intelligence (BI)
strategy. The proposed IM maturity assessment model was developed
by adapting, synthesizing and extending existing maturity models and
methods. The IM maturity assessment enables the agile assessment of
the IM capabilities at the enterprise level and across multiple domains
including People/Organization, Policy, Technology, Compliance,
Measurement and Process/Practice. Importantly, this key step
responds to one of the oft-cited criticisms of Information Systems
maturity models that is, although the primary purpose of maturity
models is to assess and identify capability gaps, many models do not
describe how to effectively perform the actions required to address
the gaps identified. AIMS-BI addresses this “knowing-doing” gap by
33
embedding the IM maturity assessment as the key foundational stage

of the overall agile methodology.
Chapter 4 goes on to describe the next step of AIMS-BI –
discovering BI opportunities. Based on the premise that it is unlikely
that organizations, especially SMEs, have the resources to address all
BI opportunities at once, this step provides guidance on how to build
and evaluate a portfolio of BI initiatives in order to identify those that
are perceived as likely to provide the greatest strategic value to the
organization. Depending on the resources available and the resources
needed for each, the top initiatives can be selected for implementation
as Proof of Concepts (PoCs). This chapter highlights the opportuni-
ties for the engagement of key stakeholders in building this portfolio
of initiatives which improves the agility of the methodology.
Chapter 5 describes the process models that can be used to build the
analytics PoCs prototypes, for both Visualization and Data Mining.
These process models are important as they provide a structured,
systematic and repeatable process for analysts to follow to ensure the
success of data-driven initiatives. Too many organizations, by taking
an ad hoc approach to BI initiatives, are undermining the value that
they can provide.
Chapter 6 describes some of the critical issues related to Data
Quality (DQ ) and Data Governance, which have been identified
as two factors that constrain the efficiency and effectiveness of data-
driven analytics initiatives and decision-making processes. The IM
maturity assessment evaluates both DQ and Governance and provides
details about particular issues that are affecting their overall scores.
This chapter outlines what both DQ and Data Governance entail,
explains why they are critical to BI and then goes on to describe, in
detail, a systematic approach to DQ standards definitions and data
quality measurement. The chapter will also demonstrate the use of
DQ Dashboards in empowering data stewards to formulate and over-
see adherence to DQ standards and policies across the organization.
Throughout the emphasis will be on the systemic approaches to mea-
surement, ranking and quantification of DQ and asset value and how
these can be maintained through proper Data Governance practices.
Finally, Chapter 7 describes data integration which provides the
decision-makers with the flexibility to explore on-demand, and
prototype various BI initiatives. Business analysts and decision-makers
N av i g atin g t he Agil e BI P r o c e s s 35
require access to data sources in a timely manner, to experiment,

prototype and evaluate various analytics initiatives. Achieving this
timely access to data requires greater agility in data integration and
the capability and tools to rapidly access and integrate data from
multiple heterogeneous sources. This chapter describes the options for
achieving this data integration and provides a detailed discussion of
how agility can be achieved.
3
I nfo rmati o n M anag ement
(IM) M aturit y A s ses sment
Evaluating Business Capabilities and Gaps
Without proper self-evaluation, failure is inevitable.

John Wooden
Introduction
There is a high failure rate for Business Intelligence (BI) initiatives

and many organizations struggle through a myriad of technology
platforms, reporting tools and localized data sources that not only
fail to deliver value-added support for timely executive and manage-
ment decision-making, but also expose the business to vulnerabilities
in terms of data security, data access, data quality and data scalability
issues (Jourdan et al., 2008).
One of the contributing factors to the high failure rate of BI ini-
tiatives is that many organizations launch initiatives without a clear
sense of the state of the IM infrastructure, organizational capabilities,
proper data quality and key business value opportunities. As orga-
nizations seek to exploit the potential of strategic BI, there is need
for an effective means of assessing the current state and maturity of
their Enterprise Information Management resources and practices
and to identify critical gaps that could present barriers and inhibit
the desired returns on, and success of, BI investments. It is important
that this mechanism provides a high-level assessment perspective for
executive management, while being granular enough to inform a sys-
tematic program of BI initiatives.
Information Management (IM) Maturity Assessment is a key
foundational step in the proposed Agile Integrated Methodology
for Strategic Business Intelligence (AIMS-BI) methodology as it
37
enables organizations to assess and benchmark their Enterprise IM

capabilities across multiple domains (i.e. People/Organization, Policy,
Technology, Compliance, Measurement and Process/Practice) as a
precursor to formulating an Agile Business Intelligence (BI) strategy.
It provides the organization with some measures of its IM maturity
across a number of domains and these measures can then be used
to identify gaps and provide some baseline values for measuring
improvements in IM capabilities when remedies are put into place.
The results of the assessment provide the organization with a better
understanding of their BI readiness. Maturity models, have increas-
ingly been used by organizations as a systematic way to assess, mea-
sure and benchmark their capabilities in various domains. Examples
include the Organizational Project Management Maturity Model
(OPM3)1 and the Capability Maturity Model Integration (CMMI).2
This chapter describes a proposed IM Maturity Model that was
implemented as a component of AIMS-BI to enable organizations
to assess and benchmark their Enterprise IM capabilities. The pro-
posed model was developed by adapting, synthesizing and extending
existing maturity models and methods and was used to develop an
online survey instrument which organizations can adopt for carrying
out their IM assessment. The model determines the organizational
context within which BI initiatives should be conceptualized, devel-
oped and deployed, in order to deliver the maximum strategic benefits
to the organization.
Importantly this key step responds to one of the oft-cited criti-
cisms of Information Systems maturity models, i.e. although the
primary purpose of maturity models is to assess and identify capa-
bility gaps, many models do not describe how to effectively per-
form the actions required to address the gaps identified. AIMS-BI
addresses this “knowing-doing” gap by embedding the Information
Maturity Assessment as the key foundational stage of the overall agile
methodology.
1 h t t p s : // l e a r n i n g. o r e i l l y. c o m / l i b r a r y / v i e w /o r g a n i z a t i o n a l - p r o j e c t-
management/9781628250305/.
2 https://cio.com/article/2437864/developer/process-improvement-capability-
maturity-model-integration-cmmi-definition-and-solutions.html.
IM M AT URIT Y A S SE S SM EN T 39
Maturity Models in Organizations
Maturity models have gained prominence as they allow organizations

to assess, measure and benchmark organizational capabilities relat-
ing to Processes, Objects and People (Mettler, 2009). By identifying
capability gaps through a formal assessment of “Where am I today?” vs
“Where would I like to be?” organizations are able to determine their
maturity (i.e. competency, capability, level of sophistication) within
a selected domain and establish a systematic basis for an informed
program of development and continuous improvement. Maturity
models have been developed for specific domains such as software
development, project management, business process management and
IT business value. Perhaps the best known is the Capability Maturity
Model Integration (CMMI), used to manage the software develop-
ment processes, which integrates several derivative models from the
very popular Capability Maturity Model (CMM).
Maturity models can generally be categorized as one of two types,
Staged and Continuous: (i) Staged or Fixed-level maturity models
distinguish a fixed number, typically five, of generic maturity levels.
Each maturity level is associated with a number of focus areas (pro-
cess areas in CMM) that have to be implemented satisfactorily or be
in compliance with all specified requirements of one maturity level,
before the organization can progress to the next level. The original
CMM for software and its successors are typical of a staged model;
(ii) Continuous or Focus-area maturity models, on the other hand,
put emphasis on a number of focus areas or dimensions that have to
be developed to achieve maturity in the target functional domain.
This type of maturity model identifies capability levels within each
focus area, thus allowing organizations to develop their capabilities or
competencies within those focus areas without being restricted to
the generic maturity levels. The overall maturity of the organiza-
tion is then expressed as a combination of the specific maturity levels
within the individual focus areas. The derivative CMMI allows for
both staged and continuous representation.
Focus-area maturity models provide the flexibility to express granu-
lar interdependencies within and between the focus areas thus allow-
ing organizations to define a balanced, incremental development path,
taking all focus areas into account. This focus area approach becomes
an important design consideration for BI maturity models, given the

complex, hierarchical and multi-order nature of strategic BI capabilities.
Implementing the IM Maturity Model
The IM maturity model to support AIMS-BI was synthesized from

a number of existing maturity models, developed for Enterprise
Information Management/BI, based on the following criteria:
• The scope of the maturity model should cover the organi-
zation’s Enterprise Information Management resources and
practices as the context for deploying an agile BI capability.
• It should target both business and technology management
and practitioners.
• It should be based on practitioner-oriented maturity models that
have been validated by industry and practitioner application.
• The model should be multidimensional addressing the matu-
rity of process (e.g. the maturity of data governance), people
(e.g. level of analytical skills of business analysts) and artifacts
(e.g. quality of data assets).
• The application of the maturity model should be based on an
assessment tool that can be self- or third-party-administered
to a combination of management and staff respondents.
Based on these criteria and a review of twelve BI Maturity Models,
the AIMS-BI maturity model was synthesized from a combination
of the MIKE2.0 IM QuickScan tool (Methodology for an Integrated
Knowledge Environment)3 and the IBM-DGC Maturity Model4
(DGCMM). The MIKE2.0 IM QuickScan tool provides coverage
of all three elements of people, process and object, while also provid-
ing detailed, granular coverage of twenty eight functional areas and
contains a comprehensive set of Capability assessment questions that
can be adapted and reused for different organizational contexts. Three
new functional areas, Value Creation, Strategic Alignment and Data
Usage Quality, along with additional questions to assess their capabil-
ity were included in the synthesized model.
3 www.openmethodology.org.
4 https://935.ibm.com/services/uk/cio/pdf/leverage_wp_data_gov_council_matu-
rity_model.pdf.
Figure 3.1 IBM data governance council maturity model.
The IBM-DGC Maturity Model (DGCMM) reflects close

agreement and concurrence with the domains and functional areas
in the MIKE2.0 IM Maturity Model. However, the DGCMM
also provides the added benefit of emphasizing the interconnect-
edness and path dependencies between the various elements (see
Figure 3.1). This is an important feature of Continuous or Focus-
area maturity models as it allows for the crafting of a balanced,
incremental development program, taking all focus areas into
account (van Steenbergen et al., 2010).
The synthesized AIMS-BI IM maturity model examines information
maturity on six dimensions: People/Organization, Policy, Technology,
Compliance, Measurement and Process/Practice as follows:
1. People/Organization considers the human side of Information
Management, how people are measured, motivated and sup-
ported in related activities. Organizations that motivate staff
to think about information as a strategic asset tend to extract
more value from their systems and overcome shortcomings in
other categories.
2. Policy considers the message to staff from leadership and
whether staff is required to administer and maintain informa-
tion assets appropriately with consequences for inappropriate
behaviors. Without good policies and executive support, it is

difficult to promote good practices even with the right sup-
porting tools.
3. Technology assesses the tools that are provided to support
IM activities. While technology on its own cannot fill gaps
in the information resources, a lack of technological support
makes it difficult to establish good practices. Given the tech-
nical nature of the content, only the IT representatives were
required to answer these questions.
4. Compliance surveys the external IM obligations of the orga-
nization and helps assess how well an organization is pre-
pared to meet targets from industry and external regulators.
A low compliance score indicates that the organization is
relying on luck rather than good practice to avoid regulatory
and legal issues.
5. Measurement looks at how the organization identifies infor-
mation issues and analyzes its data. Without measurement, it
is impossible to sustainably manage the other aspects of the
framework.
6. Process/Practice considers whether the organization has
adopted standardized approaches to IM. Even with the right
tools, measurement approaches, and policies, information
assets cannot be sustained unless processes are consistently
implemented. Poor processes result in inconsistent data and a
lack of trust on the part of stakeholders.
For example, the assessment of the maturity of data stewardship in
organizations will be measured at the level of supporting policy guide-
lines, technical infrastructure, process capability and measurement
practice. The instrument is supported by an extensive set of capability
statements that allows for a very granular assessment of organizational
capabilities, the identification of gaps and the design of incremental
programs of improvement.
The dimensional structures and functional areas evaluated by the
synthesized maturity model are shown in Figure 3.2. The two ques-
tionnaire items shown illustrate how the policy and people dimension
of data usage quality are assessed.
Figure 3.2 AIMS-BI synthesized IM maturity model.
Applying the AIMS-BI IM Maturity Assessment
A number of resources are available to support the AIMS-BI IM maturity

assessment. These include an on-line survey instrument, integrated with
a spreadsheet-based analysis and reporting tools, that is readily adaptable
to different organizations. The questionnaire is typically administered in
parallel with semi-structured interview sessions with various stakehold-
ers and from these a final IM assessment report is generated.
Example of the Application of IM Maturity Assessment
Using a contrived case of an educational institution for illustrative

purposes, the following example demonstrates how the AIMS-BI
maturity assessment tool would be applied.
The IM Assessment is typically performed in parallel with semi-
structured interview sessions and focus group workshops with a
diverse set of stakeholders. The persons selected to participate in the
IM assessment should constitute a representative cross-section of
key business stakeholders. The IM maturity assessment is conducted
using the AIMS-BI IM Maturity tool administered using the online
instrument. The following narrative describes the typical output of
such an assessment.
Figure 3.3 below shows the overall capability levels derived across
the six high-level dimensions, through the IM assessment. The
weighted result of this assessment is 1.6 out of a maximum of 5; the
highest scores being assessed for Technology (2.0) and the lowest for
Policy and Measurement (1.3).
Using the five levels of maturity for an organization, based on their
IM practices, as proposed by META Group (see Table 3.1), the over-
all score of 1.6 would be categorized as reactive – meaning that there
is minimal formalized Enterprise Information Management practices,
except in isolated silos, and the organization reacts to data quality
issues as they arise. More significantly, the higher levels of maturity
in the Technology and Compliance categories suggest the existence
of relatively sound traditional technology competencies and capabili-
ties. However, the lower maturity scores for Policy and Measurement
signal clear gaps in formal Data/Information Governance mechanisms
and the minimal use of best practice behaviors and techniques such
as data quality metrics and profiling/measurement. The overall need
for improved Data Governance practices in areas such as Data qual-
ity management, ownership and stewardship is also manifest in the
qualitative comments recorded from the various user group interviews.
Figure 3.3 Information maturity assessment.

Table 3.1 Information Maturity Model (IMM) Capability Definitions

INFORMATION MATURITY MODEL CAPABILITY DEFINITIONS
Level 1 – Aware No common information practices. Any pockets of information management
(IM) maturity that the organization has are based on the experience and
initiatives of individuals
Level 2 – Reactive Little in the way of Enterprise Information Management practices. However,
certain departments are aware of the importance of professionally
managing information assets and have developed common practices used
within their projects. At the enterprise level, a level 2 organization reacts
to data quality issues as they arise
Level 3 – Proactive Has a significant degree of IM maturity. Enterprise awareness, policies,
procedures and standards exist and are generally utilized across all
enterprise projects. At level 3, the IM practices are typically sponsored and
managed by IT
Level 4 - Managed Manages information as an enterprise asset. The business is heavily
engaged in IM procedures and takes responsibility for the quality of
information that they manage. A level 4 organization has many mature
and best-in-class practices and utilizes audits to ensure compliance
across all projects
Level 5 - Optimized Considers information to be as much an enterprise asset as financial and
material assets. A level 5 organization has best-in-class IM practices
that are utilized across all enterprise projects. The distinguishing
characteristic of a level 5 organization is the focus on continuous
improvement. At level 5, all data management practices and assets are
regularly measured and the results are analyzed as the basis for
process improvement
Source: MIKE2.0 (www.openmethodology.org).
The assessment also allows the organization to analyze the details

for each domain. For example, the organization was interested
in looking at the details for the technology category. The technol-
ogy category is further broken down and assessed according to a
number of functional areas and each is assigned its own score (see
Figure 3.4). Each functional area for Technology Management has
been described in Table 3.2. The functional areas actually reveal
some of the reasons for technology score being at reactive level even
in the presence of enterprise-wide systems. The scores for Profiling/
Measurement, Metadata Management, Data Stewardship and Data
Capture, all of which are important for harnessing the value from
the data, are all weak. This again highlights the importance of the
Information Maturity Assessment before investing significantly in
analytics because no matter how much is invested in the analytics, if
Figure 3.4 Technology assessment.
Table 3.2 Definition of Technology Functional Areas

FUNCTIONAL AREA DESCRIPTION
B2B data integration Is data received from, or sent to, other organizations consistently
handled?
Cleansing Are data cleansing tools appropriately deployed?
Common data model Considers the use of consistent data models and standards
Common data services Use of middleware and standard messages to interface data
between systems
Data analysis Extent to which data analysis technologies are appropriately
deployed
Data capture Considers the use of consistent design approaches to collecting
data from end users or customers
Data integration Are automated data integration tools used (as opposed to
inconsistent manual coding)?
Data quality metrics Considers the inclusion of automated data quality metrics
Data standardization Extent to which data integration is combined with technology to
standardize data (such as capitalizing names, splitting fields etc.)
Data stewardship Provision of monitoring technology for data quality and other
management metrics
Data usage quality Extent of utilization of the primary data assets by end users for
their decision-making
Data validation Consistent use of rules to validate data entered or supplied
Master data management Use of common platforms to manage master data across the
organization
Metadata management Appropriate technology to manage and disseminate metadata
Platform standardization Extent to which data management forms part of the technology
platform strategy
Profiling/measurement Is a data profiling technology consistently used?
Security Technology management of data aspects of security
these technology functional areas are not improved the benefits will
not be realized.
The institution’s investment in transactional and operational sys-
tems resulted in Technology being assessed the highest of the six
domains, however, the weaknesses in policy, people and practices
limit the effectiveness of the IT investments. This perspective will
enable business leadership to look beyond the technical issues to iden-
tify and address the critical areas of attention if it is to realize greater
value from its data assets and information resources.
Aggregating the same assessment results using the IBM-
DGCMM dimensions (see Figure 3.5) reflects very clearly where the
strengths and weaknesses of the organization lie within its Enterprise
Information Management capabilities. The result of this alternative
assessment is illustrated in Figure 3.5. This perspective reinforces
the earlier insights derived from the IM Assessment. The institution
is not creating/realizing maximum value from its data/information
assets and requires greater attention to organizational enablers such as
improved Information Governance practices in Data ownership and
stewardship. Executive management needs to signal the importance
of BI through more effective messaging, policy and organizational
mechanisms (roles/responsibilities).
Figure 3.5 IBM-DGC maturity model assessment perspective.

The findings of the IM Assessment are discussed, with both func-

tional and technology participants, and validated using key stake-
holder workshops. The results are reconciled with the outcomes of
semi-structured interviews that are conducted in parallel. This typi-
cally provides a qualitative consistency to support the quantitative
assessment scores. The granularity of the IM Assessment will enable
the organization to program the development of a balanced portfolio
of BI initiatives that will include technology and applications proj-
ects, as well as capacity building/training, organizational structures/
roles and policy initiatives. Additionally, the granularity provides an
important medium of engagement for executive, functional and tech-
nology management to collaborate in the formulation of a strategic BI
Roadmap.
Conclusion
In this chapter, we posit that an IM maturity assessment is an impor-

tant foundational step that is both agile and resource-efficient and
will enable organizations to assess and benchmark their Enterprise
Information Management capabilities as a precursor to formulating a
strategic, agile BI strategy.
The AIMS-BI IM maturity model supported by the assessment
instrument is flexible to adaptation and use in various types of orga-
nizations. This allows for adaptability and matching to the context of
SMEs.
One of the important “intangibles” of conducting the IM Maturity
Assessment is that it provides an important medium of engagement
and dialogue for executive, functional and technical management to
collaborate in the formulation of strategic BI initiatives. This facili-
tated dialogue can highlight, for instance, the shared responsibility
for data quality assurance including institutional mechanisms such as
data standards, metadata management and the role of data stewards.
One of the common criticisms of maturity models is that whereas
the primary purpose of maturity models is to assess and identify capa-
bility gaps, many do not describe how to effectively address the gaps
identified. The IM maturity assessment as the first stage of AIMS-BI
embeds the IM assessment in an integrated development process that
addresses this significant “knowing-doing” gap.
4
C re atin g BI P ortfoli os
Most of us spend too much time on what is urgent and not

enough time on what is important.
Stephen R. Covey
Introduction
The second and the third steps of AIMS-BI involve discovering and
prioritizing BI opportunities in order to create a prioritized BI port-
folio of possible Proof of Concepts (PoCs) initiatives. This need for a
prioritized portfolio is based on the premise that it is unlikely that
organizations, especially SMEs, have the resources to address all BI
opportunities at once. AIMS-BI allows organizations to build and
evaluate a portfolio of BI initiatives to identify those that are perceived
as likely to provide the greatest strategic value to the organization.
This chapter describes how these initiatives can be identified and pri-
oritized, then, depending on the resources available and the resources
needed for each, the top initiatives can be selected for implementation
as PoCs.
The development of this prioritized portfolio requires an under-
standing of the stakeholder needs and identifying possible opportu-
nities for the organization in terms of strategic BI. It is important to
gain a thorough understanding of the organization’s business, assess
any current BI initiatives that the organizations have in place or are
planning to invest in, and discover BI opportunities that align with
the strategic initiatives of the organization.
The Discovery of BI Opportunities
The aim of this step is to identify a set of strategic BI opportunities

for the organization. In order to ensure the completeness of this set,
multiple stakeholders (e.g. business leaders, business analysts and data
49
analysts) across several functional areas of the organization should be

engaged. This early engagement of stakeholders adds agility to the
methodology.
The engagement of the stakeholders is done using semi-structured
interviews and the objectives of these interviews are to: (i) gain a more
thorough understanding of the organization’s business; (ii) assess cur-
rent BI initiatives; (iii) understand any planned BI investments; and
(iv) discover BI opportunities. Discovering BI opportunities must be
done within the context of the specific organization, so the interviews
are used to assess the current situation of the organization including
identifying existing resources, understanding the constraints and iden-
tifying business success criteria. In parallel with understanding what
exists within the organization, an understanding of what other com-
parable organizations have achieved through the adoption of BI initia-
tives and the critical success factors they have identified are extremely
important considerations. This consideration of the achievements of
others allows the organization to be clear, not just about what they
should invest in, but, just as importantly, what others, globally, have
done successfully. This global perspective can be achieved by referring
to existing case studies. The information from the interviews together
with the findings from the analysis of the case studies are then used to
develop the organization’s portfolio of BI initiatives.
In order to better understand the value of each initiative to the
organization, further information is elicited from the stakeholders.
This additional information includes:
• the name of the initiative

• the type of BI involved (e.g. advanced vs descriptive analytics)
• the business process it supports
• the target users
• the business context
• the strategic objective the initiative aligns with
• the consumers of the information
• key business questions the initiative will answer
• the performance measures
• the constraints (i.e. what will affect the success of the initia-
tive; e.g. poor data quality)
• the data sources needed to implement the initiative.
C re atin g BI P o r t f o li o s 51
Figure 4.1 Inputs to BI portfolio.
An online form can be developed to collect this information and

follow-up interviews can be used to clarify this information. All
the information gathered is used to create the portfolio of Proof of
Concepts (PoCs). The inputs into the development of the portfolio are
summarized in Figure 4.1.
BI Portfolio Evaluation
When a portfolio of PoC initiatives has been created using the guide-
lines above, it may be necessary to prioritize the initiatives given the
likelihood of limited resources. It may be the case that some of the
PoCs are mandatory as they are prerequisites to the success of the other
initiatives. For example, based on the IM Maturity Assessment (see
Chapter 3) and the discussions with the stakeholders, the issue of data
quality may keep reoccurring. Given that the success of the initiatives
is likely to be highly dependent on the quality of the data, it may have
to be agreed that a data quality PoC is mandatory. The remaining
PoCs would then be prioritized to help in determining which should
be developed into prototypes.
Different stakeholders will have their own preferences as to which
PoCs are most important and there are a number of criteria to consider in
this prioritization process. A subjective, multi-criteria decision-making
technique is suited to this process. Based on these needs, the multi-
criteria decision making technique, Analytic Hierarchical Processing
(AHP) (Saaty, 1980), can be used as it is specially designed
for multi-criteria, subjective estimation. AHP uses a hierarchical

structure with a goal, a set of alternatives that are to be prioritized
and the set of criteria that need to be considered in the prioritization.
In this case, the goal is the prioritization of the PoCs in the portfolio,
the alternatives are the possible PoCs and the criteria are the strategic
objectives of the organization. Using the strategic objectives as the cri-
teria against which the PoCs should be assessed ensures the alignment
of the BI agenda with that of the strategy of the organization. The
pairwise comparison of the alternatives in relation to each of the cri-
teria is done by the stakeholders using the ranking guide in Table 4.1.
The criteria themselves can also be prioritized based on their impor-
tance to the organization. The pairwise comparison combines the cri-
teria importance with the alternative preference measure to derive a
numerical priority for each solution alternative. This approach checks
the ranking’s consistency in the pairwise comparison to ensure that
the comparison is being done in a consistent manner. The output from
applying AHP will be the portfolio of PoCs with a numerical priority
for each. This priority rating is used to identify those PoCs that will
provide the greatest impact in terms of achieving the strategic objectives
and therefore should be where the limited resources should be invested.
The selected PoCs are to be signed off by the PoC sponsors (i.e. the
senior manager(s) under whom the PoC falls). The inclusion of the
stakeholders in the prioritization process as well as the alignment of
each initiative with a strategic objective supports the agility of the pro-
posed methodology.
Table 4.1 Ranking Guide

COMPARED TO THE SECOND ALTERNATIVE,
THE FIRST ALTERNATIVE IS: NUMERICAL RATING
Extremely preferred 9
8
Very strongly preferred 7
6
Strongly preferred 5
4
Moderately preferred 3
2
Equally Preferred 1
Example: BI Opportunity Discovery and Portfolio Evaluation
To demonstrate the opportunity discovery and portfolio evaluation steps

we use the example of a financial institution. The IM assessment was
performed and was followed up with the semi-structured interviews
which provided a better understanding of the state of the organiza-
tion’s BI agenda. The main strategic objectives of the organization were
also elicited in these discussions. They were framed on Sales, Payments,
Efficiency and Service. The information gathered at this stage is critical
for the alignment of the PoCs with the strategy of the organization.
An example portfolio of PoCs for a financial institution is described
in Table 4.2. For each of the possible BI initiatives, a brief description
and the strategic objective it aligns with have been provided.
Through discussions with senior management, the set of BI initia-
tives was reduced to a number of possible PoCs. Based on the current
needs, some of these PoCs were considered mandatory and some were
Table 4.2 Example Portfolio of PoCs

STRATEGIC
PoC PoC DESCRIPTION OBJECTIVE
1 Market This PoC demonstrates analytics models/techniques that can Sales
segmentation be used to segment the customer market
2 Credit risk This PoC demonstrates the use of analytics modeling Sales,
analysis techniques (e.g. decision trees) for credit risk analysis. It is payments
likely that financial institutions are already doing some risk
analysis but this PoC demonstrates the importance of
considering alternative models
3 Call center Analysis of call center data set: Efficiency,
traffic • To determine how the call center operators are performing service
pattern (e.g. length of time to process various requests)
analysis • To show the traffic at the call center at various times
which can then be used to determine the optimal
allocation of call center resources
4 Market Basket Identification of products that sell well together and can help Sales
analysis to increase the effectiveness of sales campaign management
and targeting processes (e.g. cross-sell and upsell of
products)
5 Customer Many organizations have found that a small percentage of Sales,
segment customers contribute the majority of the value to the service
value organization. Therefore this PoC calculates the profitability or
analysis lifetime value of customers and show the changes that occur
as the customers move through the lifecycle (e.g. from low
profitability through medium to high)
(Continued )
Table 4.2 (Continued ) Example Portfolio of PoCs

STRATEGIC
PoC PoC DESCRIPTION OBJECTIVE
6 Segmentation It is easier to market to existing rather than new customers, Sales,
& credit risk therefore, this PoC will analyze customer data to identify payments
analysis of ways to get more value from the semi-banked
the
semi-banked
7 Banking Each distribution channel has a cost associated with it and Efficiency,
distribution there are benefits of moving customers across channels. This service
channel PoC can also assist in allocating resources (e.g. identifying
analysis and where new ATMs/branches should be located)
optimization
8 Sales Develops sales performance dashboards that provide members Sales
performance of staff with interactive drill-down visibility for sales
management performance across the enterprise
dashboard
9 Payments Explores the use of payments data for customer analytics to Payments
analytics determine how payments can provide additional predictor
variables for credit-risk scoring or customer segmentation
10 Data quality Demonstrates the importance of data profiling and how it can Sales,
analytics on be used to assess and improve the quality of organizational efficiency,
customer data. Data quality is one of the most significant factors service
data affecting the success of data-driven initiatives
11 Data Identifies benefits and issues of data integration and Efficiency
integration/ demonstrates techniques for improving data accessibility. It
data as a highlights the importance of data standardization to the
service integration process. It demonstrates that appropriate data
governance and standardization policies can lead to
significant improvements in the efficiency of decision-making
12 Visual analysis A key goal of agile analytics is to provide decision-makers with Efficiency
with Tableau the tools to access data and direct their own analysis. The
workshop organization had Tableau, a popular and powerful
analytics/data visualization tool, but it is was being used
extensively
thought to be important for short-term consideration. For example,

Data Quality is foundational to the success of analytics initiatives
but it was clear from the IM assessment and interviews that there
were data quality issues that could hamper BI and therefore the Data
Quality PoC was considered to be mandatory. Another mandatory
PoC was Data as a Service (DaaS) as this provides the ability to inte-
grate data on a need-be basis. Four other PoCs that the organiza-
tion identified as important in the short term were Market Basket
Analysis, Payment Analytics, Distribution Channel Optimization,
Segmentation and Credit Risk of Semi-Banked. It was agreed, how-

ever, given the availability and access to resources and the need for a
quick turnaround, that only two of these four should be developed as
prototypes.
A larger team was constituted to select the top two PoCs for devel-
opment. The inclusion of key decision-makers across various func-
tional areas as a part of this selection team ensures further buy-in and
agility in the methodology and the outcome. The subjective multi-
criteria group decision-making technique, AHP, was used for this
prioritization process.
This prioritization process was extremely participatory. At a half-
day workshop the stakeholders were briefed on the findings from
the application of AIMS-BI to date and were asked to prioritize the
remaining initiatives based on the impact each would have on the
strategic directions e.g. (i) Credit Expansion (ii) Sales and Service (iii)
Efficiency (iv) Payments (see Figure 4.2).
The stakeholders were grouped and each group was asked to do
a comparison of each pair of alternatives with respect to the criteria
using the ranking guide shown in Table 4.1. For an example template
see Figure 4.3.
The criteria themselves could have also been ranked, however, in
this case, they were assumed to have equal importance. On comple-
tion of the pairwise ranking, the rankings were normalized and used
to compute a score for each of the alternatives with respect to each
of these criteria. These scores were then combined to derive an over-
all score for each alternative, which was then used to prioritize the
PoCs (see Table 4.3). In the example, Market Basket Analysis and
Figure 4.2 Criteria and alternatives.

Ranking by C1 - Credit Expansion

A1 A2 A3 A4
A1 1.00 8.00 9.00 1.00
A2 0.13 1.00 0.13 0.11
A3 0.11 8.00 1.00 0.11
A4 1.00 9.00 9.00 1.00
Ranking by C2 - Sales & Service
A1 A2 A3 A4
A1 1.00 8.00 4.00 0.33
A2 0.13 1.00 0.20 0.11
A3 0.25 5.00 1.00 0.17
A4 3.00 9.00 6.00 1.00
Ranking by C4 - Payments
A1 A2 A3 A4
A1 1.00 6.00 1.00 3.00
A2 0.17 1.00 4.00 3.00
A3 1.00 0.25 1.00 2.00
A4 0.33 0.33 0.50 1.00
Ranking by C3 - Efficiency
A1 A2 A3 A4
A1 1.00 0.13 0.14 0.13
A2 8.00 1.00 7.00 5.00
A3 7.00 0.14 1.00 0.14
A4 8.00 0.20 7.00 1.00
Figure 4.3 AHP pairwise comparison.
Table 4.3 AHP Ranking of PoCs

ALTERNATIVE POC RANKING
A4 Market Basket 1
A1 Payment Analytics 2
A2 Distribution Channel Optimization 3
A3 Segmentation and Credit Risk of Semi-Banked 4
Payments Analytics were the two chosen for the development of ana-
lytics prototypes.
Conclusion
The process of the development and prioritization of BI initiatives

proved to be extremely participatory especially when it came to
applying the prioritization process. There were very lively discussions

and active participation in applying the AHP. The buy-in from stake-
holders as well as the alignment of the initiatives with the strategic
objectives of the organization ensures that the BI agenda is transfor-
mational. Thus, the chapter highlights the importance of engaging
key stakeholders in building this portfolio of initiatives as this con-
tributes to the agility of the AIMS-BI methodology.
5
The P ro ces s and Value
o f B uild in g P ro of - of -
C o n cep t P rotot ypes
If you can’t describe what you are doing as a process, you don’t
know what you’re doing.
W. Edwards Deming
Introduction
AIMS-BI consists of four steps which lead to the development of a

strategic Business Intelligence (BI) roadmap, tailored to the orga-
nization’s business context and capabilities. This chapter focuses on
Step 4 which involves developing the prototypes for the selected Proof
of Concepts (PoCs). The prototypes serve several purposes including
securing the buy-in of executive management and providing a more
thorough understanding of the amount of time and resources that
will have to be invested in a full-blown BI implementation. The pro-
totypes are also integral to the development of the BI roadmap as
they help to identify those key areas on which the organization needs
to focus and develop capabilities if it is to maximize the benefits of its
investment in a BI strategy. The agile methodology requires a quick,
low-cost, turnaround development of the prototypes which can be
achieved using open source solutions. However, many organizations
may not have the resources and expertise to developing the PoCs
effectively and therefore end up building them in a rather ad hoc way.
In this chapter, we discuss the existing process models that can be
adopted to ensure that the development of PoCs is done is a system-
atic structured way.
59
Why Develop PoCs?
Davenport and Harris (2007) provide many exemplars of organizations

that make widespread use of BI and Data Assets to create significant
value for them by employing an enterprise approach with strong senior
management advocacy and leadership. The authors propose a path that
organizations need to traverse to reach an “Analytics Competitor” state
in which the organization begins to consistently realize a stream of
rich insights from BI that translates into actionable value. As illus-
trated in Figure 5.1, reaching this state typically requires navigating
through a critical stage 2 in which functional analytics initiatives are
used to build momentum and secure executive interest. Empirical evi-
dence suggests that many enterprise BI initiatives stall at this point,
which might be characterized as a “chasm of uncertainty”, and never
make it to stage 3, where there is executive commitment and resource
allocation to a strategic BI vision. AIMS-BI provides a systematic and
rapid approach to navigating these first three stages and producing a
strategic BI roadmap that reflects the key BI opportunities and a visible
program of implementation. The organization will then be positioned
Figure 5.1 Path to becoming an analytical competitor.

P R O C E S S & VA LUE O F BUIL D IN G P o C P R O T O T Y P E S 61
to secure executive commitment and the resources required to put it on

the path to becoming a serious “Analytics Competitor”.
BI – The Different Layers of Analytics
The various layers of analytics (see Figure 5.2) include reporting, analy-
sis, monitoring and prediction. The techniques available in each layer
answer specific types of questions. The lowest level, the Reporting level,
focuses on “What happened?” questions and so includes techniques
that facilitate querying and searching (e.g. SQL). The next layer up is
the Analysis layer which focuses on the question “Why did it happen?”
and allows the decision-makers to view the data from multiple dimen-
sions using Online Analytical Processing (OLAP) and data visualiza-
tion tools. The next layer, the Monitoring layer, focuses on the question
“What is happening now?” and uses tools such as dashboards and score-
cards. The Prediction layer looks into the future to answer the question
“What might happen?” which requires data mining techniques and tools.
Finally, the Prescription layer is about finding appropriate actions, based
on the results of the prediction combined with the decision-makers’
business insights, to ensure the benefits of BI initiatives.
The techniques and technologies recommended to answer the ques-
tion related to each layer should not be applied in an ad hoc way, but
Figure 5.2 Layers of analytics.

should follow one of the process models that have been used to build
BI initiatives. Such models provide a systematic and focused approach
to the development of the BI PoCs. The following section discusses
the application of models to the Monitoring and the Prediction layers.
Monitoring Layer
The monitoring layer is focused on the question of “What is

Happening Now?” and to answer this question Visualization tech-
niques and technologies are recommended.
What Is Data Visualization?
Data visualization is a technique that allows information and data to

be represented in a graphical or pictorial format. Data can often be
overwhelming and difficult to process and since “a picture is worth
a thousand words”, visualization is used to simplify it. Visualizations
can take varying formats from key measures being presented to the
executives each morning, to big screens that show real-time data, such
as statistics in a graphical manner, to interactive charts/graphs that
allow the user to explore the data. Data visualizations are used to
tell stories through the use of the many available visuals (e.g. charts,
graphs, images, text) connected in a cohesive and coordinated man-
ner. They not only allow decision-makers to see the complete picture
as it emerges from the data, but also allow a level of interactivity across
the visuals to explore how changes in one visual may affect the others.
Visualizations are built to narrate the stories that exist in data i.e.
“What is happening now?” The insights from data are often lost if
they are not explained to the decision-makers. Visualizations use
graphics to translate large volumes of data to insights which can be
used to make informed decisions and converted into actions. The data
visualization process should follow a model to ensure that it is done in
a systematic and structured way.
Data Visualization Process Models (VPM)
Process models, generally, provide a systematic, repeatable set of steps

to guide decision-makers. The Data Visualization Process Model
(see Figure 5.3), in particular, outlines the steps that decision-makers

should follow when implementing a Data Visualization initiative.
The VPM model starts with an understanding of the business context
and ends with the enabling of decision-making and actions. To under-
stand the business context those KPIs that affect the business objec-
tive are first determined and then used to help identify the variables
in the data set that should be considered. To see how the variables
are performing, various graphs and charts are generated and finally
the designer needs to ensure that the visualizations are assisting with
some decision or action. Each of the steps is described below:
Understanding the Business Context

This step requires that the analyst identifies the business questions
that are to be answered by the visualization. Not only do these busi-
ness questions need to be identified, but it is also important that the
analyst understands the audience (i.e. the decision-maker) for whom
the visualization is being developed. To understand the audience,
Figure 5.3 Data visualization process model.

the analyst needs to interact with the business user and identify the
story that needs to be told through the visualization. Numbers con-
vey a powerful message; therefore the important metrics will need to
be identified. A good starting point for identifying these metrics is
the organization’s Key Performance Indicators (KPIs) as these met-
rics align with strategic objectives and measure actual performance
against targets, thresholds and timelines. KPIs are critical to any
performance management system so they need to be incorporated
in the dashboards or stories of the visualization. Also these metrics
should be represented in the visualizations as many leaders recognize
that “What gets measured gets done” and so the inclusion of the KPIs in
the visualization will improve accountability. The business questions
should focus on the following aspects of the organization:
• What – e.g. what are the best-selling products?
• When – e.g. at what times of the year do the sales of a certain
product spike?
• Where – e.g. in which regions/countries is a particular product
doing best?
• As compared to – e.g. how are the sales of Brand X’s 50 mL
water compared to Brand Y’s over the last 3 months?
• By how much – e.g. what is the percentage increase in the sales
of Product X over the last 3 months?
Get Data
Having identified the key business questions and metrics for the visu-
alization, the next step is to identify and retrieve the data needed
to do the visualization. It is likely that this data resides in disparate
sources throughout the organization, and some may even be external.
Therefore, identified data items will need to be integrated. As a part
of this integration process, the issue of data preparation and cleansing
will have to be addressed. In terms of preparation, some visuals may
require the data to be in a different format to the one it is stored in at
the source so the data will have to be transformed into a form suited to
the visualization. Additionally, any quality issues with the source data
must be resolved before it is included in the data for the visualization.
Chapters 6 and 7 provide details of both data integration and data
cleaning and preparation.
Visualization Design Considerations

To design and build effective visualizations the following questions
need to be considered:
1. How will the user respond to the visualization?
2. How will the visualization be used?
3. How should the data be encoded?
How Will the User Respond to the Visualization?

The user experience can be on three levels, namely:
• Visceral – the visceral experience relates to a quick emotional
reaction. This feeling is more intuitive than rational and
involves the basic perceptual tasks of distinguishing objects
and is really about forming the true first impression.
• Behavioral – this experience is about conscious awareness and
is controlled by cognition but influenced by visceral reaction.
This is the level of most human behavior.
• Reflective – this experience is solely a cognitive and intel-
lectual response; it is not influenced by the visceral reaction.
How Will the Visualization Be Used?

Visualizations and Dashboards can be used for many levels of
decision-making e.g. strategic, analytical or operational decisions.
The narrative will differ depending on the use of the visualization.
Strategic Dashboards – monitor the execution of corporate
strategic objectives at each level of the organization and
emphasize management rather than monitoring or analysis.
They are often implemented using the Balanced Scorecard
methodology and are referred to as scorecards.
Analytic Dashboards – track departmental processes and proj-
ects and emphasize analysis rather than monitoring or man-
agement. They are often implemented using portals and run
against data marts or warehouses where data is refreshed
periodically.
Operational Dashboards – track core operational processes and
often display real-time data. Operational dashboards empha-
size monitoring more than analysis or management.
How Should the Data Be Encoded?

The various dimensions of data can be encoded with appropriate
visual properties. There are a number of options for data encoding
including color, size, shape, spatial position, length, width and orien-
tation, and the selection of these options is extremely important for
the effectiveness of the visual. For example, colors bring out p eople’s
emotions, so for a visualization that is narrating murder rates the color
red automatically heightens the emotive response. Colors also simplify
complexity. In a complex graphical encoding color can draw atten-
tion to one detail while blurring others; it can be used to highlight
or alert the user to some particular event that was discovered in the
data. Colors can be monotone where the ordering shows low to high
concept (e.g. number of customers in different regions), they can be
diverging with the use of two colors and a neutral color e.g. (profit –
low, zero, high), or they can be contrasting to highlight comparisons
(e.g. difference between sales and profit). The choice and amount of
color should be considered carefully as too much color can make the
visualization confusing.
The encoding choice is dependent on the data type as particular
types are suited to particular codings as depicted in Table 5.1.
These options, when used in combination, can be extremely effective
for developing the narrative.
The design questions for visualizations (i.e. how will users respond?
how will the visualization be used? and how should data be encoded?)
are intricately intertwined. For example, the user’s response to the
visualization can be influenced by the data encoding while the visual-
ization can be designed to elicit a specific response. Since the design
for a visceral experience seeks to generate a quick affective emotional
reaction, analysts use color and size to elicit this reaction, for example,
Table 5.1 Type of Data and their Encoding

CATEGORICAL DATA ORDINAL DATA QUANTITATIVE DATA
Position Position Position
Color hues Size Length
Shape Color intensity Size
Clusters Color hues Color intensity
Boundaries Shapes -
red and big can be used to represent a highly dramatic narrative. The
behavioral experience focuses on issues such as readability and usabil-
ity. These visualizations are interactive in nature and allow users
to literally have a conversation with the data. Analytical dashboards
will have a higher degree of interactivity whereas strategic dash-
boards, which present highly summarized data primarily for senior
management decision-makers, require little interactivity. A reflective
user experience occurs in visualizations when the graphics result in
decisions and/or actions. Operational dashboards should be action-
oriented and strategic dashboards should be decision-oriented. Using
size and color for key variables on strategic dashboards will tend to
focus the attention of the decision-makers on the key performance
indicators. All these aspects of the design need to be considered in
order to ensure that the visualization is fit for its purpose and also
geared toward the specific decision-makers.
Building Visualizations
In building visualizations, the data is encoded as a visual object with
multiple components. The analysts will have to select visualization
techniques and group the individual visualizations to form a strong
narrative. The business questions, the data and the design consider-
ations will be used to create the individual graphical encodings. For
example, if the focus of an organization is on customer satisfaction,
the analyst will need to graphically encode the current customer sat-
isfaction levels, customer satisfaction across branches/regions, the
trend of customer satisfaction over years and quarters, the relation-
ship between sales and customer satisfaction etc. For each graph the
analyst will have to determine the most appropriate representation
(i.e. bar, pie, line, heat map) and connect them so that the big picture
emerges.
Business Insights
The visualizations enable decision-makers to gain evidence-based
insights into the organization. They interact with the visualizations to
derive insights based on the compelling stories that emerge from the
graphics. The graphical, summarized and multiple views of the data
enable the decision-makers to understand the complete picture that is
embedded within data.
Decisions and Actions

This step allows the analyst to share insights with other stakeholders
through views, dashboards and stories. These insights are then used
to drive business actions and make more data-driven decisions. The
impact of the visualization will be realized when the decisions and/or
actions are reflected in improvements to Key Performance Indicators
(KPI) of the organization. The visualizations may explain some of
the mysteries embedded in the data which will guide new business
questions.
Prediction Layer
The prediction layer of analytics is focused on the question “What

Might Happen?” To answer this question, data mining techniques
and tools are recommended. Again, a process model for data mining
should be followed to ensure that the required steps are taken in a
systematic and structured way.
What Is Data Mining?
Data mining has been defined as the process of extracting valid,

non-trivial, previously unknown, interesting patterns or associations
from large databases. It is routinely being used in marketing, retail,
banking, telecommunications, supply chain optimization and fraud
detection applications.
The data mining techniques can be classified as either predictive
or descriptive. Predictive techniques are considered to be supervised
learning methods as they require a labeled training set. Labeled
means that there is a training set for which there is a known value
for a target variable. For example, if customers are to be classified as
high or low risk there is a target variable (e.g. risk level) and a training
set of customer data for which their value for the risk level variable are
known to be either high or low. The technique is then used to build
a model which identifies those input variables that are important in
predicting the given target variable and so focuses on either classify-
ing data into a set of predefined classes (i.e. one for each possible value
for the target variable) or predicting future data states for that target
variable. When a new case is presented for which there is no value for
the target variable (i.e. the class it belongs to is not known), the model
is used to predict the class. Examples of predictive analytic techniques
are classification, regression and value prediction.
Classification is the most commonly applied data mining technique
in predictive modeling. It consists of three stages: model construc-
tion, model evaluation and model use. It uses a set of pre-classified
examples (i.e. training data set) to develop a model that can classify
the population in the future. In the model construction or learning
phase the training data is used by the selected classification algorithm
to build a classifier model. In the evaluation phase the generated
model is checked for accuracy using test or validation data and, if the
model is acceptable, it is used to classify new data. Common clas-
sification algorithms include decision trees, logistic regression, k-NN
and neural networks.
Descriptive techniques, on the other hand, are unsupervised learn-
ing methods. They do not have a test data set with a known target vari-
able, rather they focus on describing behavior. Common descriptive
techniques include sequence mining and association rule mining.
Association rule induction is a commonly used technique in data
analytics which is used to find recurring patterns in data of the form
X => Y (i.e. X implies Y), where X and Y are concepts or sets of
concepts which frequently occur together in the data set. Association
rules have been used successfully for market basket analysis which is
based on the theory that if a customer buys a certain group of items,
they are more (or less) likely to buy another group of items. It is a tech-
nique that has become extremely popular in the retail business and
to understand customers purchase behaviors and to identify relation-
ships between the items that people buy. The generation of association
rules is a two-step process. First all the itemsets, where each itemset
is a set of items a customer buys in one transaction, that satisfy a user-
specified minimum support criterion are extracted from the data set.
The associations between the items that occur frequently together are
then identified using a user-specified minimum confidence criterion.
The two criteria, confidence and minimum support, are significant as
they generate a large number of possible association rules and there-
fore require techniques that are able to identify the most useful ones.
As it is unlikely that there is only one possible data mining tech-
nique that can be applied to a given data set, choosing which are the
most appropriate for a given problem and a given data set is not a
trivial task. Therefore, it is important that this choice is guided by a
multiphase knowledge discovery process model such as IKDDM –
Integrated Knowledge Discovery and Data Mining.
IKDDM Process Model for Data Mining
The need for formal data mining process models that prescribe a path
from data to knowledge discovery has been recognized. These models
provide a framework that allows for the identification of all the inputs
and outputs associated with tasks as well as their dependencies within
and across the phases of the process. Following such a process model
provides a mechanism for applying data mining techniques in a sys-
tematic and structured manner thereby increasing the likelihood that
the results will be accurate and reliable and the process is repeatable.
One such process model is the IKDDM model (see Figure 5.4)
which consists of the following phases: Business (or Application
Domain) Understanding (which includes definition of business and
data mining goals), Data Understanding, Data Preparation, Data
Mining/Analytics (or Modeling), Evaluation (evaluation of results
based on Data Mining goals) and Deployment. Each of these phases
is described below:
Business Understanding Phase

This first phase focuses on the project objectives and requirements from
a business perspective, understanding them fully and then converting
this knowledge into a Data Mining problem definition. A preliminary
plan for achieving the project objectives should be developed.
Data Understanding
This phase starts with an initial data collection and proceeds to activities
which familiarize analysts with the data and allow them to identify
data quality problems, to discover first insights into the data or to
detect interesting subsets to form hypotheses for hidden information.
Data Preparation
This phase covers all activities associated with constructing the final
data set (data that will be fed into the modeling tool(s)) from the initial
Figure 5.4 Phases of the IKDDM model (Sharma and Osei-Bryson, 2010).
raw data. Data preparation tasks are likely to be performed multiple times
and not in any prescribed order. These activities include table, record and
attribute selection as well as the transformation and cleaning of data.
Modeling (Data Mining)

In this phase, various modeling techniques (e.g. Decision Trees,
Regression and Clustering) are selected and applied. Typically, there
are several techniques that could be applied to a given problem type
and some of these techniques have specific requirements in terms of
the format of the data that is needed to perform them, therefore, this
phase often requires repeating the data preparation phase.
Evaluation
This phase of the project consists of thoroughly evaluating the model
and reviewing the steps executed to construct the model to be certain
that it properly achieves the business objectives. A key objective is to
determine if there is some important business issue that has not been
sufficiently considered. At the end of this phase, a decision on the use
of the Data Mining results should be reached.
Deployment
The creation of the model is not the end of the process as simply
extracting the knowledge from the data is of little value if it is not
organized and presented in a way that the decision-makers can use
of it. The deployment can be as simple as generating a report or as
complex as implementing a repeatable Data Mining process across the
enterprise. It is important that the decision-makers are clear about the
actions that they can take to make use of the models.
Lessons Learned from Applying the Process

Models to the Development of the PoCs
The development of the PoCs using the established process model

of IKDDM or VPM as presented in the case studies have led to
several key insights. First, modeling is an iterative not a linear pro-
cess. In IKDDM the iterations between the business understand-
ing and data understanding phases improve the understanding of the
data that is required to fulfill the business objective. These iterations
help to identify the data required, the missing data and the dirty
data and lead to the selection of the appropriate input variables and
target variables, and the creation of derived variables. The iteration
between data preparation and modeling, is critical for obtaining rel-
evant results from the modeling phase. These iterations could include
modifying the measurement levels of the variable as these levels
could impact the activities in the modeling phase. However, analysts
need to be mindful that some iterations between other phases will
make the process inefficient and costly so they must be considered
carefully.
Second, data preparation/get data is a non-trivial, often under-
estimated activity, especially when working with data from dispa-
rate sources. Often the absence of business rules for the domain
obfuscates the process of data preparation and the creation of new
variables. The data preparation/get data phase is reliant on the
knowledge and experience of the data analysts, especially their
understanding of the relationship between business objectives and

the data required to achieve these objectives.
Third, the validation and the business value assessment of a model
are functions of technical knowledge of model parameters and busi-
ness domain expertise and interpretation. This insight underscores
the critical importance and the role of business analysts working
in collaboration with data analysts and BI analysts. Collaboration
between the three is essential and informs the business analyst in the
deployment phase of the process models in IKDDM and the business
insights phase of VPM.
Fourth, organizations need to emphasize data quality management
strategies and improve their data quality and metadata as part of a
portfolio of BI initiatives. Data quality is multidimensional in nature
and includes the format, accuracy, presence, meaning and relatedness
of field-level data. Assessing data quality on multiple dimensions is
difficult and therefore it is important that the organization adopts
master data management as this provides a single point of reference
for critical organizational data which ensures that everyone under-
stands and interprets the data in the same way. Data quality is an
important issue as poor data can undermine the generated knowledge.
Conclusion
Data is seen as one of an organization’s most valuable assets and

finding innovative ways to harness value from it can create a stra-
tegic advantage for the organization. The move toward data-driven
decision-making is an indication that organizations have started to
leverage the opportunities presented by their data and that data is
central to decision-making contexts.
The availability of a process model for the development of PoCs
and full-blown analytics initiatives will provide guidance to organiza-
tions that are new to the field and those that may not have a great deal
of experience in applying analytics effectively.
6
D ata G ov ernan ce and
D ata Q ualit y M anag ement
Quality is not an act, it is a habit

Aristotle
Introduction
Data Quality and Data Governance have been identified as two factors
that have constrained the efficiency and effectiveness of data-driven
analytics initiatives and decision-making processes. The Information
Management Maturity assessment, administered in the first step of
AIMS-BI, rates both Data Quality (DQ ) and Data Governance and
highlights issues surrounding them. Beyond the obvious issue of accu-
racy (i.e. degree to which data values are correct), inconsistent data
formats, missing values, same field names with different meanings
from different data sources, and the identification of logical related-
ness between specific data instances based on values are all commonly
occurring issues that highlight the need for a systematic approach to
DQ. Additionally, even if current DQ issues are addressed, without
proper Data Governance and the accountability and ownership it pro-
vides, these errors will reoccur.
This chapter outlines what both Data Governance and DQ entail,
explains why they are critical to Business Intelligence (BI) and then
goes on to describe, in detail, a systematic approach to DQ stan-
dards definition and DQ measurement. The chapter will also dem-
onstrate the use of DQ Dashboards for empowering data stewards in
formulating and overseeing adherence to DQ standards and policies
across the organization. The emphasis will also be on the systemic
approaches to measurement, ranking and quantification of DQ and
75
asset value and how these can be maintained through proper Data
Governance practices.
Data Governance and Data Quality
Data Governance is a set of processes and practices that are estab-

lished to support the formal management of data assets within the
organization. Its aim is to establish consistent DQ , improve data
integrity, control data access and address data security and retention.
These processes and practices are used to manage the execution and
enforcement of authority over the definition, production and usage of
data and data-related resources.
Data Governance covers a number of areas that help the organiza-
tion to gain better control over its data assets, including DQ , Data
Stewardship, Security and Privacy, Integrity, Usability, Integration,
Compliance, Availability and Roles and Responsibilities. DQ is a
mandatory part of the overall Data Governance Strategy, as without
it the organization will not be able to successfully manage and govern
its most strategic asset: its data. Therefore this chapter will focus on
the important issue of DQ.
One of the issues with DQ is that organizations are often unclear
about who is ultimately responsible for its management. A survey of
a group of senior managers in an organization asked, “Who should
be responsible for Data Quality in your organization?” The following
options were presented to them:
• Board of Directors
• Corporate Executives
• Business unit heads and managers
• Front-line workers
• IT department/Data Warehousing team
• Cross-functional team
• Data Quality analysts/Data stewards
• Others
• No one
The results are represented in Figure 6.1 below.
The diverse responses demonstrates that this management team
was divided on who is ultimately responsible with the most popular
DATA G O V ERN A N C E & DATA Q UA LIT Y M A N AG EM EN T 77
Figure 6.1 Survey of responsibilities for data quality.
choice being the DQ Analysts and Data Stewards at 28%. However,

TDWI, an educational and research company for BI and data ware-
housing states the following:
To succeed, a data quality program must be initiated by the CEO, over-
seen by the board of directors, and managed by a chief data quality
office or senior-level business managers in each area of the business.1
The lack of clarity in terms of responsibility may arise because many

business leaders do not understand the critical role the organization’s
DQ plays in their strategy formulation and how it can affect the strat-
egy execution. Making strategic decisions based on poor quality data
can have severe financial and credibility implications.
To overcome this lack of clarity about responsibility PwC provides
the following advice:
Define a Data Strategy and link it to the business and risk strategy such
that it leads to Commitment from Senior Management and other stake-
holders for implementing a central Data Management strategy; a con-
sistent basis for solutions which support business goals.2
1 http://download.101com.com/pub/tdwi/Files/DQReport.pdf.
2 https://pwc.fr/fr/assets/files/pdf/2016/05/pwc_a4_data_governance_results.pdf.
As organizations seek to develop and leverage core Information Assets

and BI capabilities as key enablers of their strategic ambitions, the
importance of DQ emerges as a recurrent issue and is perceived by busi-
ness users as a constraint on the efficiency and effectiveness of various
data-driven analytics initiatives and decision-making processes. Experts
agree that data needs to be managed like a business critical issue and this
requires an enterprise-wide Data Management strategy.
The Information Management (IM) Maturity Assessment conducted
as the first step of AIMS-BI identifies gaps in DQ m anagement func-
tions and Data Governance practices. These may include, for example,
a lack of DQ metrics, data stewardship, data standardization, metadata
management, and profiling and measurement. An example of the IM
assessment of the DQ Management functions is shown in Figure 6.2.
The assessment shows a detailed breakdown of some of the specific
DQ and Governance issues that organizations face. While many orga-
nizations may recognize that they have DQ issues, they may not be
clear about the specific nature of these issues. The example a ssessment
below identifies a number of areas in which the organization is par-
ticularly weak, including Data Stewardship, Data Standardization,
Profiling Measurement.
Organizations are recognizing significant business challenges
and opportunity costs that arise as a result of data issues, including
ineffective data use; Data Governance focused only on security and
Figure 6.2 IM maturity assessment: data quality management functions.

Figure 6.3 Cost pyramid for data quality.
control; a lack of understanding of Data and its impact on business

outcomes; reactive, isolated Customer DQ projects and periodic data
cleanup initiatives to improve DQ.
Some organizations have invested significantly in comprehensive
cleanup exercises, for example, to confirm/update customer’s contact
information. While these cleanup efforts may yield measurable DQ
improvements, these improvements are often short-term. Without the
systematic use of specific tools and processes for measuring (profiling)
the quality of data assets as a means of monitoring and maintaining
progressive DQ improvements, the same errors will reoccur.
The cost pyramid for DQ shown in Figure 6.3 is often used to
illustrate the cost impact of different approaches to DQ. It highlights
the comparative cost implications of addressing DQ issues at source
versus the consequences of cleaning data after capture or working
with erroneous data. It clearly demonstrates the importance of a sys-
tematic approach to addressing DQ issues early in the process.
Data Quality Management Process
A DQ management process, such as that outlined in Figure 6.4,

provides an organization with a set of suggested stages and activities
that should be followed if the organization is to treat DQ with the
strategic importance that it deserves.
The stages of the DQ Management process, as well as the activities
for each of these stages, described in detail below:
Stage 1: D
efining Data Standards Quality Metrics
The first stage of the process involves the definitions of data standards
and quality metrics. It requires the engagement of key stakeholders
Figure 6.4 Data quality management process.
within the organization through an established Data Management

Group.
The output of the data standards development process is a data
standards catalog for the data set. The catalog is intended to provide
a comprehensive listing of the data items that must be maintained/
governed across various systems, according to the agreed standards.
The data standards should be specified in a business rather than tech-
nical language. Additionally, Data Standards development is seen
as an evolutionary process. It is expected that the standards will be
reviewed and updated. The composition, size and type of data entities
may change due to changes in business requirements and technol-
ogy. Hence, the standards outlined in the catalog must be subject to
change management and version control.
Developing the standards requires a review of existing data entities,
formats and validation rules. The standard should be assessed against
known Compliance data requirements and data issues and should be
defined to ensure that known quality gaps are addressed. A sample
data standard definition template is outlined in Table 6.1.
DQ is a multidimensional concept and therefore if an organization
is to tackle their quality challenges, the important dimensions of that
quality must first be identified. The proposed dimensions should be
discussed and agreed on by the Data Management Group. Example
DQ dimensions and their classification as either objective or subjec-
tive are described in Table 6.2.
For each of the data standards the specific objective/subjective DQ
indicators which apply to the data element are defined. An example of
Table 6.1 Data Standard Definition Template

METADATA VALUE
Data Item (DI) name The full name of the data element
DI description A simple but unambiguous definition of the data element
DI type Either string, integer, date/time
Data steward The role who maintains this data element
Date published The date this version was published as a data standard <YYYY/MM/DD>
Is part of The parent element of the data items
Syntax The required format of the data from the business perspective. This
will include the minimum and maximum number of characters, if
appropriate, and the structure of the data type or item
e.g. National ID business format is NN-NNN-NNN
where each N represents a digit from 0 to 9
Validation Generic for types and specific for items. The validation rules to be
applied for acceptance of data
Values List of the acceptable values (e.g. male, female)
Default value For any list of values, the default value to be used unless otherwise
stated
Verification Steps taken to establish the correctness of the data type or item
Comments Additional notes
Data quality dimensions The specific objective/subjective data quality indicators which apply to
and minimum quality the data element (e.g. validity, completeness, usability), the metric
standards relevant to each quality indicator and minimum measurement value
one such data item standard (First Name of the Customer) is shown
in Table 6.3.
The data standards defined in the catalog must have stakeholder
buy-in, therefore they must be signed-off by the members of the Data
Management Group. It is important to note that different stakehold-
ers may have different notions of acceptable quality standards. For
instance, 100% completeness of customer contact information is seen
as a required standard by the Compliance group, due to KYC (Know-
Your-Customer) Compliance obligations, while the Marketing/Sales
department may perceive a 75% completeness as an acceptable stan-
dard for effective customer targeting.
Stage 2: D
ata Quality Assessment
The process of DQ assessment involves systematic data profiling and

quality assessment and the generation of DQ reports and visualization
dashboards. This process requires that, the tables that have the primary
Table 6.2 Data Quality Dimensions

DIMENSIONS OF
DATA QUALITY DESCRIPTION
Accuracy The degree to which data correctly reflects the real-world object being described
Validity The degree to which the data conforms to a standard and business rules
Completeness The extent to which data is not missing and is of sufficient depth and breadth.
The data can be missing at multiple levels:
• population – percentage of population represented
• schema – attributes/tables missing
• data value – missing field values
Consistency The degree to which the data that exists in multiple locations is similarly
represented and/or structured
Integrity The degree to which data conforms to data relationship rules
• Referential integrity
• Uniqueness of primary key
• Cardinality
Currency The degree to which data reflects the real-world concept that it represents
Accessibility The extent to which data is available or easily and quickly retrievable
Uniqueness The degree to which each data record is unique
Usability The extent to which business process(es) and/or individuals understand and
are able to use this data
Relevancy The extent to which the data is applicable to one or more business process(es)
or decision(s)
Believability The extent to which data is deemed credible by those using it
Table 6.3 <Salary> Data Standard

ATTRIBUTE VALUE
Data Item (DI) name Salary
DI description Represents an employee’s salary
DI type (e.g. string, numeric) Numeric
Data steward HR manager
Date published 2010/04/03
Is part of Employee profile
Syntax Minimum X, maximum X numeric value
Validation 1. No dollar sign
2. No commas
Values None
Default value Based on job title
Verification Should be the person’s gross monthly salary
Comments All employees must have a salary
Data quality dimension, DIMENSION METRIC MINIMUM STANDARD
metrics and minimum
Currency Binary value (0, 1) based 1 – salary must be
quality standards
on checks against job within a given range
description for the job title
source of data under consideration should be identified and used for the
DQ assessment phase. A good starting point, in terms of which fields
in the tables to use for the profiling, is the unique identifier key (e.g.
Customer_ID in a Customer table or Employee_ID in an Employee
table). The quality of the unique identifier field is critical to the success
of many of the analytics Proof of Concepts as it is often the common
field used to join the tables needed for the analytics from across mul-
tiple systems. Errors, inconsistencies and missing values in these unique
identifier fields will severely limit the results (i.e. number of rows) real-
ized by joining this table with others and thus can significantly reduce
the size of the data set to which the analytics will be applied.
The Data Profiling activity conducts an assessment of actual data
and data structures and helps to identify DQ issues at the individual
attribute level, at the table level, and between tables. Profiling also
captures metadata as a by-product of the process.
Two examples of types of profiling analyses that should be carried
out on the tables are column profiling and table profiling:
Column Profiling involves describing and analyzing the data
1.
found in each single column/field of a data table. For example,
assume an organization wants to ensure that it can contact its
customers and therefore decides to focus on the data relating
to “Right Party Contact”, Table 6.4 shows the profile for key
contact fields in a possible set from aggregated tables. While
the column analysis should be performed across all the columns
Table 6.4 Example Profile of Contact Data
VARIABLE ROWS FILLED (%) MISSING (%) UNIQUE (%)
First name 69,884 100.00 0.00 26.44
Last name 69,884 100.00 0.00 18.34
Address line 1 69,818 99.91 0.09 60.33
Address line 2 56,519 80.88 19.12 27.95
Country of residency 69,664 99.69 0.31 0.05
Nationality 69,664 99.69 0.31 0.16
Branch 69,884 100.00 0.00 0.06
Primary email address 34,227 48.98 51.02 97.93
Primary contact 52,736 75.46 24.54 97.43
Secondary contact 41,974 60.06 39.94 59.53
Customer_TRN 68,630 98.21 1.79 99.83
ID type 69,884 100.00 0.00 0.01
in the table, of special interest are those columns that facilitate

“Right Party Contact”. It can be seen in the example table below
that approximately 75% of primary contact numbers are filled,
while email address is just under 50% populated. This gives the
organization a good feel of how contactable their customers are.
Table Profiling involves analyzing data across the rows of
2.
a single table to establish dependencies between attributes
within the table. For example, Table Profiling can be used to
investigate the validity/uniqueness, format accuracy and com-
pleteness of the Customer_TRN (a Tax Registration Number
which is a unique nine-digit identification number assigned
to each individual tax payer, business enterprise or organiza-
tion by way of an automated system) field of a data set, given
its likely importance to the effectiveness of customer analytics
initiatives that can span systems and subsidiaries. Figure 6.5
shows as an example of some possible results of profiling the
results of the Customer_TRN field. Although this field is
expected to be the customer’s TRN number which should
conform to the specific format (nine digits), the p rofiling
analysis shows multiple formats of the values in the table –
most prevalent being a numeric value of length 9, a numeric
value of length 7, a value starting with an alpha character fol-
lowed by seven digits, a numeric value of length 6, as well as a
number value of length 9 separated by “–”. Additionally, this
Figure 6.5 Example: profiling Customer_TRN with ID type.

field had a number of null and duplicate values which should

never be the case if this is the primary key of the table. If this
is to be the field that will be used to join other tables, which
is highly likely if it is the key, then the NULL and duplicate
values can lead to inconsistencies and inaccuracies. The attri-
butes that are to be used in the analytics PoCs should always
be profiled before being included, as poor quality data will
lead to poor quality results.
The number of distinctly different formats for the Customer_TRN
value would seem puzzling and erroneous, however, it is possible that
other unique IDs (for example driver’s licence, passport or national ID
numbers) were also captured in the field. The existence of metadata
would have made this data understanding process less time consum-
ing. Subsequently adding the field ID_Type to the analysis allowed
the profiling to identify the number of different ID_Types contained
in the field Customer_TRN (see Figure 6.5). Validating the format
of the Customer_TRN field against the ID_Type using Regular
Expression analysis indicated that 91% of the formats were valid.
Such a variety of personal IDs used as the unique identifiers of
customers can become quite confusing in the absence of proper meta-
data standards. This variety will confound efforts to join the data to
other data sources that may have different unique identifier types,
adversely impacting the effectiveness of enterprise-level Analytics.
This discussion highlights the importance of using data profiling
as an initial step in the data preparation activities associated with
analytics initiatives in order to fully understand the contents, inter-
pretation and use of the variables being included in the analytics, as
inaccurate or misinterpreted data will lead to inaccurate results and
flawed strategic decision-making.
Stage 3: D
ata Maintenance/Stewardship
The third stage of the DQ management process, data maintenance/

stewardship, includes batch data cleaning efforts to address quality
issues detected during DQ assessment. More importantly, a critical
aspect of sustainable DQ is to identify and empower data stewards
to help formulate and oversee adherence to DQ standards and poli-
cies. This data stewardship role requires both technical and business
inputs as the steward must also ensure that standardized data element
definitions and formats are being adhered to and that the metadata
is being maintained. Stewards also manage the process of identify-
ing and resolving any emerging DQ issues and oversee periodic batch
cleansing activities.
DQ dashboards have increasingly become an effective means of
creating the essential visibility needed to facilitate monitoring and
reporting on DQ metrics and compliance over time. They provide
an important tool for enabling and empowering data stewards to
assess the state of DQ , and address root cause process issues within
their designated jurisdiction. Key to the design of DQ dashboards is
understanding and identifying the key quality indicators (KQI) that
the organization needs to track, monitor and manage. Figure 6.6 dis-
plays an example DQ dashboard that emphasizes the quality status
for a key “Unique Identifier” and “Rights Party Contact” fields that
are essential for maintaining effective customer analytics and con-
tact management. As highlighted on the “Dashboard”, an organiza-
tion would be able to establish a baseline, then monitor progressive
improvements in these key data items over time, based on various
DQ interventions. These quality attributes can be stored as part of an
enterprise metadata repository, in order to facilitate historical trend-
ing of DQ improvements as part of a comprehensive data governance/
information life-cycle management process.
For DQ dashboards to be most effective, especially for Executives,
it is useful to combine the data profiling information from individ-
ual variables into aggregate business metrics. For example, in the
prototype dashboard in Figure 6.6, a simple business metric has been
created called RPC_ndx, which provides a quantitative indicator
of the organization’s “ability to contact” its customers. The metric is
computed from Name, Address, Telephone#, Email as follows:
RPC_ndx:  Name*1 + ( Address ) *1 + Telephone*1 + Email*2  5
In this case, it has been determined that having valid email contact
information for a customer contributes twice as much value to this
business metric than any other contact attribute.
Figure 6.6 Example: data quality dashboard.

Stage 4: D
ata Life-Cycle Process Review
The final stage of the DQ management process facilitates root cause

analysis and systematic data life-cycle process improvements to mini-
mize the recurring incidence of new quality issues and the need for
extensive periodic batch cleaning exercises. It begins with the map-
ping of the existing data life cycle which provides stakeholders with a
better understanding of the critical data points from its initial creation,
including integration across systems, access, retention and eventual
disposal. For example, it will show where the various data items are
captured and by which roles, how the data items are being used and
the possible points at which the quality can be compromised. A criti-
cal objective of establishing this mapping is to provide input for the
root cause analysis. As previously pointed out, it is inefficient for an
organization to undergo extensive periodic batch cleaning exercises
without addressing the root cause of the problem as these problems, if
not addressed at the source, will reoccur. Many organizations do not
have a systematic approach to root cause analysis for information/ DQ
issues and, if this is the case, it will be reflected in the results from the
IM maturity assessment.
Once the existing process is mapped and the root cause analysis
performed, the appropriate revisions to problematic business processes
can be identified and addressed. In this more systematic approach to
data governance, the emphasis is on modifying people’s attitudes and
improving processes and the strategy is focused on establishing effec-
tive DQ policies and control processes based on DQ standards and
rules. Supporting mechanisms include automating the monitoring
and reporting of DQ metrics and communicating knowledge about
the value of the data and the importance of DQ across the business.
One common and useful DQ PoC that organizations should
consider as they build out their portfolios is the mapping of the data
life-cycle processes around key customer contact data.
Conclusion
DQ is a recurrent issue across organizations, and is perceived by

business users as a constraint on the efficiency and effectiveness of var-
ious data-driven analytics initiatives and decision-making processes.
The effectiveness of the analytics PoCs is often limited by several

aspects of DQ. Other than the obvious issue of accuracy (i.e. degree
to which data values are correct), inconsistent data formats, missing
values, same field names with different meanings from different data
sources, identification of logical relatedness between specific data
instances based on values are all common issues encountered across
organizations that highlight the need for a more systematic approach
to DQ. A DQ PoC is important as it demonstrates the benefits of a
systematic approach to DQ standards definition and DQ measure-
ment. The use of a DQ Dashboard can help empower data stewards
in formulating and overseeing adherence to DQ standards and poli-
cies across the business. The following specific recommendations/
observations are highlighted:
• Establishing a formal Data Governance framework, both in
terms of the core discipline of DQ Management as well as
institutional approaches to Data Stewardship, should be an
immediate priority for organizations that do not currently
have this in place.
• Formal Metadata Management and Data Standards Catalogs
are critical for the efficient development of PoCs and subsequent
analytics initiatives. The absence of metadata can lead to reduced
throughput in data preparation activities, misinterpretation of
data items and inaccurate interpretation of analytics results. The
Data Standards catalog will include all of the primary Data
Sources and key fields and the Metadata catalog will provide a
centralized, searchable electronic repository of data assets
• It is recommended that organizations invest in an Enterprise
data integration platform, e.g. Talend, that can be employed in
the Data Integration/DQ PoCs. Such platforms enable more
systematic approaches to the logical and physical definition
and standardization of data across the enterprise, and enhance
the efficiency of the definition, storage, movement and deliv-
ery of data to business analysts and decision-makers.
• Special emphasis should be placed on monitoring, tracking
and managing the quality of a relatively small set of critical
data items for integration purposes, using DQ Dashboards and
Stewards. Null data values, missing data and inconsistent data,
especially in the primary key fields, (e.g. Customer ID) can

significantly impact the analytic benefits/value of integration.
• Related to the previous point, Master Data Management for
key data entities, for example Customer, will become impor-
tant for managing data consistency across the subsidiaries and
systems throughout the organization.
• Effective analytics are dependent on the availability and
currency of data over time. If the data values are not being
updated, the analytics will reflect inaccurate trends and will
greatly impact the decision-making. For example, the custom-
er’s home address, marital status and occupation may emerge
as significant predictor fields but if they are not current and
do not reflect how they have changed over time, erroneous
inferences could be drawn. Therefore, part of the data gov-
ernance mechanism should seek to maintain the c urrency of
key data items.
• The primary rationale for BI should be to realize maxi-
mum returns from data assets. Returns may take the form
of increases in operational efficiency, employee productiv-
ity, time to market, sales and revenue, as well as mitigating
costs which may be realized as process failures, opportunity
costs, scrap and rework costs, and the failure to sense and
act on business opportunities. Developing economic models
for the value of information and opportunity costs of DQ
can be demonstrated in a PoC but should be extended and
institutionalized to provide the organization with systemic
approaches to measurement, ranking and quantification of
data/information value creation activities.
• Establish a formal DQ Management process, facilitated by
an Enterprise DQ Platform that integrates all of the requi-
site functions, including profiling, cleansing, standardization
and DQ Dashboard functions to monitor DQ metrics and
support data governance initiatives.
• Institutionalizing Data Stewardship includes defining the
necessary roles, procedures and support mechanisms and
identifying and empowering Data Stewards to help formulate
and oversee adherence to DQ standards and policies.
7
D ata I nteg r ati on
Managing Data Sources for Agility
It’s difficult to imagine the power that you’re going to have when
so many different sorts of data are available.
Tim Berners-Lee
Introduction
As organizations seek to increase their use of analytics, their busi-

ness analysts and decision-makers require access to additional,
non-conventional sources of data, in a timely manner, in order to
experiment, prototype and evaluate various analytics initiatives.
Achieving this timely access to data requires greater agility in data
integration and the capability and tools to rapidly access and inte-
grate data from multiple heterogeneous sources. This integration will
provide decision-makers with the flexibility to explore, and prototype
various Business Intelligence (BI) initiatives, and better control the
scope, cost and timeline of their implementation/evaluation.
The emerging discipline(s) of “Data Virtualization” and “Data as a
Service” together with a new breed of Open Source platforms pres-
ent alternative, lower cost approaches to the traditional ETL/Data
Warehousing applications for Data Integration within the enterprise.
They provide a range of functions (e.g. Federation, Transformation,
Profiling, Cleaning, Interface handling) and Data Delivery Services
(e.g. model-driven Metadata services, Data access Web service
APIs),1 to enable agile approaches to BI. These approaches to low-
cost, agile integration are critical for strategic BI adoption by Small
and Medium-Sized Enterprises (SMEs).
1 Application Program Interface.
91
This chapter will evaluate the approaches, challenges and benefits

of data integration, demonstrate the emerging “data virtualization”
techniques and highlight the importance of (meta)data standardiza-
tion to enhance this process. It provides further discussions of how
agility is achieved in the data integration step of AIMS-BI.
Opportunities through Data Integration
As organizations seek to harness more value from their data assets,

they recognize the importance of integrating existing, disparate, het-
erogeneous sources of data for improved decision-making. In organi-
zations, data often exists in silos and needs to be connected for its full
potential to be realized. Figure 7.1 shows various types of customer
data that may exist in different locations and which, if integrated,
can provide a 360° view of the customer. For example, a financial
institution may have customers’ demographics in one data source
and details of their transactions, such as daily/monthly withdrawals/
deposits or financial instruments purchased, in another data store. If
the customer has a debit/credit card, their behavioral data is captured
in the point of sales transactional database while their queries and
complaints may be stored in a customer center database. Integrating
Figure 7.1 Connecting customer data.

Data In t eg r ati o n 93
these disparate sources of customer data provides a core platform for

enhanced analytics.
The traditional approach of building a Data Warehouse for integra-
tion purposes is now being seen as too rigid and too resource intensive
for some types of organizations. For example, SMEs may not have the
human and capital resources that are associated with the development
of a Data Warehouse. With this in mind, the main objectives of this
chapter are first to explore the various integration techniques and then
to demonstrate non-traditional opportunities that may be suited to
SMEs. These opportunities include:
1. Improving business analysts’ productivity/efficiency by
increasing their access to various internal/external data sets
through increased self-service data access and improved turn-
around on data service requests. More specifically by:
a. Enabling efficient access to data across multiple heteroge-
neous data stores;
b. Enabling on-demand access for analysts to data that is not
already in an intermediary data store;
c. Increasing visibility (publishing) of available intermediary
stores and facilitating wider, controlled access to analysts;
d. Facilitating the creation of data mashups 2 which can be
published to other analysts.
2. Demonstrating the effective use of new information infra-
structure solutions/tools to support key information manage-
ment functions and bolster business analytics capabilities.
Key Deliverables from Realizing Integration Opportunities
The opportunities above, if met, will lead to a number of deliverables

that the organization can adopt for their strategic BI roadmap. These
deliverables include:
1. The identification and installation of the appropriate informa-
tion infrastructure/tool that can be used to (i) demonstrate
to key stakeholders the key capabilities listed from 1a. to 1d.
above and (ii) make required data available for any analytics
2 Data mashups integrate heterogeneous data from multiple data sources.

PoCs the organization selects based on Steps 2 and 3 of

AIMS-BI (i.e. BI opportunity portfolio development and
prioritization).
2. Data integration technology selection and implementation
guidelines documentation published for the organization’s
technical team.
3. Recommendations for possible continued use of the open
source tools (or other existing tools) as part of the BI roadmap.
As a part of the integration process, an important Key Performance
Indicator (KPI) to be considered is to improve data usage quality. Data
usage quality measures the extent to which users trust, have ready
access to, and make effective use of organizational data.
Common Approach to Data Integration
For many organizations, their primary data assets reside in several enter-
prise systems. For example, a financial institution may have some data
in the core banking application, some with the credit card application,
some with loan application and some with the Customer Relationship
Management System. These systems may be using, for example, Oracle
or SQL Server database platforms. It is likely that, in the existing setup,
various mechanisms (e.g. Redpoint, Microsoft SQL Server Integration
Services (SSIS) and native SQL queries) are being used to extract
data from the sources to serve business analytics and decision-support
requirements. In addition, end-user departments may be using a vari-
ety of their own customized tools (e.g. Excel spreadsheets) to access,
retrieve, clean and analyze data relevant to their needs. However, this
practice often constrains the flexibility and timeliness with which the
organization can respond to new requests for access to, and integration
of, data from multiple heterogeneous sources (internal and external).
Possible Integration Approaches
There are a number of approaches to data integration that have been

classified in various ways. One possible way to view this is to consider
two broad classes that are distinguishable based on whether the data
is physically integrated or logically integrated:
Table 7.1 Data Integration Techniques

EXAMPLE
CLASS TECHNIQUE DESCRIPTION
Physical integration Data consolidation Brings data together from several separate systems
into one physical data store
Data warehouse Data from heterogeneous sources is integrated in
(DW)/data mart advance and stored in a DW
Virtual integration Data federation Uses a virtual database and creates a common data
“virtualization” model for heterogeneous data from different systems
Enterprise Service Used to develop a layer of services that allow access
Bus (ESB) to data that remain at their source
i. Physical integration – this involves the integration of the data

from several sources into a new physical data source whether
termed a database, data repository or a data warehouse (DW).
ii. Logical/Virtual integration – this involves the integration of
data without creating another physical data store, but rather
by providing an interface to a unified view of the data that
remains in these disparate data stores. Through this interface,
the technical aspects of the stored data are hidden from the
data consumer, i.e. to the consumer it is as if they are accessing
data from a single data source. There are various ways in which
both the physical and logical integration approaches can be
realized, some examples of each are identified in Table 7.1 and
are described further in the following section.
Examples of Data Integration Techniques

Physical Integration
a. Data Consolidation – Data consolidation physically brings

data together from several separate systems thus creat-
ing a version of the consolidated data in one data store. The
Extraction, Transformation and Loading (ETL) technology
supports data consolidation. The ETL process is responsible
for pulling the data from the disparate sources, transforming
it into an understandable format and loading it into another
database or DW.
b. Data Warehouse (DW) – A DW is a physical repository of his-
torical, enterprise-wide data, cleaned to ensure that it is of a high
quality, obtained from internal and possibly external sources. The

data loaded into the warehouse is perceived as being critically
important for strategic decision-making. The data is organized
and structured to ensure that it can be accessed efficiently for the
strategic type of analysis for which the warehouse is designed.
The building of this DW will include the consolidation of data.
Virtual Integration
1. Data Federation: A form of data virtualization which utilizes a

virtual database to create a common data model for heteroge-
neous data from disparate systems. The data is viewable from
a single point of access which can be implemented using tech-
nologies such as Enterprise Information Integration (EII).
2. Enterprise Service Bus (ESB): The ESB is an abstraction layer
which can be used to develop a set of services that allow access
to the data without the users knowing the details of where and
how it is stored. This approach does not require a common data
model as the data federation approach does, rather, each source
provides a service by exposing its features and communicating
with other services offered by other sources. The application
invoking the services does not need to know where the data is
stored, what the original source interface is and how the data is
stored at the source; it will only see an Application Programming
Interface (API) such as SOAP (Simple Object Access Protocol)
or REST (Representational State Transfer) which is the mech-
anism used for sources to expose their services.
Physical vs. Virtual Integration
There are several factors to be considered when choosing between

physical and virtual integration. The physical integration approach is
considered to be update-driven as the data from the disparate sources is
loaded in advance and stored in the destination source for direct que-
rying and analysis. Additionally, when the source changes, the data
in the destination is refreshed periodically to reflect these changes.
On the other hand, the logical integration approach is viewed as
query-driven as the data is retrieved from the source on an as need
be basis. Each time a query is posed it is translated to queries on the

source where the relevant data is stored.
There are a number of advantages to the physical approach that have
made, for example, the DW Warehouse approach a popular choice for
integration. These benefits include the ability to design the DW to
ensure efficient query performance for strategic type queries. However,
because the data sources are designed to ensure the efficiency of trans-
actional type (e.g. SQL) queries, this design is not suitable for the ana-
lytic type queries. Keeping the two sets of stores separate allows each
to be designed to optimize the types of queries that are being posed
on them. This means that the local transactional type processing at
the source will be unaffected by the demands being placed on it by the
strategic queries and that the integrated source can be designed for the
specific needs of strategic analytical queries (e.g. data summarization).
Additionally, if the source is down for any reason, the data is still
available at the integrated source (i.e. there is some data replication).
Finally, when loading data into a physical store it is possible to include
additional data including data from external sources which can pro-
vide added value when integrated with the already existing internal
data and historical data which is not typically kept at the source given
that transactional query processing largely involves the updating of
the existing data. Updating the data means that the historical perspec-
tive of how the data changes over time is not typically captured at the
source. In the case of the DW approach, as new data is loaded into the
warehouse, rather than updating the existing data, it is appended to
the existing data thereby facilitating the historical perspective needed
for BI and analytics. The warehouse design facilitates the storing of
aggregated and summarized data which is what is primarily used for
BI and analytics; thus precomputing and storing the aggregated data
improves the efficiency of the analytics. Finally, another major benefit
of the physical integration approach is that in the development of the
destination source, especially in the case of the DW, the data that
is loaded would have gone through an extensive cleaning process to
remove data quality issues and a transformation process to ensure the
data is in a format suited for analytics. This means that the destination
source is a clean, integrated, transformed, summarized data set that
is suited for analytics purposes. In the case of logical integration, the
data is extracted from the source and can have quality issues which
must be addressed on a need-be basis with a logical approach and can

become difficult to manage.
However, even with all the benefits the physical approach offers,
there are some major drawbacks which make it infeasible for certain
types of organizations. The most obvious is that there is an additional
physical store of data and so the storage needs increase significantly
especially as the integrated data store includes historical data. Another
issue with the physical approach is that the data in the integrated
source may not always be reflective of the data at the source. This is
due to the fact that the Extraction Transformation and Load (ETL)
process is done periodically and therefore the integrated data source
does not always have the most up-to-date data. In the logical approach
this issue would not arise, data that is needed for processing a query is
retrieved from the source at the time requested. The DW approach is
also seen as complex, rigid, costly, time consuming and resource inten-
sive and there are many cases of failed DW initiatives even with sig-
nificant initial investments. Therefore, the physical integration option
is seen as impractical for some organizations, particularly Small and
Medium-Sized Enterprises (SMEs). Thus, lower cost, less resource
intensive, more agile approaches have been sought which gave rise to
a number of alternative approaches, such as logical data integration.
Organizations, such as SMEs, that are looking to these alterna-
tives as options can also consider the Open Source solutions that are
now available. These Open Source solutions require a smaller initial
investment of resources and are extremely beneficial for PoCs that can
be used to demonstrate the importance and power of data integration
as they provide a quick turnaround without a significant cost.
Selecting a Data Integration Approach
Given the number of approaches that are now available for Data
Integration, an organization needs to be deliberate in determining
which option is most appropriate for its needs. An example of the
functionality for Data Integration is illustrated in Figure 7.2.
This target functionality can be achieved by:
1. Evaluating available Open Source platforms using an estab-
lished Open Source Maturity Model to select an appropriate
Data Integration technology solution.
Figure 7.2 A high level architecture of an agile BI system.
2. Evaluating the performance of the Data Integration platform

connectivity to core organizational data sources to deter-
mine feasible configuration options ranging from traditional
Extract, Transform and Load (ETL) to virtual integration.
3. Evaluating various data management functionalities, such as:
a. Data Integration: Federation, Interface handling,
Transformation, Profiling.
b. Data services delivery: Model-driven Metadata services,
Web service APIs for data access, Master Data Management.
There are a number of options for Open Source tools (e.g. Talend and
Pentaho) that can be evaluated. The next section provides a summary
of some of the features that are seen in these platforms. The features
are classified based on the following themes and are displayed in
Table 7.2.
• Data Integration – Ability to connect data from multiple sources.
• Data Management – Understanding, organizing and
storing data.
• Data Delivery – Flexible delivery of data.
• Technical Infrastructure – Servers, networks and software.
10 0 BUSINE S S IN T EL LI G EN C E F O R SM E s
Table 7.2 Themes and Features for Managing Data

CATEGORY/THEME PRODUCT FEATURE
Data integration Data integration: connect to multiple data sources (traditional,
Big Data & NoSQL)
Data integration Data integration: support Extract Transform and Load (ETL),
data federation
Data integration Dynamic data modelling & integration: easy-to-use, graphical,
drag-and-drop tools to build processes and transformations
Data management Metadata management: model-driven metadata services &
directory search using a single metadata repository
Data management Data quality/data profiling: tools that profile and cleanse data,
parse and standardize data, and match, merge, etc.
Data management Master data management: create a unified view of information
and manage that master view over time
Data delivery Data as a service: deliver data to consumers via Web service
APIs RESTful/SOAP)
Data delivery Reports and dashboards: pre-built and customizable reports/
dashboards that show key data quality metrics
Technical infrastructure Multi-platform runtime support: ability to deploy to Linux-based
and windows systems, cloud and virtualization environments
Technical infrastructure Web-based monitoring: ability to monitor resources and
deployments from any browser
Recommendations for SMEs
The design and implementation of the classic data warehouse archi-

tecture that underpins many BI Systems can be technically and
financially daunting for organizations without the requisite technical
resources. Smaller organizations in particular are often challenged by
the complexity and resource requirements of traditional BI solutions.
New data integration techniques combined with cloud services pro-
vide more flexible and more agile options for SMEs.
The data-as-a-service platforms improve the efficiency of accessing
existing intermediary stores as the web services allow for on-demand
access to data, thereby eliminating the need for the input of an IT
data specialist. For example, a web service can expose the data in a
data store so that an analyst can access it directly. Additionally, the
tools for data integration enable analysts to create data mashups (i.e.
integration of two or more data sets in a single graphical interface)
themselves, without relying on a data specialist. This means that the
existing data can be used more flexibly, thus increasing its use.
Using a tool to connect to a diverse set of sources and that allows

easy extraction and manipulation of data directly from the core data
sources increases an organization’s ability to be agile. This ability to
extract data, using a simple web service, will greatly improve the effi-
ciency of the analysts as they can now access data as needed, rather
than having to wait until it is moved to an intermediary store.
Conclusion
Data integration is a fundamental issue for any organization consid-

ering investing in BI initiatives. While the importance of this step is
recognized, the actual commitment, in terms of time and resources,
required to carry it out successfully is often underestimated. The suc-
cess of data integration is highly dependent on data quality and gover-
nance. Any issues related to poor quality and governance, will reduce
the integration opportunities and increase the cost and resources
needed to perform it. The application of the IM assessment as the first
step of the methodology will provide some important insights in terms
of the readiness of the organization for effective data integration.
Diverse options and the tools available for data integration allow
SMEs and similar organizations that were previously disadvantaged
due to the investments required by traditional methods, to now con-
sider integration as an important part of their portfolio as they adopt
BI. The newer opportunities also provide a more flexible and agile
approach to integration which is important given the environment in
which these organizations are operating.
Part III
A B lueprint
for A ction
This section describes the development of the Business Intelligence

(BI) roadmap, which identifies the activities that organizations
should focus on if they hope to become serious contenders in the
strategic BI space.
Chapter 8 describes how the outcomes and findings from the
preceding steps of AIMS-BI are integrated into a strategic BI r oadmap.
This strategic BI roadmap reflects a balanced portfolio of BI initiatives
and a coherent implementation sequence, and provides a sound basis
for ensuring that BI initiatives are aligned to organizational strategic
priorities.
Chapter 9 addresses the systematic means by which organizations
realize value from their information assets and BI initiatives. It begins
by emphasizing the need for them to estimate the business value that is
embedded in their data assets. Once organizations ascribe a business
value to their data assets, it becomes easier for executive management
to appreciate and elevate the strategic importance of managing these
assets. Thereafter, Benefits Realization Management (BRM) can be
employed as a collective set of processes, practices, and tools that help
the organization to increase the likelihood that benefits are realized
from BI and other ICT-enabled initiatives. Finally, in the epilogue,
we highlight, based on our experience, some of the factors we con-
sider to be essential for the successful implementation of AIMS-BI
in SMEs and those activities which are relevant to the needs of these
enterprises but that have not been considered in more traditional BI
approaches.
10 3
8
D e v elo pin g a Road map
fo r S tr ateg i c B usines s
I ntelli g en ce
If you don’t know where you are going, any road will get
you there.
Lewis Carroll
Introduction
This chapter describes how the results of the preceding steps of

AIMS-BI are integrated into a strategic Business Intelligence (BI)
roadmap. Both the IM maturity assessment and the PoCs are integral
to the development of the strategic BI roadmap as both identify key
areas on which the organization needs to focus and develop capabili-
ties if it is to maximize the benefits of its investment in strategic BI.
The preceding steps also ensure that the roadmap developed through
AIMS-BI is tailored to each particular organization’s strategic BI
requirements. Figure 8.1 identifies the various activities of AIMS-BI
that inform the strategic BI roadmap.
The roadmap provides a balanced portfolio of BI initiatives; these
initiatives include technology and analytics projects, as well as capac-
ity building/training, creating organizational structures/roles and
policy initiatives. The initial roadmap sequencing reflects the concur-
rent need for both developmental and applied initiatives that address
critical gaps identified from the IM maturity assessment, while gen-
erating early returns from existing or enhanced capabilities.
This strategic BI roadmap reflects a balanced portfolio of BI
initiatives and a coherent implementation sequence, and provides a
sound basis for ensuring that BI initiatives are aligned to organiza-
tional strategic priorities. This chapter highlights the importance of
10 5
Figure 8.1 Interventions to inform the strategic BI roadmap.
a strategic BI roadmap in placing the organization in a position to

secure executive commitment and resources and, in so doing, put-
ting the organization on the path to becoming a serious “Analytics
Competitor”.
Information Management (IM) Maturity

Assessment (Capability Gaps)
The IM maturity assessment report is a key input into the devel-

opment of the strategic BI roadmap. As described in depth in
Chapter 3, the assessment selects six domains across an organization:
People/Organization, Policy, Technology, Compliance, Measurement and
Process/Practice, and assesses various functional competencies (e.g. Data
Quality Strategy, Data Ownership, Executive Sponsorship) within each
of these domains. The breadth of the coverage of the IM maturity
assessment is representative of the scope of the IM capability required.
An example of the results of the assessment are visually represented in
Figure 8.2, using the IBM Data Governance Council (IBM-DGC)
reference model. This representation has the advantage of highlight-
ing the interconnectedness and path dependencies between the vari-
ous aspects of IM maturity and the most critical capability gaps to
be addressed. For instance, Data Quality Management, Information
Lifecycle Management and Information Security and Privacy are the
D E V EL O PIN G A R OA D M A P F O R S T R AT EGI C BI 10 7
Figure 8.2 Information maturity assessment – IBM-DGC model perspective.
three core disciplines required for organizations to be able to real-

ize desirable value outcomes from BI. However, having key enablers,
such as Organizational Structures and Awareness, Policy, Strategic
Alignment and Stewardship, in place will enhance the effectiveness of
the core disciplines. Based on the IM maturity assessment, initiatives
are identified and included in the BI roadmap.
Proof of Concept (PoC) Prototypes
The BI (PoCs) prototypes highlight the potential for the application

of BI analytics techniques on the significant data resources and assets
owned by the organization. The PoCs deliver working prototypes that
provide the basis for the accelerated deployment of production appli-
cations which, with incremental development effort, could become
valuable components of the organizations analytics portfolio.
Data Quality
Data Quality (DQ ) emerges as a recurrent issue in many organiza-

tions and is perceived by organizations as a constraint on the effi-
ciency and effectiveness of various data-driven analytics initiatives and
decision-making processes. The effectiveness of the analytics PoCs is

often limited by several aspects of DQ. Other than the obvious issue of
accuracy (i.e. degree to which data values are correct), inconsistent data
formats, missing values, same field names with different meanings
from different data sources, and the identification of logical relatedness
between specific data instances based on values are all issues that may
be encountered when developing the prototypes. Initiatives to address
the DQ issues identified must be included as part of the roadmap.
Chapter 6 discusses the benefits of a systematic approach to DQ stan-
dards definition and DQ measurement. The use of a DQ Dashboard
can help to empower data stewards in the formulation and oversight
of DQ standards and policies across the business. The principal DQ
and Governance lesson to be learned is the importance of systemic
approaches to measurement, ranking and quantification of DQ and
asset value, so that the organization can begin to better manage and
maximize returns on its data investments. Thus, DQ and Governance
is likely to become a critical aspect of the BI roadmap.
Data Integration
The Data Integration stage (see Chapter 7) demonstrates the value of

utilizing a specialized data integration platform. Data Integration at
the prototype development stage can be performed using open source
options (e.g. Open Source community edition of the Talend Open
Studio Platform). The use of open source options at this stage allows
for rapid development without the need for significant investment in
resources. However, when considering a full-blown implementation,
a subscription-based Enterprise Platform might be more suitable for
the Data Integration needs identified. The organization should con-
sider deploying this option to enable more systematic approaches to
the logical and physical definition and standardization of data across
the enterprise and the on-demand delivery of data services to a range
of information consumers.
Any platform should also support (i) the Data standards
Cataloguing (ii) Metadata Management and (iii) DQ Management
(Profiling/Dashboard) functionalities identified in Chapter 6 in order
to facilitate the formalizing of Data Governance/Stewardship roles,
procedures and supporting mechanisms.
D E V EL O PIN G A R OA D M A P F O R S T R AT EGI C BI 10 9
Architecture of a Business Intelligence (BI) Roadmap
Gartner emphasizes the complex orchestration of program manage-

ment, technology, institutional mechanisms and skills required for the
strategic use of BI. They also underscore the need for a framework that
defines the layers and components that are to be integrated and aligned
in order to deliver a more strategic vision and plan for implement-
ing BI, analytics and performance management (PM) initiatives (see
Figure 8.3). The Gartner Business Analytics framework,1 is a high-level
framework that provides a practical basis for representing an enterprise
BI Roadmap, is used to situate the various BI initiatives derived from
applying the AIMS-BI methodology. The BI roadmap will be guided
by the following summarized adaptation of Gartner’s recommendations
on the use of this framework:
• The framework should be used to develop a strategy and an
implementation plan, and to highlight key decisions, integra-
tion points, gaps, overlaps and biases;
• A portfolio of IM, analytic and decision-making capabili-
ties will be needed to meet the diverse requirements of the
organization;
Figure 8.3 The Gartner business analytics framework.
1 https://gartner.com/imagesrv/summits/docs/na/business-intelligence/gartners_
business_analytics__219420.pdf.
• Formal Program Management (PM) will be required to bal-

ance resources and investments across multiple BI, Analytics
and PM initiatives and projects;
• BI initiatives are best suited to iterative development and
should be driven by a BI competency center (BICC) that
gathers requirements, prioritizes needs and delivers solutions
in phases.
The components of the Gartner Framework and their application to
an organization’s BI roadmap are as follows:
Performance
PM links the strategic goals of the business with the execution of

these goals and seeks to align operational activities and processes with
an appropriate enterprise metrics framework. The explicit statement
of strategic priorities and key performance indicators (KPIs) in the
BI Roadmap help to ensure that BI initiatives are aligned to organi-
zational strategic priorities. During Steps 2 and 3 of AIMS-BI the
selected PoCs are aligned with the strategic objectives of the orga-
nization, as identified by key stakeholders. As well as the explicit
statement of strategic priorities and KPIs, an Enterprise Performance
Management system/Dashboard is one of the recommended early BI
initiatives that should be contained in a BI Roadmap, as it will help
executives to measure and track the impact and effectiveness of the BI
program as it is rolled out.
People
This layer of the framework addresses BI initiatives that focus on the

development and institutionalization of people capabilities and roles
based on the task perspectives of producers, consumers and enablers:
Producers – Domain experts/business analysts who analyze
data, specify and develop analytic models, produce reports
and dashboards, provide insights into business performance
and support the decision-making process. Effective data
integration and the availability of data self-service plat-
forms make the process of acquiring relevant data easy and
D E V EL O PIN G A R OA D M A P F O R S T R AT EGI C BI 111
efficient and allow producers to conduct analytical tasks inde-

pendently of IT support. The producers need knowledge and
expertise across a wide competency spectrum ranging from
technical skills associated with modeling and analysis tools
to a deep understanding of business issues and their corre-
sponding performance impact, along with good communi-
cation skills.
Consumers – Persons within the organization who consume
the results and insights of analytics to either make deci-
sions or take actions. Consumers are instrumental in ensur-
ing that the results of analytics are operationalized in the
institution; they include operational workers, such as sales
representatives who are making the day-to-day decisions
within a particular product domain; functional manag-
ers making operational and tactical decisions; and senior
executives responsible for strategic decision-making. BI
initiatives targeting consumers include function-specific
performance dashboards and scorecards as well as decision-
support applications and tools.
Enablers – The information technology professionals within
the organization and external supporting vendors and/or
consultants who are responsible for designing, building and
maintaining the platforms and systems used by producers
and consumers. The IM maturity assessment will help to
identify areas in which the organization needs to invest, in
terms of the technical platforms and capabilities, in order
to more effectively deliver the requisite IM disciplines and
services, e.g. Data Architecture, Metadata Management and
Data Quality Management. However, the requirements for
enablers go beyond technology competencies to encompass
subject area domain expertise and business communications
skills. Among the enablers should be someone with the
designated role of Data Steward who is provided with the
supporting structures and processes required to establish
and maintain data quality accountability across the busi-
ness. It is unlikely that all these capabilities will reside in
a single individual; therefore the organization may need to
consider the establishment of a BI Competency Center to
consolidate internal core BI expertise and capabilities to be

augmented, where n ecessary, with resources from outside
the organization.
Process
The Gartner framework highlights three interconnected groups of

processes: decision processes, analytical processes and information gov-
ernance processes. The Information governance processes enable the
analytical processes through supporting mechanisms such as data
governance, data stewardship, data integration and data quality
management which are essential for BI initiatives to be successful.
The business impact of the analytical processes is realized through
effective decision processes, which ensure that the outputs of the ana-
lytics are deployed to enhance decision-making and business value.
Several of the BI initiatives in the roadmap are likely to focus on
improved information governance and analytical processes. A standard-
ized template for specifying analytics initiatives will ensure that key
questions are addressed when formulating BI initiatives. This tem-
plate includes explicitly defining a deployment strategy as part of
each analytic initiative as this will enhance the effectiveness of the
associated decision processes.
Platform
The platform layer integrates the technology required to manage

data and information with the tools to build analytical capabilities
as well as the technology that improves decision-making capabilities.
The information capability layer is foundational and the data integra-
tion and data quality PoCs, which in many cases will be identified
as mandatory, will inform key BI platform initiatives. Built on top
of this layer is the analytical capability layer which includes descrip-
tive, predictive and prescriptive analytics, several of which are iden-
tified and evaluated during Steps 2 and 3 of AIMS-BI. Applying
the insights and output of BI and analytics will require technology
enablers to work closely with information producers and consumers in
order to understand and model the business decisions. This end-to-
end BI process requires the integration of information capabilities,
analytic capabilities and decision capabilities to facilitate acquiring

good quality data, performing advanced analytics and enabling the
sales teams to target the right customers with the right products.
Example Roadmap – Portfolio of BI Initiatives
An example portfolio of BI initiatives and strategic BI roadmap for a

generic financial institution is represented in Figure 8.4. This road-
map, which is derived from applying the steps of AIMS-BI, provides a
balanced portfolio of BI initiatives that includes technology and ana-
lytics projects, as well as capacity building/training, organizational
structures/roles and policy initiatives. Key features of this derived BI
roadmap include:
• Enterprise scope, incremental execution: The BI portfolio
reflects a full spectrum of initiatives that will impact all
areas of the organization, including the IT division, from
branch operations to the various subsidiaries, and from
executive management to line staff. However, it is designed
in such a way that its execution can be planned incremen-
tally based on the organization’s absorptive capacity for
change and without overcommitting financial or human
resources.
• Bootstrap initiatives: The completed PoC prototypes pro-
vide the basis for the accelerated deployment of BI initiatives
that can lead to important early wins and demonstrable value
in the execution of the roadmap.
• Development, leverage and application of BI capabili-
ties: The roadmap provides a mix of development and appli-
cation initiatives that allows for concurrently addressing
critical gaps identified from the IM maturity assessment
while generating early returns from existing or enhanced
capabilities.
• Intrinsic Progress Monitoring: The built-in annual IM
maturity assessment provides a systematic means by which
the organization measures improvements in its IM and BI
capabilities. The Enterprise Performance Management
System/Dashboard is likely to be one of the BI initiatives
Figure 8.4 Example of a strategic roadmap.

recommended for early implementation based on the BI

roadmap. It will help executives to track the realized business
impact and effectiveness of the BI program as it is rolled out.
Navigating the BI Roadmap
The organization’s strategic BI roadmap (Figure 8.4) is designed

as a dynamic portfolio of initiatives that covers an initial timespan
(e.g. 18 months) and is amenable to enterprise-level Project Portfolio
Management (PPM) best practices. The association of each initia-
tive with the organization’s stated strategic priorities allows for the
dynamic allocation of financial and human resources in alignment
with shifting or emergent business priorities.
The initial roadmap sequencing reflects the concurrent need for
both developmental and applied initiatives that address critical gaps
identified by the IM maturity assessment, while generating early
returns from existing or enhanced capabilities. Based on each annual
scheduled maturity assessment, the roadmap can be recalibrated
to reflect developmental progress in addressing maturity levels
across the dimensions of People/Organization, Policy, Technology,
Compliance, Measurement and Process/Practice. It is important that
the organization maintains a balanced effort across the five layers
reflected in the roadmap until the initial portfolio of initiatives is
completed.
The organization’s team that is assigned responsibility for the
execution of the BI roadmap should undertake a more detailed
profiling of each of the individual initiatives in terms of estimated
costs and projected benefits, manpower/expertise resource require-
ments and technology dependencies. This will facilitate more effective
PPM and ultimately help the organization to realize an effective BI
strategy, develop core BI capabilities, align with any ongoing changes
in strategic priorities and optimize its BI investments.
Conclusion
A strategic BI roadmap provides a big picture of where the organi-

zation wants to go in order to achieve its strategic objectives. The
development of the BI roadmap takes into consideration not just the
organization’s priorities but also the information management gaps,

along with industry practices, and the lessons learned from the devel-
opment of the PoCs. Each phase of AIMS-BI identifies key areas that
the organization should focus on and this culminates in a strategic BI
roadmap that reflects a balanced portfolio of BI initiatives, a coher-
ent implementation sequence, and provides a sound basis for ensuring
that BI initiatives are aligned to organizational strategic priorities.
9
C re atin g B usines s Value
from D ata A s se ts
[Organizations] most valuable asset, although not shown in the

financial statements, is data.
PwC1
Introduction
Data has become such an important asset and organizational resource

in the digital economy, that it is now being regarded as the new fuel or
the new raw material of the 21st century. But data as an asset has some
unique attributes that distinguish it from other traditional physical
assets. Specifically, unlike natural raw materials, data is not dimin-
ished when it is consumed. Indeed, the value of data assets increase in
direct relationship to the number of people who are able to make use
of the asset.
One of the important steps in building the commitment to Strategic
Business Intelligence (BI) is to be able to demonstrate the value of the
data assets. Although important, it is often not done as it is difficult
to quantify the value of the data asset. More recently, there has been a
focus on this and a number of options have been proposed. This issue
of measuring the value of data assets will be outlined in this chapter.
Edward Wilson made this perennial assertion:
We are drowning in information, while starving for wisdom. The world
henceforth will be run by synthesizers, people able to put together the
right information at the right time, think critically about it, and make
important choices wisely.
Wilson (1999)
1 https://pwc.fr/fr/assets/files/pdf/2016/05/pwc_a4_data_governance_results.pdf.
117
Wilson’s “Synthesizers” forecast the ascendancy of today’s Analytics

Competitors, organizations such as Google, Amazon, Capital One
and Netflix that make widespread use of BI and data assets to create
significant business value. Across industry sectors, Synthesizers mani-
fest two key business attributes that have become significant enablers
of their competitive advantage and strategic ambitions.
• Quality Data Assets: The distinctive quality of the data pos-
sessed by the organization that cannot be easily replicated by
its competitors.
• Agility: The ability of the organization to rapidly sense and
respond to volatile environmental changes by maximizing the
use/leverage of its data assets.
Measuring Value of your Data Assets
While systematic methods and tools have emerged for assessing

and measuring data quality, consensus on standardized methods for
the valuation of data assets has continued to be elusive. In general,
data quality needs to be examined, planned and managed within a
larger Information Management context. The following definition
of Information Management is a useful one: “a program that man-
ages the people, processes, and technology in an enterprise toward
the goal of maximizing the leverage, value, and investment in data
and k nowledge, while avoiding increased risk and cost due to data
and knowledge misuse, poor handling, or exposure to regulatory
scrutiny”.2 This definition underscores the importance to organi-
zations of establishing a means of assessing the value of their data
assets in order to provide relevant context to the efforts and resources
allocated to data quality management activities. This is not, by any
means, a trivial task. Information theorists have long struggled with
the challenge of establishing a mechanism for the valuation of the
intangible organizational data, information and knowledge assets.
Gartner defines Infonomics3 as follows:
2 Ladley, J. (2005). Beyond the data warehouse: Plain English. DM Review,

September.
3 http://gartner.com/it-glossary/infonomics.
C RE ATIN G BUSINE S S VA LUE F R O M DATA A S SE T S 119
Infonomics is the emerging discipline of managing and accounting

for information with the same or similar rigor and formality as other
traditional assets (e.g., financial, physical, intangible, human capital).
Infonomics posits that information itself meets all the criteria of formal
company assets, and, although not yet recognized by generally accepted
accounting practices, increasingly, it is incumbent on organizations
to behave as if it were to optimize information’s ability to generate
business value.
Infonomics is the emerging discipline of managing and accounting for

information with the same or similar rigor and formality as other tra-
ditional assets and posits that information itself meets all the criteria
of formal company assets. While not yet recognized by generally
accepted accounting practices, increasingly, it is incumbent on orga-
nizations to behave as if it were to optimize information’s ability to
generate business value, in the same way that other more traditional
physical assets are acquired, valued, accounted for and leveraged.
Unfortunately, while some companies recognize and account for
the value of intangible assets, such as brands and intellectual prop-
erty, relatively few have any sense of the value of their data assets.
According to Gartner Research, 90% of surveyed executives list data
as one of their most important competitive assets, but fewer than 10%
of them actually quantify it.
Method of Information Value Assessment
Within the global Data ecosystem, analysts (e.g. McKinsey) estimate

a combined realizable value of over $20 trillion for Open Data, Big
Data, Digitization and IoT Data (Internet of Things). At the level of
the individual firm, methods for estimating the value of data assets
include:
Cost (Historical Cost): This is the traditional cost accounting
1.
approach to valuing assets based on how much was originally
paid to acquire the asset (purchase price and/or development
cost).
2.
Market (Current Cash Equivalent): Using this method, an
asset is valued based on how much other people or organiza-
tions are prepared to pay for it. One conceptual method for
this value assessment is to determine the market value of the

business (i.e. share capital), then assess how much an inde-
pendent third party (such as a potential buyer) would value
the organization if it was offered without the historical infor-
mation about its products, customers, staff and risk. Informal
estimates suggest this drop in market value could range from
more than 50% to less than 30%, depending on the infor-
mation intensity of the organization (consider for instance,
Amazon’s market value without its data assets).
Utility (Present Value): This method estimates the value of
3.
the data assets based on the net present value of expected
future economic benefits. Theoretically, this is the best indi-
cator of the value of data but it is often more practical to apply
this method in the cost-benefit analysis and justification of
discrete Analytics project initiatives rather than for assessing
the overall value of the enterprise information assets.
We used the second method (market value) with several companies to
develop theoretical estimates of their data assets. Figure 9.1 illustrates
the computations for a hypothetical company on the local Jamaica
Stock Exchange (JSE) Junior Market. Company X’s business model
is heavily dependent on data as a key business resource, so we estimate
their data assets to be worth 25% of the market value.
This exercise helps organizations to put a tangible value to initia-
tives such as systematic data quality management to arrest the rate
of information value depreciation. Even with conservative estimates
Figure 9.1 Estimating the value of the company’s data assets.

C RE ATIN G BUSINE S S VA LUE F R O M DATA A S SE T S 121
of 20% or less of the business’ market value, this hypothetical valua-

tion provides an important point of reference that will help to do the
following:
1. Increase Awareness: Increasing awareness of the value of data
as an organizational asset, which may lead to commensurate
attention being given to its management and leverage in com-
parison to more modestly valued physical assets.
2. Improve accountability: Measuring the value of data will give
greater impetus to the institution of roles and responsibilities
such as data stewardship/custodians.
3. Cost-justification of BI Initiatives: There is typically greater
difficulty in cost-justifying BI projects relative to more
conventional operational systems. An information valuation
paradigm is much more appropriate for such systems.
Managing the Value of your Data Assets
Once organizations ascribe a business value to their data assets, it

becomes much easier for Executive Management to appreciate and
elevate the strategic importance of improved Data Governance prac-
tices because the loss of business value due to poor data quality and
ineffective data use become quantifiable business outcomes rather
than anecdotal narratives.
The components of such a systematic Data Governance program
could include the following:
• Data Standards catalog & Metadata Management: Establish
a Metadata catalog of all the key Data and Information Assets
across the business to provide a centralized, searchable elec-
tronic repository of assets, not dissimilar to any well-managed
financial assets register.
• Data Quality Management: Establish a formal Data Quality
Management process for Enterprise Data, facilitated by an
Enterprise Data Quality Platform that integrates all of the
requisite functions, including profiling, cleansing, standard-
ization and Data Quality Dashboard functions to monitor
data quality metrics and support data governance initiatives
(see Chapter 6).
• Establish Data Stewardship: Institutionalizing Data

Stewardship by defining the necessary roles, procedures, sup-
port and accountability mechanisms to identify and empower
Data Stewards to help formulate and oversee adherence to
Data Quality standards and policies.
• Information Value Models/Metrics: Work on developing
economic models for the value of information and opportu-
nity costs of data quality in order to inform management on
how to assure returns on investments in data assets.
Returning to Edward Wilson’s prescient assertion, two decades hence,
the Synthesizers that run the world of business and commerce today,
are those that have accumulated enormous repositories of data assets
coupled with distinctive analytic capabilities that cannot be readily
reproduced by their competitors.
Benefits Realization Management (BRM)
Like any other business asset, estimating and ascribing a value to data
assets helps organizations to determine the appropriateness of mea-
sures for managing and protecting the asset. However this activity
alone does not provide a basis for realizing the future economic bene-
fits from the utilization of the asset. Benefits Realization Management
(BRM) is a collective set of processes, practices and tools that can help
managers to increase the likelihood that benefits are realized from BI
and other ICT-enabled initiatives.
BRM is a continuous process that includes investment decision-
making using cost-benefit analysis, project management, implementa-
tion, monitoring and continuous adjustment. It is a disciplined approach
that is based on the central tenets that technology alone cannot deliver
business benefits, and benefits rarely happen completely according to
plans or intent. Thus BRM prescribes a systematic process of manage-
ment activities that consists of the following stages (Figure 9.2):
Figure 9.2 Stages of management activities.

C RE ATIN G BUSINE S S VA LUE F R O M DATA A S SE T S 12 3
Benefits Identification: At the inception of each project ini-

tiative, business benefits are clearly identified and quantified
using cost-benefit analysis. A business manager is assigned
as the owner of each benefit, with full responsibility for its
realization. The collection of the planned benefits are usu-
ally captured and managed in a benefits register. In addition,
a Benefits Realization Roadmap is sometimes developed to
provide a visual illustration that shows when and how benefits
are expected to be enabled for the business over the course of
the project.
Benefits Analysis and Planning: Benefits are prioritized, and
metrics are developed to allow each Benefit to be monitored
and tracked throughout the project. It is important that the
expected realization of Benefits is integrated into the overall
program plan.
Benefits Delivery: This involves continuous monitoring of the
project implementation, and maintaining and updating the
Benefits register as the program progresses. The realization of
benefits or any changes to the benefits expectations should be
reported to the management team.
Benefits Transition/Sustainment: Once benefits are realized
through the project, ongoing responsibility for maintaining
the benefits should be handed over to operational management
to continue to monitor benefits performance and ensure that
the emergent results are continued. A Benefits Sustainment
Plan can be established to identify risks, processes, measures
and tools necessary to ensure the continued realization of the
benefits.
Ultimately the implementation of BI and other ICT-enabled initia-
tives provide organizations with enhanced capabilities. However, for
benefits realization to be effective and sustainable, business managers
must assume activist responsibility for orchestrating the key business
changes and processes necessary to take advantage of these capabili-
ties to realize the benefits.
Over time, Organizations need to reassess their Information matu-
rity capability to determine the impact of the BI initiatives within the
organization. In Chapter 3, we had discussed two key measures in
the Information maturity assessment namely value creation and data

usage quality. By tracking these measures over progressive assessment
cycles as they execute their BI Roadmap, business managers will have
a qualitative and quantitative assessment of whether the BI Initiatives
are delivering real business value for the organization.
Conclusion/Recommendations
The primary rationale for BI should be to realize maximum returns

from data assets. Returns may take the form of increases in opera-
tional efficiency, employee productivity, time to market, sales and
revenue as well as mitigating costs which may be realized as process
failures, opportunity costs, scrap and rework costs and the failure to
sense and act on business opportunities. Developing economic mod-
els for the value of information and opportunity costs of data quality
can be demonstrated in a PoC but should be extended and institu-
tionalized to provide the organization with systemic approaches to
measurement, ranking and quantification of data/information value
creation activities.
E pilo gue : L es sons
L e arned
Think big and don’t listen to people who tell you it can’t be done.
Life’s too short to think small.
Tim Ferriss
Small and Medium-Sized Enterprises (SMEs) must not be intimidated
by the notion of becoming data-driven organizations, rather they
should recognize the opportunity to become leaders in this space.
They should not be stymied by the perception that it is extremely
resource intensive to become data-driven. This books seeks to dispel
the perception by providing a methodology, AMIS-BI, that SMEs
can adopt and adapt, taking into account nuances that are specific to
their own enterprises.
Critical Success Factors
There is a plethora of case studies that describe the effect that strategic
BI initiatives have had on organizations and the areas in which these
initiatives have been applied. These case studies identify a number of
factors that have been critical to the success of the BI initiatives, many
of which were supported by our own experiences working with SMEs.
12 5
12 6 E PIL O GUE: L E S S O NS L E A RN ED
Many organizations have invested significantly in BI initiatives

without fully understanding and implementing the critical success
factors; consequently they have not been able to reap the expected
benefits. A part of the roadmap must include a plan for ensuring that
the critical success factors will be in place to support the implementa-
tion of the BI initiatives. From the literature and our own experiences,
we have compiled the following list of critical success factors SMEs
must consider as they invest in BI initiatives:
1. The alignment of data-driven initiatives and strategic objectives.
2. Changing the mind-set of employees. Decision-making must
be based on data, and not, as previously, on intuition. So move
from a culture of intuition-based decision to one of data-
driven decision-making.
3. Identifying a champion at the senior level – We have found that
those organizations that have an influential champion are those
that have had the most success applying AIMS-BI. The role of
the senior level designated champion in ensuring stakeholder
engagement throughout the process can not be underestimated.
4. Communicating with all stakeholders. Stakeholders must be
kept up to date on the progress of AIMS-BI during its appli-
cation. The champion has a critical role here.
5. Data quality. This is foundational to BI success. Therefore,
unless the organization has previously done so, as part of the
BI roadmap a formal data quality process must be developed
to maintain the quality of the data.
6. Identifying the right team. A BI project is not an IT only
initiative, it requires a mix of team members including business
analysts, data analysts, BI analysts.
Based on our experience and feedback from applying AIMS-BI to

organizations, this methodology is suited to SMEs as it includes a
number of activities that are relevant to their needs but that have not
been considered in the more traditional BI approaches. These activities
include the following:
1. The IM assessment – This assessment has proved to be an
extremely useful first step in identifying the state of strategic
BI readiness of an organization as it helps to identify the areas
Epil o gue: L e s s o ns L e a rned 12 7
that need improvement if the organization is to maximize

the benefits of its BI investment. There is a close alignment
between the IM assessment areas and the critical success fac-
tors for BI implementation which means that the assessment
is measuring areas key to the success of BI. We found that
the self-evaluation is an important feature of this IM assess-
ment as it makes the stakeholders more open to the findings
than if an external assessor acting independently was to carry
this out. The stakeholders felt that this step validated their
own views on organizational weaknesses (e.g. data quality).
The results of the Information Maturity Assessment form
an important step for BI capability benchmarking and the
results form an important input to the development of the BI
roadmap.
2. Proof of Concepts (PoCs) and Prototypes. The PoCs
and, more specifically, the development of the prototypes
demonstrated to the executive management team and all
key stakeholders the business value of advanced analyt-
ics, systematic approaches to better data quality manage-
ment, and the agility of data as a service. The prototypes
also helped stakeholders to better understand the process
of implementing BI as a whole, and not just the analytics
component, as well the factors that are critical to its suc-
cess. All of this secured/reinforced executive commitment
to strategic BI.
3. Engagement. Engaging the senior executives and other key
stakeholders early in the process proved to be invaluable for
getting the buy-in for the project. It was interesting to see the
level of interest and commitment of the senior executives and
all other key sponsors.
4. Agility. Deployment agility was enabled through the use of
Open Source software. There are a number of benefits includ-
ing reduced lead-time, cost, time-to-value and, importantly,
increased opportunity for experimentation. The agile meth-
odology of AIMS-BI helps to ensure that the BI initiatives
are enterprise in scope, strategically aligned, executive-driven
and portfolio-based, which are all key success factors for such
initiatives.
Appendix: Case Studies
12 9
13 0
Academic Analytics for an Educational Institution

A leading higher educational institution in the Caribbean is faced with the challenge of how to get more value from its data and is considering investing in analytics.
Senior management and key decision-makers have expressed their concern about the lack of access to clean data that are essential to their critical decision-making
and they want to identify opportunities for the application of academic analytics.
AIMS-BI was applied to the academic institution
Business Needs Our Approach Client Results
• The institution wanted to identify academic • The Information Management (IM) assessment was administered. • AIMS-BI is useful for educational institutions
analytics opportunities. • Low maturity scores for policy and measurement signaled clear gaps in • AIMS-BI provides a systematic approach to academic
• They needed to identify their barriers to formal data governance mechanisms and the minimal use of best practice analytics and in so doing can maximize the benefits
strategic Business Intelligence (BI) adoption. behaviors such as data quality metrics and profiling/measurement. analytics can provide.
• The finding from IM assessment was corroborated by the qualitative • The PoCs must be aligned with the strategic objectives of
comments recorded from the various user group interviews. the institution.
• The main strategic objectives of the organization were also elicited. They • Capability gaps that are essential for successful analytics
were based on student experience, finance, efficiency and research. are identified from the IM assessment, plans to address
• A portfolio of initiatives was developed that included the development of a these become a part of the roadmap.
student life-cycle dashboard, the establishment of data stewardship, a • A significant amount of time and effort was spent to get
research productivity dashboard and data quality management processes. the data from the form in the heterogeneous databases to
App en d i x : C a se S t ud ie s
• Senior management was asked to prioritize those Proof of concepts (PoCs). a form suitable for the various modeling techniques.
• The student life cycle was chosen as the prototype to be developed. This highlighted the need for data standardization policies.
• This initiative was considered to be critical as it aligned with two key • The primary output of this process is the roadmap; however,
strategic objectives: other important outputs are the prototypes that do not require
• Improving academic and administrative process efficiency. much more effort for full deployment.
• Improving student experience. • The student life cycle provides a basis for managing and
• The Knowledge Discovery and Data Mining (KDDM) process model was used optimizing the student experience by tracking students as
to develop this prototype. they progress through the institution.
• A number of dashboards, that focused on application processing and
student throughput were developed.
Data-Driven Credit Risk Models for a Financial Institution
A leading financial institution in the Caribbean is interested in analyzing payments data to determine if this data can provide a further understanding of customer
behavior e.g. can the size and/or frequency of utility bill payments or retail transactions provide additional predictor variables for customer risk profiling? This
institution is interested in reviewing its credit card scoring models to determine if payments data can be used to improve the quality of these models. They think reliable
proxies from nontraditional sources could be used to determine the customers’ ability to repay and willingness to pay. Six sources were identified as reliable:
telecommunications providers, utilities, wholesale suppliers, retailers, government and the institutions’ own previously overlooked data.
AIMS-BI was applied in the institution and data driven credit risk models were identified as one of the top priority PoCs that needed to be
prototyped.
• The institution wants to consider other reliable • AIMS-BI was applied and in prioritizing the portfolio of PoCs, improving • These non-traditional sources proved to be important in
and high-quality data they have access to with credit risk models was identified as a priority. determining the customers’ ability to repay.
the aim of improving the credit card scoring • The key stakeholders were interviewed and their concerns about the existing • Analysis of the payments data can provide a further
models currently being used. scoring models were discussed. understanding of customers.
• These additional sources can be used as • Included business analysts from payments and credit risk divisions • Payment data can be used to provide additional predictor
reliable proxies for customers’ ability to repay. variables to improve the performance of credit-risk scoring
• Key decision-makers identified potential sources of data that could be used
• They need to build new credit card scoring as proxies. models.
models using variables from payments data. • This data represents behavior which can be a better predictor
• This data was profiled to determine its quality and transformed so that it
than the salary/profit data reported by persons.
could be integrated into the scoring models.
• Derived variables were created to build these models which
• The predictive modeling technique (decision trees) was used to build
require a good understanding of the domain knowledge and data.
profiles based on existing data.
• The institution recognized that they had issues with
• These prototype models were developed and verified with existing data.
organizational metadata and data quality and this led to the
inclusion of initiatives to address this being included in the BI
roadmap.
• The prototype developed was near implementation ready so in
the development of the BI roadmap, this was included as a
131
short-term activity.
13 2
Transactors and Revolvers for a Financial Institution

A leading financial institution in the Caribbean needs to understand the credit card portfolio at the customer level by identifying the attributes associated with
transactors (i.e. those that pay full amount of bill each month) and revolvers (i.e. those that pay a part of the amount with interest each month), and also identify the
profiles of revolvers who are likely to become delinquent.
AIMS-BI was applied in the institution and this came out as one of the top-priority PoCs that needed to be prototyped.
• The institution wants to determine whether • AIMS-BI was applied to this institution and in building the portfolio of PoCs, • Non-traditional sources proved to be important in
customers bill payment patterns determine this need to understand the credit card portfolio was identified as one of determining the customers’ ability to repay.
their propensity to either be a transactor or the top priorities. • Credit card portfolio management was previously being
revolver and to determine how to make their • Further discussions were held with the affected decision-makers to identify done at the portfolio level this model supports, describing
transactors into revolvers yet ensure that potential sources of data that could be used as proxies. the portfolio at the customer level as each customer can
revolvers will not become delinquent. be profiled.
• This data was profiled to determine its quality and transformed so that it
• The institution needs to consider other reliable would be useful for integration in the scoring models. • The organization identified three groups – transactor, normal
and high-quality data they have access to that revolver and delinquent revolver.
• Prototype models were developed and verified with existing data.
can improve the understanding of the credit
• The models can be used to determine the credit worthiness of
card portfolio.
a particular customer/potential customer whose credit risk is
• Institutions want to be able to determine how to not yet known.
make their transactors into revolvers yet ensure

• A great deal of time had to be spent in identifying the
that their revolvers are not likely to become
sources of the data needed to build this model.
delinquent.
• The prototype developed from this was near implementation
ready so in the development of the BI roadmap this could be
included as a short-term activity.
• The ability to identify customers’ propensity to pay can
personalize the interactions with them.
Market Basket Analysis for a Financial Institution
A leading financial institution in the Caribbean is focusing on building a sales culture to meet one of its strategic objectives of improving sales. The fact that some of
its products have moderate or low penetration suggests that both the sales strategy and products should be reviewed. Additionally, there is an awareness of repeat
customer take-up of related products, however, the evidence of how this functions is merely anecdotal. There is no systematic means of identifying related products or
directing sales personnel to offer product bundles.
AIMS-BI was used in the institution and this came up as one of the most important PoCs that needed to be demonstrated to the stakeholders.
• The focus of this customer analytics initiative • AIMS-BI was applied to this institution and this initiative was included in • Both association rules and sequential mining help to
was to improve marketing strategies. the portfolio of PoCs. increase the effectiveness of sales campaign management
• Maximize the customer value to the • The business leader and business analyst were interviewed to get a better and targeting processes.
organization by cross-selling/upselling various business understanding. • The concept hierarchies developed were extremely useful in
products to existing customers. • To improve the success rate of current targeted marketing campaigns, classifying products into subgroups which were then used in
• This will ultimately contribute to the bank’s market basket analysis was employed as a customer analytic technique. the development of the models.
revenue targets by improving the efficiency of • This technique uses association rule mining to identify the products that • The inclusion of demographic and product data together in a
sales initiatives customers currently purchase together, which can be used to identify basket was novel and facilitates the discovery of
bundles (i.e. those products that go well together) and therefore that should multidimensional rules and frequent patterns in buying
be marketed accordingly. products.
• Customers don’t necessarily buy financial products all at one time, • The data preparation phase is reliant on the knowledge and
therefore a basket contains products bought over time. experience of the data mining analysts and importantly their
understanding of the business objectives and the
• Sequential rule mining was used as it shows not only which products were
corresponding data required.
bought together but also the sequence in which they were bought.
• The prototype developed was near implementation ready so
• Concept hierarchies for the domain were developed to enable analysis at
this could be included as a short-term activity in the
different levels of granularity.
development of the BI roadmap.
• Numeric variables were discretized in consultation with financial business
analysts who provided the linguistic terms used to describe each of the
data ranges (e.g. age values were discretized to young, middle-aged, etc.).
13 3
Glossary
Agility: A methodology that promotes regular interactions amongst

stakeholders, building prototypes and responsiveness to
change.
AIMS-BI: Agile Integrated Methodology for Strategic Business
Intelligence.
Analytical Hierarchical Processing (AHP): A Multi-Criteria
Decision-Making technique that reduces complex decisions
to a set of evaluation criteria, and a set of alternative options
from which the best decision is to be made.
Analytics Competitors: Organizations that improve performance
through the application of data analytics to their business
processes.
Business Intelligence: Business Intelligence (BI) can be described as
a set of techniques and tools for the acquisition and transfor-
mation of raw data into meaningful and useful information/
knowledge for business analysis purposes.
Business Intelligence Competency Centre (BICC): A team of peo-
ple that develop the overall strategic plan and priorities for BI.
Capability Gaps: Assess the current capabilities of the enterprise and
identify the gaps that prevent it from meeting business needs
and achieving desired outcomes.
13 5
13 6 G l o s s a ry
Case Study: An evaluation method in Design Science research which

entails studying the designed artifact (i.e. methodology) in
depth in a business environment.
Corporate Data Warehouse: A central storage system for enterprise-
wide data.
Critical Success Factors: Capabilities or circumstances necessary
to enable a positive outcome for a business program or a
strategy.
Cross-Industry Standard Process for Data Mining (CRISP-DM):
A structured approach for implementing a data mining
project.
Data as a Service (DaaS): A platform for data integration and man-
agement on a Cloud.
Data Assets: Data that is expected to generate future revenues.
Network Data Delivery Services (NDDS): NDDS provides trans-
parent network connectivity and data ubiquity to a set of pro-
cesses possibly running in different machines.
Data Standards Cataloging: Data standards are the rules by which
data are described and recorded. In order to share, exchange
and understand data, the format as well as the meaning must
be standardized.
Data Stewardship: Data stewardship is the management and over-
sight of an organization’s data assets according to established
data governance practices.
Data Virtualization: Data virtualization integrates data from dis-
parate sources, locations and formats, without replicating the
data, to create a single “virtual” data layer that delivers unified
data services to support multiple applications and users.
Design Science Research Methodology (DSR): A set of guide-
lines for the development and evaluation of Information,
Communication and Technology (ICT) artifacts to address
real-world problems.
Enterprise Information Management (EIM): Enterprise informa-
tion management (EIM) is a set of business processes, disci-
plines and practices used to manage the information created
from an organization’s data as an enterprise asset. EIM is an
integrative discipline for structuring, describing and govern-
ing information assets across organizational and technological
G l o s s a ry 13 7
boundaries to improve efficiency, promote transparency and

enable business insight – Gartner.
Extraction Transformation and Loading (ETL): Extract, transform
and load (ETL) is the process of integrating data from mul-
tiple, typically disparate, sources and bringing them together
into one central location.
IBM-DGC Maturity Model (DGCMM): A model developed to
ensure consistency and quality control in governance through
proven business technologies, collaborative methods and best
practices.
Information Management (IM) Maturity Model: A model that pro-
vides a standardized tool for consistent point-in-time assess-
ment of the maturity of overall Information Management
(IM) capabilities within an organization.
Information Management (IM) Maturity Assessment: Assesses
the maturity levels for various organizational aspects related
to Information Management and thus provides a benchmark
for future assessments.
Key Performance Indicators (KPIs): A set of quantifiable measures
that a company uses to gauge its performance over time.
Knowledge Discovery and Data Mining (KDDM) Process Model:
A model that organizes Knowledge Discovery and Data
Mining (KDDM) projects within a common framework. This
helps organizations to understand the Knowledge Discovery
process and to provide a road map to follow while planning
and carrying out the projects.
Enterprise Metadata Management (EMM): The business discipline
for managing the metadata about the information assets of
the organization. Metadata is “information that describes
various facets of an information asset to improve its usability
throughout its life cycle.” (Gartner).
Maturity Model: A tool that helps management determine the cur-
rent effectiveness of a process and supports identifying the
capabilities they need to acquire next in order to improve per-
formance e.g. CMMI, OPM3.
MIKE2.0 IM QuickScan Tool (Methodology for an Integrated
Knowledge Environment): An open source methodology for
EIM that provides a framework for information development.
13 8 G l o s s a ry
Multi-Criteria Decision-Making (MCDM): A branch of opera-

tional research dealing with finding optimal results in
complex scenarios including various indicators, conflicting
objectives and criteria.
Organizational Capabilities: An organizational capability is a com-
pany’s ability to manage resources, such as employees, effec-
tively to gain an advantage over competitors.
Proof of Concepts (PoCs): It is a realization of a certain method or
idea in order to demonstrate its feasibility, or a demonstration
in principle with the aim of verifying that some concept or
theory has practical potential. A PoC is usually small and may
or may not be complete.
Prototyping: A draft version of a solution that allows for the explora-
tion of the solutions and demonstrates the intention behind
the solution before investing time and money into its full-
blown development.
Return on Investment (ROI): A financial ratio used to calculate the
benefit of the investment in relation to the investment cost.
Semi-Structured Interviews: A meeting in which the interviewer
does not strictly follow a formalized list of questions. The
questions are open-ended.
SMEs: SMEs are small and medium-sized enterprises that are typi-
cally classified according to employee count and/or annual
revenues. While there is no universal classification, what
is not disputed is that in most economies they account for
the majority of enterprises, contribute considerably to GDP,
export and employment. They produce a significant portion of
the private sector output.
Strategic Analytics: Detailed, data-driven, analyses supporting out-
comes of strategic importance for a company.
Strategic BI Roadmap: Describes a proposed path of Strategic BI
progression to meet the implementation of a delivery strategy.
It reflects a balanced portfolio of BI initiatives and a coher-
ent implementation sequence, and provides a sound basis for
ensuring that BI initiatives are aligned to organizational stra-
tegic priorities.
Bibliography
Aberdeen-Group. (2011). Agile BI: Complementing traditional BI to address

the shrinking decision window. Retrieved from http://montage.co.nz/
assets/Brochures/Aberdeen-Agile-BI.pdf
Anderson-Lehman, R., Watson, H. J., Wixom, B. H., & Hoffer, J. A. (2004).
Continental airlines flies high with real-time business intelligence. MIS
Quarterly Executive, 3(4), 163–176.
Apte, C., Liu, B., Pednault, E., & Smyth, P. (2002). Business applications of
data mining. Communications of the ACM, 45(8), 49–53.
Ariyachandra, T. & Watson, H. J. (2010). Key organizational factors in
data warehouse architecture selection. Decision Support Systems, 49(2),
200–212.
Becker, J., Knackstedt, R., & Pöppelbuß, J. (2009). Developing maturity
models for IT management. Business and Information Systems Engineering,
1(3), 213–222.
Bhambri, V. (2011). Application of data mining in banking sector. International
Journal of Computer Science and Technology, 2(2), 199–202.
Breslin, M. (2004). Data warehousing battle of the giants. Business Intelligence
Journal, 9(1), 6–20.
Brown, D. H. & Lockett, N. (2004). Potential of critical e-applications for
engaging SMEs in e-business: A provider perspective. European Journal
of Information Systems, 13(1), 21–34.
Chandler, N., Hostmann, B., Rayner, N., & Herschel, G. (2015). Gartner’s
business analytics framework. Gartner Inc.
Chapman, P., Clinton, J., Kerber, R., Khabaza, T., Reinartz, T., Shearer,
C., & Wirth, R. (2000). CRISP-DM 1.0.
Chen, H., Chiang, R. H. L., & Storey, V. C. (2012). Business intelligence and
analytics: From big data to big impact. Management Information Systems
Quarterly, 36(4), 1165–1188.
13 9
14 0 Bib li o g r a p h y
Cios, K. J., Teresinska, A., Konieczna, S., Potocka, J., & Sharma, S. (2000).
Diagnosing myocardial perfusion from PECT bull’s-eye maps-A
knowledge discovery approach. IEEE Engineering in Medicine and
Biology Magazine, 19(4), 17–25.
Cockburn, A. & Highsmith, J. (2001). Agile software development, the peo-
ple factor. Computer, 34(11), 131–133.
Cooper, B., Watson, H., Wixom, B., & Goodhue, D. (2000). Data ware-
housing supports corporate strategy at first American corporation. MIS
Quarterly, 24(4), 547–567.
Davenport, T. H. & Harris, J. G. (2007). Competing on Analytics: The New
Science of Winning. Boston, MA: Harvard Business School.
De Bruin, T., Freeze, R., Kaulkarni, U., & Rosemann, M. (2005).
Understanding the main phases of developing a maturity assessment
model. Paper Presented at the 16th Australasian Conference on Information
Systems (ACIS), Australia, New South Wales, Sydney.
Demirkan, H. & Delen, D. (2013). Leveraging the capabilities of service-
oriented decision support systems: Putting analytics and big data in
cloud. Decision Support Systems, 55(1), 412–421.
Eck, A., Keidel, S., Uebernickel, F., Schneider, T., & Brenner, W. (2014).
Not all information systems are created equal: Exploring IT resources
for agile systems delivery.
Elzinga, J., Horak, T., Lee, C.-Y., & Bruner, C. (1995). Business process man-
agement: Survey and methodology. IEEE Transactions on Engineering
Management, 42(2), 119–128.
Evelson, B. (2011). Trends 2011 and Beyond: Business Intelligence, Vol. 31.
Cambridge MA: Forrester Research, Inc.
Evelson, B. (2014). The Forrester Wave™: Agile business intelligence plat-
forms, Q3 2014. Forrester Research Inc.
Fayyad, U., Piatetsky-Shapiro, G., & Smyth, P. (1996). From data mining to
knowledge discovery in databases. AI Magazine, 17(3), 37–54.
Gregor, S. & Hevner, A. R. (2013). Positioning and presenting design science
research for maximum impact. MIS Quarterly, 37(2), 337–356.
Hannula, M. & Pirttimaki, V. (2003). Business intelligence empirical
study on the top 50 Finnish companies. Journal of American Academy of
Business, 2(2), 593–599.
Hevner, A. R., March, S. T., Park, J., & Ram, S. (2004). Design science in
information systems research. MIS Quarterly, 28(1), 75–105.
Highsmith, J. (2009). Agile Project Management: Creating Innovative Products.
London, UK: Pearson Education.
Hostmann, B., Rayner, N., & Friedman, T. (2006). Gartner’s Business
Intelligence and Performance Management Framework. Stamford, CT:
Gartner.
IBM. (2007). The IBM Data Governance Council Maturity Model: Building a
Roadmap for Effective Data Governance. IBM Software Group. Retrieved
from https://studylib.net/doc/8219376/the-ibm-data-governance-council-
maturity-model--building-a
Bib li o g r a p h y 141
Inmon, W.H. (2005). Building the Data Warehouse. New York: John Wiley &
Sons.
Jacobs, A. (2009). The pathologies of big data. Communications of the ACM,
52(8), 36–44.
Jourdan, Z., Rainer, R. K., & Marshall, T. E. (2008). Business intelligence:
An analysis of the literature. Information Systems Management, 25(2),
121–131.
Jukic, N. (2006). Modeling strategies and alternatives for data warehousing
projects. Communications of the ACM, 49(4), 83–88.
Katal, A., Wazid, M., & Goudar, R. H. (2013, August). Big data: Issues,
challenges, tools and good practices. 2013 Sixth International Conference
on Contemporary Computing (IC3), Noida, New Delhi (pp. 404–409).
IEEE.
Kimball, R. & Ross, M. (2011). The Data Warehouse Toolkit: The Complete
Guide to Dimensional Modeling. New York: John Wiley & Sons.
Knabke, T. & Olbrich, S. (2013). Understanding information system
agility—the example of business intelligence. Paper Presented at the 46th
Hawaii International Conference on System Sciences, Hawaii.
Kotelnikov, V. (2007). Small and medium enterprises and ICT, United
Nations Development Program-Asia Pacific Development Information
Program. Asian and Pacific Training Center for Information and
Communication Technology for Development.
Kurgan, L. A. & Musilek, P. (2006). A survey of knowledge discovery and
data mining process models. The Knowledge Engineering Review, 21(1),
1–24.
Lee, Y. & Koza, K. A. (2006). Investigating the effect of website quality on
e-business Success: An Analytic Hierarchy Process (AHP) approach.
Decision Support Systems, 42, 1383–1401.
Loveman, G. (2003). Diamonds in the data mine. Harvard Business Review,
81(5), 109–113.
Mansingh, G., Osei-Bryson, K.-M., & Asnani, M. (2015). Exploring the
antecedents of the quality of life of patients with sickle cell disease:
Using a knowledge discovery and data mining process model based
framework. Health Systems: Palgrave McMillan, 5(1), 52–65.
Mansingh, G., Osei-Bryson, K.-M., & Reichgelt, H. (2010). Using ontolo-
gies to facilitate post-processing of association rules by domain experts.
Information Sciences, 181(3), 419–434.
Mansingh, G. & Rao, L. (2014). Enhancing the decision making process: An
ontology based approach. Paper Presented at the International Conference
on Information Resources Management (Conf-IRM), Ho Chi Minh City,
Vietnam.
Mansingh, G., Rao, L., Osei-Bryson, K.-M., & Mills, A. (2015). Profiling
internet banking users: A knowledge discovery in data mining process
model based approach. Information Systems Frontiers, 17(1), 193–215.
March, S. T. & Smith, G. F. (1995). Design and natural science research on
information technology. Decision Support Systems, 15(4), 251–266.
Mariscal, G., Marban, O., & Fernandez, C. (2010). A survey of data min-
ing and knowledge discovery process models and methodologies. The
Knowledge Engineering Review, 25(2), 137–166.
Melon, M. G., Beltran, P. A., & Cruz, M. C. G. (2008). An AHP-based
evaluation procedure for innovative educational projects: A face-to-face
vs. computer-mediated case study. Omega, 36(5), 754–765.
Meredith, R., Remington, S., O’Donnell, P., & Sharma, N. (2012).
Organisational transformation through Business Intelligence: Theory,
the vendor perspective and a research agenda. Journal of Decision Systems,
21(3), 187–201.
Mettler, T. (2009). A design science research perspective on maturity models
in information systems tech. report BE IWI/HNE/03: Universität St.
Gallen.
Moin, K. I. & Ahmed, Q. B. (2012). Use of data mining in banking.
International Journal of Engineering Research and Applications, 2(2),
738–742.
Mooney, J., Beath, C., Fitzgerald, G., Ross, J., & Weill, P. (2003). Managing
information technology for strategic flexibility and agility: Rethinking
conceptual models, architecture, development, and governance. Paper
Presented at the International Conference on Information Systems (ICIS),
Seattle, Washington.
Muntean, M. & Surcel, T. (2013). Agile BI–the future of BI. Revista
Informatica Economică, 17(3), 114–124.
Negash, S. (2004). Business intelligence. Communications of the Association for
Information Systems, 13, 177–195.
Negash, S. & Gray, P. (2008). Business intelligence. In Handbook on Decision
Support Systems, Vol. 2, pp. 175–193. Berlin, Heidelberg: Springer.
Ngai, E. W. T. (2003). Selection of web sites for online advertising using the
AHP. Information and Management, 40, 1–10.
Olson, D. L., Dursun D., & Yanyan M. (2012). Comparative analysis of data
mining methods for bankruptcy prediction. Decision Support Systems,
52(2), 464–473.
Osei-Bryson, K. M., Mansingh, G., & Rao, L. (Eds.). (2014). Knowledge
Management for Development: Domains, Strategies and Technologies for
Developing Countries, Vol. 35. Springer Science & Business Media.
Overby, E., Bharadwaj, A., & Sambamurthy, V. (2006). Enterprise agility
and the enabling role of information technology. European Journal of
Information Systems, 15(2), 120–131.
Paranape-Voditel, P. & Deshpande, U. (2013). A stock market portfolio
recommender system based on association rule mining. Applied Soft
Computing, 13, 1055–1063.
Popovič, A., Hackney, R., Coelho, P. S., & Jaklič, J. (2012). Towards business
intelligence systems success: Effects of maturity and culture on analyti-
cal decision making. Decision Support Systems, 54, 729–739.
Rouse, W. B. (2007). Agile information systems for agile decision making. Agile
Information Systems, KC DeSouza, Editor, Elsevier, New York, 16–30.
Bib li o g r a p h y 14 3
Rygielski, C., Jyun-Cheng, W., & Yen, D. C. (2002). Data mining tech-
niques for customer relationship management. Technology in Society, 24,
483–502.
Saaty, T. (1980). The Analytic Hierarchy Process: Planning, Priority Setting,
Resource Allocation. New York: McGraw-Hill.
Sambamurthy, V., Bharadwaj, A., & Grover, V. (2003). Shaping agility
through digital options: Reconceptualizing the role of information tech-
nology in contemporary firms. MIS Quarterly, 27, 237–263.
Sein, M. K., Henfridsson, O., Purao, S., Rossi, M., & Lindgren, R. (2011).
Action design research. MIS Quarterly, 35, 37–56.
Sharma, S. & Osei-Bryson, K.-M. (2010). Toward an integrated knowledge
discovery and data mining process model. The Knowledge Engineering
Review, 25(1), 49–67.
Sharma, S., Osei-Bryson, K.-M., & Kasper, G. M. (2012). Evaluation of an
integrated knowledge discovery and data mining process model. Expert
Systems with Applications, 39, 11335–11348.
Steinmueller, W. E. (2001). ICTs and the possibilities for leapfrogging by
developing countries. International Labour Review, 140(2), 193–210.
Thompson, S. & Brown, D. (2008). Change agents intervention in e-business
adoption by SMEs: evidence from a developing country. AMCIS 2008
Proceedings, Paper# 242.
Thomsen, C. & Pedersen, T. B. (2009). A survey of open source tools for busi-
ness intelligence. International Journal of Data Warehousing and Mining
(IJDWM), 5(3), 56–75.
Trkman, P., McCormack, K., Valadares de Oliveira, M. P., & Ladeira, M. B.
(2010). The impact of business analytics on supply chain performance.
Decision Support Systems, 49, 318–327.
Van der Lans, R. (2012). Data Virtualization for Business Intelligence Systems:
Revolutionizing Data Integration for Data Warehouses. Waltham, MA:
Elsevier.
van Steenbergen, M., Bos, R., Brinkkemper, S., van de Weerd, I., &
Bekkers, W. (2010). The design of focus area maturity models. In:
Winter, R., Zhao, J. L., Aier, S. (eds.) Global Perspectives on Design Science
Research. DESRIST 2010. Lecture Notes in Computer Science, Vol.
6105. Berlin, Heidelberg: Springer.
Watson, H. J. (2009). Tutorial: business intelligence-past, present, and future.
Communications of the Association for Information Systems, 25(1), 39.
Williams, S. & Williams, N. (2006). The Profit Impact of Business Intelligence.
San Francisco, CA: Morgan Kaufmann.
Watson, H. J. & Wixom, B. H. (2007). The current state of business intel-
ligence. Computer, 40(9), 96–99.
Watson, H. J., Fuller, C., & Ariyachandra, T. (2004). Data warehouse gov-
ernance: Best practices at blue cross and blue shield of North Carolina.
Decision Support Systems, 38(3), 435–450.
Wilson, E. O. (1999). Consilience: The Unity of Knowledge, Vol. 31. New York:
Vintage.
Wilson, N., Summers, B., & Hope, R. (2000). Using payment behaviour
data for credit risk modelling. International Journal of the Economics of
Business, 7(3), 333–346.
Witter, M. & Kirton, C. (1990). The informal economy in Jamaica: Some
empirical exercises (No. 36). Institute of Social and Economic Research,
The University of the West Indies.
Index
A in data integration, 34–35, 91

operational, 3
Academic analytics, 130 AHP, see Analytical hierarchical
Agile Integrated Methodology processing (AHP)
for Strategic Business AIMS-BI, see Agile Integrated
Intelligence (AIMS-BI), Methodology for Strategic
11, 19, 20, 31, 32 Business Intelligence
application and findings, 30–31 (AIMS-BI)
case studies, 130–133 Analysis layer, 61
description of, 23–24 Analytical competitor state, 18, 19,
design of, 22–23 60–61
end user and stakeholder Analytical hierarchical processing
engagement, 22 (AHP), 27, 51–52, 56
IM assessment process, 26 Analytical processes, 112
Inmon vs. Kimball approach, Analytic dashboards, 65, 67
20–21 Artificial intelligence, 10
“knowing-doing” gap, 22–23, 38 Association rule induction, 69
opportunity discovery, 26–27
portfolio evaluation, 27 B
proof of concept prototypes,
28–29 Behavioral user experience, 65
strategic BI roadmap, 29 Benefits realization management
Agility, 118, 127–128 (BRM), 122–124
business, 2–3 Big data analytics, 10
business intelligence and, 3–5 Business agility, 2–3
14 5
14 6 In d e x
Business analytics, 1, 19 Data as a service (DaaS), 28, 31, 54,

Business intelligence (BI) 91, 100
definition of, 3 Data (as an) asset, 1, 19
introspection session, 26–27 management of, 76
roadmap, see Strategic business notion of, 13
intelligence roadmap value of, 117–124
Business intelligence competency Data consolidation, 95
center (BICC), 110, 111 Data delivery services, 28, 91, 100
Business understanding phase, Data-driven credit risk models, 131
IKDDM model, 70 Data-driven decision-making
culture, 17
C Data encoding, 66–67
Data federation, 95, 96
Call center traffic pattern analysis, Data governance, 13, 76, 78
53 Data integration, 54, 91, 100, 101,
Capability definitions, IM 108
assessment, 44–45 data from multiple heterogeneous
Capability gaps, 106–107 sources, 94
Capability Maturity Model data warehouse, 93
Integration (CMMI), 39 deliverables, 93–94
Case studies, 130–133 functionality for, 98–99
“Chasm of uncertainty,” 60 opportunities, 92–93
Classification, predictive modeling, physical integration, 95–98
69 recommendations for SMEs,
Cloud computing, 5, 9–10, 12 100–101
Column profiling, 83, 84 themes and features for managing
Continuous maturity models, 39–41 data, 99–100
Corporate DW, 20 virtual integration, 95, 96
Cost-justification of BI initiatives, Data life-cycle process review, 88
121 Data maintenance, 85–87
Credit decision analysis, 14 Data Management Group, 80
Credit risk analysis, 53 Data mart, 95
Critical success factors, 125–128 Data mining, 14
Cross-industry standard process for definition of, 68
data mining (CRISP-DM), descriptive techniques, 69
20 predictive techniques, 68–69
Customer segment value analysis, 53 Data preparation phase, IKDDM
model, 70–71
D Data profiling, 83–85
Data quality (DQ ), 31, 54, 73,
Dashboards 107–108, 126
data quality, 75, 86–87, 89 cost pyramid for, 79
visualization and, 65 dashboards, 75, 86–87, 89
In d e x 147
dimensions, 80, 82 DW, see Data warehouse (DW)

functionalities, 78 Dynamic data integration, 100
issues with, 76
management process, 79, 80, 121 E
assessment, 81–85
data life-cycle process review, E-business applications, 12
88 Enablers, 111–112
data maintenance/stewardship, Encoding (data), 66–67
85–87 End user engagement, 22
data standards definition and Enterprise ICT adoption, 5, 11–13
quality metrics, 79–83 Enterprise information management
proof of concept (DQ PoC), 88, (EIM), 38, 47, 48
89 Enterprise-level business
recommendations/observations, intelligence, 17–18
89–90 Enterprise performance
responsibilities for, 76–77 management, 110, 113
Data standard(ization), 46, 79–83, Enterprise service bus (ESB), 95, 96
92, 121 Evaluation phase, IKDDM model,
Data standards cataloging, 89, 121 71–72
Data stewardship, 46, 85–87, 90, Extraction, transformation and
122, see also Stewards loading (ETL) process, 28,
Data understanding phase, IKDDM 95, 98
model, 70
Data usage quality, 46, 94, 124 F
Data virtualization, 28, 91
Data visualization, see Visualization Financial institution
Data warehouse (DW), 20, 93, 95–97 data-driven credit risk models, 131
Decision-making, 32, 65, 73, 126 market basket analysis, 133
Decision processes, in Gartner transactors and revolvers, 132
framework, 112 Fixed-level maturity models, 39
Deployment phase, IKDDM model, Focus-area maturity models, 39–41
72
Descriptive techniques, 69 G
Design science research, 22
Design science research Gartner business analytics
methodology, 22 framework, 109–110
DGCMM, see IBM-DGC Maturity consumers, 111
Model (DGCMM) enablers, 111–112
Digital business trends, 1–2 people, 110–111
Digital economy, 1 performance, 110
Distribution channel optimization, 54 platform, 112–113
Distribution logistics, 14 processes, 112
DQ , see Data quality (DQ ) Graphical encodings, 66
14 8 In d e x
H evaluation phase, 71–72

modeling phase, 71
Historical cost, 119 Intrinsic progress monitoring, 113,
115
I Introspection session, 26–27
IBM Data Governance Council K

(IBM-DGC), 106, 107
IBM-DGC Maturity Model Key performance indicators (KPIs),
(DGCMM), 41 63, 64, 94
ICT adoption, barriers and Key quality indicators (KQIs), 86
challenges in, 11–13 Kimball approach, 20
IM maturity models, see Maturity Knowledge Discovery and Data
models Mining (KDDM), 20
Infonomics, 119 Knowledge economy, 9
Information governance processes,
112 L
Information management, definition Layers of analytics, 61–62, see also
of, 118 specific layer of analytics
Information management (IM) Logical data integration, see Virtual
maturity assessment, 24, data integration
26, 30–31, 37–38, 43, 48,
78, 127 M
BI roadmap development,
106–107 Machine learning, 10
capability levels, 44–45 Market-basket analysis, 13, 53, 54,
granularity of, 48 56, 69, 133
IBM-DGCMM dimensions, 47 Market segmentation, 53
technology assessment, 45–47 Market value of business, 119–120
Information systems maturity Master data management, 46, 73,
models, 38 90, 100
Information value assessment, Maturity models, 21–22
119–121 Capability Maturity Model
Information value models/metrics, Integration, 39
122 continuous, 39–41
Inmon approach, 20, 21 criteria for, 40
Integrated knowledge discovery and criticisms of, 48
data mining (IKDDM) IBM-DGC Maturity Model, 41
model MIKE2.0 IM QuickScan tool, 40
business understanding phase, 70 Organizational Project
data preparation phase, 70–71 Management Maturity
data understanding phase, 70 Model, 38
deployment phase, 72 in organizations, 39–40
In d e x 14 9
staged, 39 Physical data integration, 95–96

synthesized maturity model, 41–43 advantages, 97–98
Merchandise planning and drawbacks, 98
allocation, 14 update-driven, 96
Metadata management, 46, 89, 100, vs. virtual integration, 96–98
121 PoCs, see Proof of concepts (PoCs)
Micro, small and medium-sized Portfolio evaluation, 27, 51–52
enterprises (mSMEs), 11 opportunity discovery and,
MIKE2.0 IM QuickScan tool, 40 53–56
Mobile computing, 10 Practitioner-oriented maturity
Modeling phase, IKDDM model, 71 models, 40
Monitoring layer of analytics, 61, see Prediction layer of analytics, 61, see
also Visualization also Data mining
Multi-criteria decision-making Predictive life-cycle value, 14
(MCDM), 51 Predictive techniques, 68–69
Multi-platform runtime support, 100 Prescription layer of analytics, 61
Prioritization process, 51–52
N Prioritized portfolio, 49
Program management, 29, 109, 110
Netflix, 1, 13 Project manager, 31, 126
Proof of concepts (PoCs), 27, 31
O AHP ranking of, 56
data quality (DQ PoC), 88, 89
Open-source data integration development, 60–61
software, 28 layers of analytics, 61–62
Open source solutions, 98 portfolio of, 51–54
Operational agility, 3 process models, 72–73
Operational dashboards, 65 IKDDM model, 70–72
Opportunity discovery, 26–27, VPM model, 62–63
49–51 prototypes, 28–29, 59, 107, 127
and portfolio evaluation, 53–56
Opportunity Discovery Workshop, Q
27
Organizational capability, 2, 4, 39 Quality data assets, 118
Organizational Project Management
Maturity Model (OPM3), R
38
Reflective user experience, 65, 67
P Repeat clients, 14
Reporting level of analytics, 61
Payments analytics, 54, 56 Return on investment (ROI), 4
Performance management (PM), 18, Roadmap, see Strategic business
64, 109 intelligence roadmap
15 0 In d e x
S Synthesized IM maturity model,

41–43
Sales forecasting, 14 “Synthesizers,” 15, 117–118, 122
Sales performance management
dashboard, 54 T
Segmentation and credit risk of
semi-banked, 54, 55 Table profiling, 84–85
Semi-structured interviews, 26, 50 Tactical approach, 20
Social media, 10 Talend, 89
Software-as-a-service (SaaS), Technology assessment, 45–47
12–13
Staged maturity models, 39 V
Stakeholder engagement, 22, 31,
Value creation, 124
49–50, 126, 128
Value of data assets, 117–124
Stewards, 34, 75, 86, 90, 111,
Virtual data integration, 95, 96
122, see also Data
physical vs., 96–98
stewardship
query-driven, 96
Strategic approach, 20
Visceral user experience, 65–67
Strategic business intelligence
Visual analysis with Tableau
roadmap, 29, 105–106,
workshop, 54
115–116
Visualization
data integration, 108
business insights, 67
data quality, 107–108
and dashboards, 65
example of, 113–115
data encoding, 66–67
features of, 113, 115
data processing, 64
Gartner business analytics
decisions and actions, 68
framework, 109–110
definition of, 62
consumers, 111
design considerations, 65–67
enablers, 111–112
process models, 62–63
people, 110–111
understanding the business
performance, 110
context, 63–64
platform, 112–113
user experience, 65
processes, 112
Visualization process model (VPM),
IM maturity assessment report,
62–63
106–107
navigating, 115 W
PoC prototypes, 107
Strategic dashboards, 65 Web-based monitoring, 100
Strategic objectives of organization, Wilson’s “Synthesizers,” 15, 117–118,
52–54 122

Lila Rao-Graham (Author)_ Maurice L. McNaughton (Author)_ Gunjan Mansingh (Author) - Business Intelligence for Small and Medium-Sized Enterprises_ an Agile Roadmap Toward Business Sustainability-Auerb

Uploaded by

Copyright:

Available Formats

Lila Rao-Graham (Author)_ Maurice L. McNaughton (Author)_ Gunjan Mansingh (Author) - Business Intelligence for Small and Medium-Sized Enterprises_ an Agile Roadmap Toward Business Sustainability-Auerb

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lila Rao-Graham (Author)_ Maurice L. McNaughton (Author)_ Gunjan Mansingh (Author) - Business Intelligence for Small and Medium-Sized Enterprises_ an Agile Roadmap Toward Business Sustainability-Auerb

Uploaded by

Copyright:

Available Formats

Business Intelligence for

Small and ­Medium-Sized

Lila Rao-Graham, ­Maurice ­L. McNaughton,

© 2020 by Taylor & Francis Group, LLC

No claim to original U.S. Government works

Printed on acid-free paper

International Standard Book Number-13: 978-1-138-58421-1 (Hardback)

Trademark Notice: Product or corporate names may be trademarks or registered ­trademarks,

Visit the Taylor & Francis Web site at

and the CRC Press Web site at

The Need for a Strategic, yet Agile Perspective 19

Pa r t III A B l u e p r i n t for Action

Information Management (IM) Maturity Assessment

Drs Lila Rao-Graham, Gunjan Mansingh and Maurice L.

have spent a large part of their scholarly pursuits studying, assimilat-

high. But more importantly the emphasis of this p ­ ublication (which is

The book is tremendously well organized. It effectively introduces

Lila Rao-Graham is the deputy executive director and a senior

er research interests include Business Intelligence, Data Quality.

Maurice L. McNaughton is the director of the Centre of Excellence

is research interest spans the domain of emerging Open ICT ecosys-

Gunjan Mansingh is head of the Department and a Senior Lecturer

er research interests include Business Intelligence, Data Mining,

Digital Business Trends

Digital is rapidly becoming the dominant mode of interaction for

• Adopting a self-service paradigm for human resource services

Business Intelligence as an Enabler of Agility

Business Intelligence, Data Analytics, Business Analytics and Data

This definition situates BI as an organizational capability that enables

SMEs, Business Intelligence and Agility

Does this mean that the potential operational efficiency improve-

on their size, when contemplating the implementation of advanced

About this Book

This book was conceptualized out of the authors’ combined years of

The book is based on the Agile Integrated Methodology for Strategic

Business Intelligence (BI) has become a strategic initiative for many

traditional ICT technologies and methods may not be suited to their

A point of view can be a dangerous luxury when [used in place of]

Introduction (Digital Economy – Implications for Business)

traditional attendant capital and operating costs or manage-

SMEs and ICT in Developing Economies

While AIMS-BI is highly adaptable and designed to be broadly

SMEs and ICT Adoption – Barriers & Challenges

A number of factors have been identified as barriers to ICT adoption

of small firms are different from those of their larger counterparts.

alternative for SMEs seeking sophisticated ICT services at relatively

BI – Value Opportunities for SMEs

In our introductory chapter, we discuss our use of the term business

Distribution and Retail

Market basket analysis: It increases understanding of customers’

Sales forecasting: Examining time-based and location-based sales

Credit and Micro-Finance Services

Credit Decision Analysis: Default and bad debt can be enormously

We are drowning in information, while starving for wisdom. The world

Wilson’s “synthesizers” forecast the ascendancy of today’s Analytics

Art and science have their meeting point in method.

Organizations that aspire to become data-driven competitors by

even recognize that it is their own organizational weaknesses (e.g.

in this chapter provides a sequence of steps that will produce, as its

Why the Need for New Methodologies?

Small and Medium-Sized

Lila Rao-Graham, Maurice L. McNaughton,

Trademark Notice: Product or corporate names may be trademarks or registered trademarks,

high. But more importantly the emphasis of this p ublication (which is

Step 1: Information Maturity Assessment

require access to data sources in a timely manner, to experiment,