International Journal of Information Management 31 (2011) 460–468
Contents lists available at ScienceDirect
International Journal of Information Management
journal homepage: www.elsevier.com/locate/ijinfomgt
A proactive balanced scorecard
Panagiotis Chytas a , Michael Glykas b,∗ , George Valiris a
a
b
University of Aegean, Department of Business Administration, 8 Michalon Street, Chios 82 100, Greece
University of Aegean, Department of Financial and Management Engineering, Kountouriotou Street, Chios 82 100, Greece
a r t i c l e
i n f o
Article history:
Available online 1 February 2011
Keywords:
Performance measurement
Balanced scorecard
Fuzzy cognitive maps
Simulation
a b s t r a c t
This paper describes a methodology for the development of a proactive balanced scorecard (PBSCM). The
balanced scorecard is one of the most popular approaches developed in the field of performance measurement. However, in spite of its reputation, there are issues that require further research. The present
research addresses the problems of the balanced scorecard by utilizing the soft computing characteristics
of fuzzy cognitive maps (FCMs). By using FCMs, the proposed methodology generates a dynamic network
of interconnected key performance indicators (KPIs), simulates each KPI with imprecise relationships
and quantifies the impact of each KPI to other KPIs in order to adjust targets of performance.
© 2010 Elsevier Ltd. All rights reserved.
1. Introduction
Today, companies are evolving in turbulent and equivocal
environments (Drucker, 1993; Grove, 1999; Kelly, 1998). This
requires companies to be alert and watchful so as to detect weaknesses (Ansoff, 1975) and discontinuities in regard to emerging
threats and opportunities and to initiate further probing based on
such detections (Glykas, 2004). The strategic role of performance
measurement systems has been widely stressed in management
literature. These systems provide managers with useful tools to
understand how well their organisation is performing and to assist
them in deciding what they should do next (Neely, 1998; Glykas &
Valiris, 1999).
Performance measurement systems have grown in use and
popularity over the last twenty years. Organisations adopted performance measurement systems for a variety of reasons, but mainly
to achieve control over the organisation in ways that traditional
accounting systems do not permit (Kellen, 2003). A review of the literature shows that traditional performance measurement systems
(based on financial measures) have failed to identify and integrate
all those factors that are critical in contributing to business excellence (Eccles, 1991; Fisher, 1992; Hayes, Wheelwright, & Clark,
1988; Kaplan, 1983, 1984; Maskell, 1992).
During the last decade, a number of frameworks, that help in
designing and implementing performance measurement systems,
has been identified in the literature, such as the balanced scorecard (Kaplan & Norton, 1992), the performance prism (Kennerley &
Neely, 2000), the performance measurement matrix (Keegan, Eiler,
∗ Corresponding author.
E-mail addresses: p.chytas@chios.aegean.gr (P. Chytas), mglikas@aegean.gr
(M. Glykas), gval@aegean.gr (G. Valiris).
0268-4012/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijinfomgt.2010.12.007
& Jones 1989), the results and determinants framework (Fitzgerald
et al., 1991), and the SMART pyramid (Lynch & Cross, 1991). These
frameworks aim to assist organisations in defining a set of measures that reflects their objectives and assesses their performance
appropriately. The frameworks are multidimensional, explicitly
balancing financial and non-financial measures (Kennerley & Neely,
2002). Furthermore, a number of researchers have proposed a wide
range of criteria for designing performance measurement systems
(Globerson, 1985; Maskell, 1992; Morris, 2002).
Despite, the existence of numerous approaches (frameworks,
criteria, etc.) it is evident, from the literature, that the need for
a broader research in the field of performance measurement is
required. The criticism about the static nature of performance measurement systems as well as the relationships and trade-offs that
exist among different measures is the catalyst for this research.
Furthermore, the software applications that have been developed
so far lack of an analytic capability and they cannot do predictive
modelling (Morris, 2002). Despite the many attempts in this area
(EIS, decision support tools), it is claiming that these tools do not
necessarily advance the decision-making process.
The main objective of this research is to propose a methodology (not a new performance measurement framework) that will
support existing measurement framework(s) during the process
of performance measurement systems’ design, implementation
and use, and to advance the decision-making process. Conforming to the most favoured approach, we have adopted the balanced
scorecard, to explore the existence of trade-offs among measures
within the dynamic nature of performance measurement systems
that provide predictive modelling capabilities. The use of FCMs in
the development of a Balanced Scorecard, will allow prospective
decision-makers to incorporate their insights into the model. They
may select the most preferable measures, add new ones, test the
relationships between them, and visualise holistic outcomes.
P. Chytas et al. / International Journal of Information Management 31 (2011) 460–468
This paper consists of five sections. Section 2 provides a literature review and research background; Section 3 presents the
proposed methodology. Section 4 discusses the applicability of the
proposed methodology. Finally, Section 5 concludes this paper.
2. Literature review
Senge (1992) argues that, in today’s complex business world,
organisations must be able to learn how to cope with continuous
change in order to be successful. In this changing environment,
the need for adequate design, implementation and use of performance measurement systems, is greater than ever. Eccles (1991)
claims that it will become increasingly necessary for all major businesses to evaluate and modify their performance measures in order
to adapt to the rapidly changing and highly competitive business
environment.
The introduction of a performance measurement system is
based on a three-stage process: design, implementation and
use. Failing to implement any of these stages will result into a
non-robust performance measurement system. When attempting
to improve organisational performance by utilising performance
measurement systems a critical point is the selection of appropriate measures. Anticipating this, several approaches have been
introduced (frameworks, criteria, etc.). However, in spite the availability of such approaches, the need to further research the area of
performance measurement is necessary.
Several authors have recognised that much more has to be
done in order to identify the relationships among measures (Bititci
and Turner, 2000; Flapper, Fortuin, & Stoop, 1996; Neely, 1999).
Kaplan, when interviewed by de Waal (de Waal, 2003), argued
that cause-and-effect relationships should be tested further. Nevertheless, in almost all cases, organisations ignore the dynamic
interdependencies and trade-offs among measures. Furthermore,
criticism exists regarding performance measurement systems and
their static nature. According to Kennerley and Neely (2002), consideration is being given to what should be measured today, but
little attention is being paid to the question of what should be measured tomorrow. They suggest that measurement systems should
be dynamic and must be modified as circumstances change. A radical rethink of performance measurement is now necessary more
than ever (Corrigan, 1998; Takikonda & Takikonda, 1998). In an
attempt to describe and test cause-and-effect relationships, Kaplan
and Norton (2001) proposed the use of strategy maps. However,
the causal relationships that strategy maps claim to model are not
always linear and one-way (Kaplan and Norton refer only to linear
and one-way cause and effect chain), but mostly a fuzzy mess of
interactions and interdependencies.
Kellen (2003) argues that in the area of executive management
only 6 in 10 executives place confidence in the data presented to
them. He points out that one of the main factors that prevent measurement is the fuzzy objectives. By the same token, Xirogiannis,
Chytas, Glykas, and Valiris (2008) explains that in a performance
measurement system a large number of multidimensional factors can affect performance. Integrating those multidimensional
effects into a single unit can only be done through subjective, individual or group judgement. It is impossible to have an objective
measurement and scale system for each different dimension of
measurement that can facilitate objective value trade-off between
different measures. They argue that techniques, which are suited
to fuzzy paradigms, should be considered.
Identifying the relationships and trade-offs that exist among
measures will be a great step towards the design of a robust performance measurement system. However, the robustness of the
performance measurement system is also based on its successful
implementation and use. According to Neely et al. (2000), imple-
461
mentation is not a straightforward task due to fear/resistance,
politics and subversion. Dumond (1994) claims that the main problems in the implementation of performance measurement systems
are raised due to the lack of communication and dissemination
of performance information. According to De Geus (1994) even a
simplified but credible (causal) model can be a powerful communication and learning tool. In the same token, Morecroft (1994) argues
that models are more effective when they become integral parts of
management debate, communication, dialogue and experimentation. It is possible for managers to gain insights about how their
actions might affect outcomes if they work with models. Furthermore, experimentation with models creates a cycle of increased
learning and improved models.
Finally, further to all the aforementioned issues, Morris (2002)
argues that software applications that have been developed so far,
lack of an analytic capability and they cannot carry out predictive
modelling. Despite the many attempts in this area (EIS, Decision
Support tools), it is claimed that these tools do not necessarily
advance the decision-making process.
2.1. Balanced scorecard
According to Kaplan and Norton (1996a), the balanced scorecard supplements traditional financial measures with criteria that
measure performance from three additional perspectives—those of
customers, internal business processes, and learning and growth
(Fig. 1).
• Customer perspective
Since companies create value through customers, understanding how they view performance becomes a major aspect of
performance measurement.
• Internal business process perspective
According to Kaplan and Norton (2000), in the internalbusiness-process perspective, executives identify the critical
internal processes in which the organisation must excel.
• Learning and growth perspective
According to Kaplan and Norton (2000), this perspective of the
balanced scorecard identifies the infrastructure that the organisation must build to create long-term growth and improvement.
Learning and growth come from three principal sources: 1. People; 2. Systems; and 3. Organisational procedures.
• Financial perspective
Within the balanced scorecard, financial measures remain an
important dimension. Financial performance measures indicate
whether a company’s strategy, implementation, and execution
are contributing to bottom-line improvement.
• Limitations of the balanced scorecard
Balanced scorecard (Kaplan & Norton, 1992), briefly described
previously, is the most popular framework in the area of performance measurement. The introduction of the balanced scorecard
was mainly based on a transition from the traditional financial
performance measurement systems towards a more balanced
approach (financial and non-financial measures) that includes
several measures in a multi-dimensional structure. In spite of
its “reputation”, there are several issues related to the balanced
scorecard, which need further research. More particularly:
• Cause and effect consider to be one-way in nature
The cause and effect concept is a very important element
to consider in an attempt to construct a Balanced Scorecard.
However, the way cause and effect is illustrated is rather problematic. Measures in the balanced scorecard are placed in a cause
and effect chain rather a systemic approach. Kaplan and Norton
(1996b) argue that ‘the financial objectives serve as the focus for
the objectives and measures in all the other scorecard perspectives’.
462
P. Chytas et al. / International Journal of Information Management 31 (2011) 460–468
Fig. 1. The balanced scorecard.
This statement ignores any feedback loops that might exist.
• Trade-offs among measures and among the four perspectives are
ignored
Ignoring the trade-offs among measures as well as among the
four perspectives is rather not an efficient approach. By doing so,
the communication of strategy and dissemination of performance
information is restricted because users are not in position to identify and learn why and how certain things have occurred.
• Measures are equally weighted
All the measures in balanced scorecard are given the same
weighting. This is not what happens in reality. Some measures
may be more important and have greater impact compared to
others. Weighting the measures among each other is critical on
decision-making.
• Design techniques used for the development of a balanced scorecard are rather poor in illustrating the dynamics of a system
(absence of feedback loops)
Two of the most usual design techniques used for the development of the balanced scorecard are the bubble diagram and
the generic value chain model (Fig. 2(a) and (b)). Recently Kaplan
and Norton (2001) introduced a new model; the strategy maps
(Fig. 2(c)). However, as it has been observed, these models lack
the ability of representing feedback loops. This is not very suitable
for communicating strategy as well as exploring the interrelationships among measures and in turn objectives. Ignoring the
feedback loops (two-way cause and effect) at the design stage of
a performance measurement system will lead to a non-effective
representation of the organisation and the dynamics that are
involved. Introducing new measures in this way restricts the possibility to identify the consequences that might be raised in the
whole system.
2.2. A fuzzy logic view—FCMs
Fuzzy logic was introduced in 1965 by Zadeh as a means of representing data and manipulating data that was not precise, but rather
fuzzy. The theory of fuzzy logic provides a mathematical strength
to capture the uncertainties associated with human cognitive processes, such as thinking and reasoning. Since its first appearance,
fuzzy logic has been used in a variety of applications, such as image
detection of edges, signal estimation, classification and clustering.
A fuzzy logic technique represents an alternative solution to the
design of intelligent engineering systems. Thus, fuzzy rule-based
experts systems are widely applied nowadays, this being supported
by the fact that fuzzy logic is linguistic rather than numerical, something which makes it similar to human thinking and hence simpler
to understand and put into practice. It is not within the scope of
this paper to present an overview of fuzzy logic and the reader is
directed to the seminal work on the subject by Zadeh (1997) and
in the more recent non-mathematical text by Kosko (1998). In this
paper, the concept of an FCM is used to define the state of a set of
variables/objectives.
FCMs are soft computing tools which combine elements of fuzzy
logic and neural networks. They are fuzzy signed directed graphs
with feedback loops, in which the set of concepts (each concept
represents a characteristic, state or variable of the system/model;
concepts stand for events, actions, goals, values and/or trends of the
system being modelled as an FCM), and the set of causal relationships is modelled by directed arcs (Fig. 3). FCM theory developed
recently (Kosko, 1986) as an expansion of cognitive maps that had
been employed to represent social scientific knowledge (Axelrod,
1976), to make decision analysis (Zhang, Chen, & Bezdek, 1989)
and to analyse extend-graph theoretic behaviour (Zhang & Chen,
1988).
Fig. 3 illustrates an FCM which is used to simulate the behaviour
of Company Profitability in terms of other factors that positively or
negatively affect its state (behaviour). In the figure above, Company
Profitability is directly affected by the following factors: Customer
Satisfaction (positive effect), Sales Volume (positive effect) and Internal Cost (negative effect). Directed, signed and weighted arcs, which
represent the causal relationships that exist between the concepts,
interconnect the FCM concepts. For example, in Fig. 3, there is a
strong positive relationship from the Customer Satisfaction concept
to the Company Profitability concept. Each concept is characterised
by a numeric value that represents a quantitative measure of the
concept’s presence in the model. A high numeric value indicates the
strong presence of a concept. The numeric value results from the
transformation of the real value of the system’s variable, for which
this concept stands, to the interval [0,1]. All the values in the graph
are fuzzy, so weights of the arcs are described with linguistic values (such as: “strong”, “weak”, etc.) that can be “defuzzified” and
transformed to the interval [−1,1].
P. Chytas et al. / International Journal of Information Management 31 (2011) 460–468
463
Fig. 2. Design techniques for the development of a balanced scorecard.
Studying this graphical representation, one can conclude which
concept influences other concepts and their interconnections. This
representation makes updating the graph structure easy, as new
information becomes available or as more experts are asked. This
can be done, for example, by the addition or deletion of an interconnection or a concept.
Between concepts, there are three possible types of causal
relationships expressing the type of influence of one concept on
another. The weight of an interconnection, Wij , for the arc from
concept Ci to concept Cj , can be positive (Wij > 0), which means that
an increase in the value of concept Ci leads to the increase of the
value of concept Cj , and a decrease in the value of concept Ci leads to
464
P. Chytas et al. / International Journal of Information Management 31 (2011) 460–468
3. A proactive balanced scorecard methodology (PBSCM)
3.1. Successful execution of strategy: a new component
According to Kaplan and Norton (2004) successful execution of
a strategy (Breakthrough Results) requires two components:
{Breakthrough Results}
= {Describe the Strategy} + {Manage the Strategy}
(3)
The philosophy of the two components is simple:
• You cannot manage (second component) what you cannot measure (first component).
• You cannot measure what you cannot describe (Breakthrough
Results).
Fig. 3. A simple FCM.
the decrease of the value of concept Cj . Or there is a negative causality (Wij < 0), which means that an increase of the value of concept
Ci leads to the decrease of the value of concept Cj and vice versa.
When there is no relationship from concept Ci to concept Cj , then
(Wij = 0). In Fig. 3, the weight of the interconnection between the
concepts, Company Profitability and Sales Volume is positive (represented by a blue arc and a positive value) and is illustrated as
follows: if Sales Volume is high then Company profitability will be
high—if Sales Volume is low then Company Profitability will be low.
An expert defines the main concepts that represent the model
of the system, based on his knowledge and experience on the
operation of the system. At first, the expert determines the concepts that best describe the system. He knows which factors are
crucial for the modelling of the system and he represents each
one by a concept. Moreover, he has observed which elements
of the system influence other elements and for the corresponding concepts he determines the positive, negative or zero effect
of one concept on the others. He describes each interconnection
with a linguistic value that represents the fuzzy degree of causality between concepts. The linguistic weights are transformed into
numerical weights using the methodology proposed by (Glykas,
2010).
When the FCM starts to model the system, concepts take their
initial values and then the system is simulated. At each step, the
value of each concept is determined by the influence of the interconnected concepts on the corresponding weights:
ait+1 = f
⎛
⎝
n
/ i
j=1,j =
⎞
wji atj ⎠
(1)
where at+1
is the value of concept Ci at step t + 1, atj the value of
i
the interconnected concept Cj at step t, Wji the weighted arc from
concept Cj to Ci , and f a threshold function. Three threshold functions have been identified in the literature (Kosko, 1998) and are
described below:
si (xi ) = 0, xi ≤ 0
si (xi ) = −1, xi > 0
bivalent
si (xi ) = −1, xi ≤ −0.5
si (xi ) = 0, − 0.5xi < xi < 0.5
si (xi ) = 1, xi ≥ 0.5
trivalent
1
si (xi ) =
1 + e−cxi
logistic signla, c = 5
(2)
According to Kaplan and Norton (2004), their first book, The
Balanced Scorecard, has addressed the first component by showing
how to measure strategic objectives in multiple perspectives. It also
presented the early ideas regarding the second component, how
to manage the strategy. Their second book, The Strategy-Focused
Organisation, has provided a more comprehensive approach for
how to manage the strategy. It has also introduced strategy maps
for the first component, how to describe the strategy. Their third
book, Strategy Maps, goes into much more detail on this aspect,
using linked objectives in strategy maps to describe and visualize
the strategy. They rewrite the above “equation” as follows:
{Breakthrough Results} = {Strategy Maps} → [Describe]
+ {Balanced Scorecard} → [Measure] + {Strategy
− Focused Organisation} → [Manage]
(4)
However, it is our belief that both Eqs. (2) and (3) omit an important
component: Simulate the Strategy. Hence, we rewrote the Eq. (2)
as follows:
{Breakthrough Results} = {Simulate the Strategy}
+ {Describe the Strategy} + {Manage the Strategy}
(5)
By incorporating this new component (Simulate the Strategy) in
the above “equation” we aim to overcome all the limitations identified in the literature review (in particularly in Section 2.2.1) and
view performance measurement and in particularly the balanced
scorecard within a systemic approach. In order to address this new
component, we suggest the use of fuzzy cognitive maps (FCMs). As
it was described previously, FCMs are fuzzy signed directed graphs
with feedback loops, in which the set of objects is modelled by the
nodes, and the set of causal relationships is modelled by directed
arcs. The FCM theory, was developed recently (Kosko, 1986) as an
expansion of the cognitive maps that had been employed to represent social scientific knowledge (Axelrod, 1976), to make decision
analysis (Zhang et al., 1989) and to analyse extend-graph theoretic
behaviour (Zhang & Chen, 1988). FCMs combine the strengths of
cognitive maps with fuzzy logic. By representing human knowledge in a form more representative of natural human language
than traditional concept mapping techniques, FCMs ease knowledge engineering and increase knowledge-source concurrence. The
characteristics and the structure of FCMs allow us to re-write Eqs.
(3) and (4) as follows:
{Breakthrough Results} = {FCMs} → [Simulate]
+ {FCMs} → [Describe] + {Balanced Scorecard} → [Measure]
+ {Strategy-Focused Organisation} → [Manage]
(6)
P. Chytas et al. / International Journal of Information Management 31 (2011) 460–468
465
Mission
Strategy
Objectives
Align People with Strategy
Focus and Develop Performance
Describe
KPI
FCMs
Simulate
Targets
What-if scenarios
Forecast
Initiate Corrective Actions
Adjust Targets
Fig. 4. The PBSCM.
In the above “equation”, in the first instance (simulate),
we use the simulation characteristics of the FCMs theory. The
FCM approach involves forward-chaining (what-if analysis). The
forward-chaining provides business domain experts with the capability to reason about the map they have constructed (nodes,
relationships and weights) and examine different scenarios. In the
second instance (describe), we utilise the representation capabilities of the FCMs theory. FCMs are illustrated as causal-loop
diagrams. This is very suitable for communicating strategy as well
as exploring the interrelationships among measures and in turn
objectives.
The PBSCM goes through a series of stages that involve:
(1) inputs to be provided, and (2) outcomes to be generated.
Business domain experts and/or professionals of performance measurement/balanced scorecard are people with specific business
expertise that contribute towards providing the business knowledge for the PBSCM. The following table (Table 1) indicates the
stages of a PBSCM together with the inputs and outcomes of each
stage.
Before proceeding to each of the aforementioned stages a kickoff meeting takes place between the domain experts. The aim of
this meeting is for all participants to contribute towards:
3.2. Overview of the PBSCM
•
•
•
•
The methodology for the development of a proactive balanced
scorecard is depicted in the figure below (Fig. 4). PBSCM is capable
of illustrating non-linear interactions and feedback loops through
the use of FCMs as a causal-loop diagram and performing what-if
scenarios through the use of FCMs simulation.
Establishing the PBSCM team.
Clarifying the objectives of the team.
Identifying the context of the PBSCM.
Selecting the reference material to be used for the construction
of the PBSCM.
• Anticipating possible user benefits.
• Preparing further actions for the participants.
Table 1
Inputs and outcomes of the PBSCM.
Stage
Input
Outcome
1. Establishing the mission, vision, strategic
objectives, perspectives and critical success
factors (CSF)
1. Interviews with middle and top management
1. Mission
2. Internal company data
1. CSF
1. KPI
1. KPI
2. Vision
3. Strategic objectives
4. Perspectives
5. CSF
1. KPI in each perspective
1. Target for each KPI
1. FCM with no weights
1. FCM with no weights
1. Final FCM with weights and concept values
1. Final FCM
1. Adjust targets
2. Identify key performance Indicators (KPI)
3. Establish targets
4. Define relationships among the identified
KPI
5. Assign linguistic variables to weights and
concepts-(KPI)
6. Continuous improvement
466
P. Chytas et al. / International Journal of Information Management 31 (2011) 460–468
Financial
Internal Process
Customer
Fig. 5. Define relationships among the identified KPIs.
The PBSCM methodology is composed of the following stages:
1. Establishing the mission, vision, strategic objectives, perspectives and CSF
In this stage the focus is on understanding the organisation’s
strategy, culture and capabilities in order to specify the strategic
objectives (which state the specific goals/directions the organisation aims to achieve), perspectives and critical success factors
(things the organisation must do well to achieve its strategic
objectives).
2. Identify key performance indicators (KPI)
This stage aims to narrow down the list of all possible measures
into a shortest one that provides the KPIs, which will be used in
each perspective.
3. Establish targets
Measurement alone is not good enough. We must drive
behavioural changes within the organisation if we expect to execute strategy. This requires establishing a target for each KPI
within the balanced scorecard. Targets are designed to drive and
push the organisation as to meet its strategic objectives. Targets
need to be realistic so that people feel comfortable about trying
to execute on the target.
4. Define relationships among the identified KPI
As the KPIs constituting the different perspectives have been
derived, the relationships between these KPIs (KPIs are represented as concepts in the FCM) have to be defined. An edge
connecting two KPIs represents a relationship. The direction of
the relationship (i.e. which KPI affects the other) is denoted by
the direction of the arrow on this edge. The FCM that has been
constructed (Fig. 5) using the method mentioned above does
not contain any information except that there are relationships
between abstract concepts (KPI). The next step is to enrich the
map with numerical values, which are assigned to the concepts
and relationships.
5. Assign linguistic variables to weights and concepts (KPI)
Knowledge on the behaviour of a system is rather subjective
and in order to construct a model of the system it is proposed to utilise the experience of experts. Experts are asked to
describe the causality among concepts using linguistic notions
(A fuzzy logic perspective). They will determine the influence
of one concept to the other as “negative” or “positive” and then
they will describe the grade of the influence with a linguistic
variable such as “strong” and “weak”. Influence of one concept
over another, is interpreted as a linguistic variable in the interval [-1,1]. Its term set T(influence) is: T(influence) = {negatively
very-very high, negatively very high, negatively high, negatively
medium, negatively low, negatively very low, negatively veryvery low, zero, positively very-very low, positively very low,
positively low, positively medium, positively high, positively
very high, positively very-very high}.
We propose a semantic rule M to be defined at this point.
The above-mentioned terms are characterized by the fuzzy sets
whose membership functions are shown in Fig. 6.
• M(negatively very-very high) = the fuzzy set for “an influence
close to −90%” with membership function nvvh .
• M(negatively very high) = the fuzzy set for “an influence close
to 80%” with membership function nvh .
• M(negatively high) = the fuzzy set for “an influence close to
65%” with membership function nh .
• M(negatively medium) = the fuzzy set for “an influence close
to −50%” with membership function nm .
• M(negatively low) = the fuzzy set for “an influence close to
−35%” with membership function nl .
• M(negatively very low) = the fuzzy set for “an influence close
to −20%” with membership function nvl .
• M(negatively very-very low) = the fuzzy set for “an influence
close to −10%” with membership function nvvl .
• M(zero) = the fuzzy set for “an influence close to 0” with membership function z .
• M(positively very-very low) = the fuzzy set for “an influence
close to 10%” with membership function pvvl .
• M(positively very low) = the fuzzy set for “an influence close to
20%” with membership function pvl .
• M(positively low) = the fuzzy set for “an influence close to 35%”
with membership function pl .
• M(positively medium) = the fuzzy set for “an influence close to
50%” with membership function pm .
• M(positively high) = the fuzzy set for “an influence close to
65%” with membership function ph .
• M(positively very high) = the fuzzy set for “an influence close
to 80%” with membership function pvh .
• M(positively very-very high) = the fuzzy set for “an influence
close to 90%” with membership function pvvh .
The membership functions are not of the same size since it is
desirable to have finer distinction between grades in the lower
and higher end of the influence scale. As an example, three
experts propose different linguistic weights for the interconnection Wij from concept Ci to concept Cj : (a) positively high, (b)
positively very high, and (c) positively very-very high. The three
suggested linguistics are integrated using a sum combination
method and then the “defuzzification” method of the centre of
gravity (CoG) is used to produce a weight Wij = 0,73 in the interval [−1,1]. This approach has the advantage that experts do not
have to assign numerical causality weights but to describe the
degree of causality among concepts.
A similar methodology can be used to assign values to concepts. The experts are also asked to describe the measurement of
each concept using linguistic notions once again. Measurement
of a concept is also interpreted as a linguistic variable with values
in the interval [−1,1]. Its term set T(Measurement) = T(Influence).
A new semantic rule M2 (analogous to M) is also defined and
these terms are characterized by the fuzzy sets whose membership functions 2 are analogous to membership functions
.
6. Continuous improvement
The purpose of this phase is to continuously update the usability of the FCMs in order to provide improved user support.
The continuous improvement cycle requires the users to run a
P. Chytas et al. / International Journal of Information Management 31 (2011) 460–468
467
µ
µ nvvh µ nvh µ nh µ nm
µ nl
µ nvl µ nvvl µ z
µ pvvl µ pvl
µ pl
µ pm
µ ps
µ pvs
0,1
0,35
0,5
0,65
0,8
µ pvvs
1
0,5
-1
-0,9
-0,8
-0,65
-0,5
-0,35
-0,2
-0,1
0
0,2
0,9
1
influence
Fig. 6. Membership functions of linguistic variable influence.
simulation exercise on the FCM (using weight and concept values defined in the previous stage) and test its effectiveness in
response to the targets defined previously. The adjustment will
be based on the behaviour of the FCM during simulation and on
the results it delivers.
4. Discussion
As far as the theoretical value is concerned, the PBSCM methodology extends previous research attempts by (a) allowing fuzzy
definitions in the cognitive maps, (b) introducing a specific interpretation mechanism of linguistic variables to fuzzy sets, and (c)
allowing dynamic decomposition and reconfiguration of a balanced
scorecard strategy-map. As far as the practical value is concerned,
preliminary evaluation results indicate that when compared to the
expert estimates, the methodology provides reasonably good activities.
4.1. Added value
Having established the theoretical and practical value of the proposed methodology, it is useful to discuss also the added value of
incorporating such a methodology during the development of the
balanced scorecard. It is presumed in this paper that the resulting
methodology provides real value to the principle beneficiaries and
stakeholders of balanced scorecard projects. For example:
• The methodology eases significantly the complexity of deriving
expert decisions concerning the balanced scorecard development.
• The proposed methodology serves as a back-end to provide holistic strategic performance evaluation and management.
◦ Shared performance measurement enables business units to
realize how they fit into the overall business model of the enterprise and what their actual contribution is.
◦ Senior management receives valuable input from the business
units (or the individual employees) who really comprehend
the weaknesses of the current strategic model as well as the
opportunities for performance change.
• Shared culture
◦ All business units at the enterprise feel that their individual
contribution is taken under consideration and provide valuable
input.
◦ All business units and individuals at the enterprise feel confident and optimistic; they realize that they will be the ultimate
beneficiaries of the balanced scorecard exercise.
◦ The information sharing culture supports the enterprise’s competitive strategy and provides the energy to sustain this by
exploiting group and individual potential to its fullest.
• Shared learning
◦ The enterprise realizes a high return on its commitment to
human resources.
◦ There is a constant stream of improvement within the enterprise.
◦ The entire enterprise becomes increasingly receptive to strategic changes, since the benefits can be easily demonstrated to
individual business units.
• Shared information
◦ All business units and individuals have the necessary information needed to set clear objectives and priorities.
◦ Senior management can effectively control all aspects of the
strategic process
◦ The enterprise reacts rapidly to threats and opportunities.
5. Conclusions
4.2. Preliminary usability evaluation
Senior managers of two major IT enterprises have evaluated
the usability of the proposed methodology and have identified a
number of benefits that can be achieved by the utilization of the
proposed mechanism as a methodology for balanced scorecard
development. A summary of major business benefits (as identified
by senior managers) is provided to improve the autonomy of this
paper:
• Shared goals
◦ Concept-driven simulation pulls individuals together by providing a shared direction and determination for strategic
change.
This paper proposes a proactive balanced scorecard
methodology (PBCSM). The proposed decision aid may serve
as a back end to Balanced Scorecard development and implementation. By using FCMs, the proposed methodology draws a causal
representation of KPIs; it simulates the KPIs of each perspective
with imprecise relationships and quantifies the impact of each KPI
to other KPIs in order to adjust performance targets. The underlying
research addressed the problems of the current balanced scorecard
development process. The main objective of this research is to
propose a methodology (not a new performance measurement
framework) that will support existing measurement framework(s)
during the process of performance measurement systems’ design,
implementation and use, and to advance the decision-making
process. Future research will focus on conducting in depth studies
468
P. Chytas et al. / International Journal of Information Management 31 (2011) 460–468
to test and promote the usability of the methodology and to
identify potential pitfalls.
References
Ansoff, H. I. (1975). Managing strategic surprise by response to weak signals. California Management Review, XXVIII(2), 21–33.
Axelrod, R. (1976). Structure of decision: The cognitive maps of political elites. Princeton, NJ: Princeton University Press.
Bititci, U. S., & Turner, T. (2000). Dynamics of performance measurement systems.
International Journal of Operations & Production Management, 20(6), 692–704.
Corrigan, J. (1998). Performance measurement: Knowing the dynamics. Australian
Accounting, 68(9), 30–31.
De Geus, A. (1994). Modeling to predict or to learn? Portland: Productivity Press.
de Waal, A. A. (2003). The future of the balanced scorecard: An interview with
Professor Dr Robert S Kaplan. Measuring Business Excellence, 7(1), 3.
Drucker, P. F. (1993). Managing in turbulent times. Harper Collins Publishers.
Dumond, E. J. (1994). Making best use of performance measures and information.
International Journal of Operations & Production Management, 14(9), 16–31.
Eccles, R. G. (1991). Performance measurement manifesto. Harvard Business Review,
69(January–February), 131–137.
Fisher, J. (1992). Use of non-financial performance measures. Journal of Cost Management, 6, 31–38.
Fitzgerald, L., Johnston, R., Brignall, S., Silvestro, R., & Voss, C. (1991). Performance
measurement in service businesses. London: The Chartered Institute of Management Accountants.
Flapper, S. D., Fortuin, L., & Stoop, P. P. (1996). Towards consistent performance management systems. International Journal of Operations & Production Management,
16(7), 27–37.
Globerson, S. (1985). Issues in developing a performance criteria system for an
organisation. International Journal of Production Research, 23(4), 639–646.
Glykas, M. (2004). Workflow and process management in printing and publishing
firms. International Journal of Information Management, 24(6), 523–538.
M. Glykas (Ed). Fuzzy Cognitive Maps: Advances in Tehories, Methodologies, Applications and Tools. ISBN: 3642032192, Springer, Germany, 2010.
Glykas, M., & Valiris, G. (1999). Formal methods in object oriented business modelling. Journal of Systems and Software, 48(1), 27–41.
Grove, A. (1999). Only the paranoid survive: How to exploit the crisis points that challenge every company. Bantam Books.
Hayes, R. H., Wheelwright, S. C., & Clark, K. B. (1988). Dynamic manufacturing: Creating the learning organization. New York: Free Press.
Kaplan, R. S. (1983). Measuring manufacturing performance: A new challenge for
managerial accounting research. The Accounting Review, 58(4), 686–705.
Kaplan, R. S. (1984). Yesterday’s accounting undermines production. Harvard Business Review, 62(July–August), 95–101.
Kaplan, R. S., & Norton, D. P. (1992). The balanced scorecard—measures that drive
performance. Harvard Business Review, January–February, 71–92.
Kaplan, R. S., & Norton, D. P. (1996a). Using the Balanced Scorecard as a Strategic
Management System. Harvard Business Review, January–February.
Kaplan, R. S., & Norton, D. P. (1996b). The balanced scorecard—translating strategy into
action. Boston, MA: Harvard Business School Press.
Kaplan, R. S. & Norton, D. P. (2000). Why does business need a balanced scorecard?
http://www.corpfinance.riag.com/.
Kaplan, R. S., & Norton, D. P. (2001). Translating the balanced scorecard from performance measurement to strategic management: Part 1. Accounting Horizons,
15(1), 87–104.
Kaplan, R. S., & Norton, D. P. (2004). Strategy maps: Converting intangible assets into
19 tangible outcomes. Boston, MA: Harvard Business School Press.
Keegan, D. P., Eiler, R. G., & Jones, C. R. (1989). Are your performance measures
obsolete? Management Accounting (US), 70(12), 45–50.
Kellen, V. (2003). Business performance measurement, at the crossroads
of strategy, decision-making, learning and information visualization.
http://www.kellen.net/bpm.htm.
Kelly, K. (1998). New rules for the new economy: Ten ways the network economy is
changing everything. London: Fourth Estate.
Kennerley, M. P., & Neely, A. D. (2000). Performance measurement frameworks—a
review. In Proceedings of the 2nd international conference on performance measurement Cambridge,
Kennerley, M. P., & Neely, A. D. (2002). A framework of the factors affecting the
evolution of performance measurement frameworks. International Journal of
Operations and Production Management, 22(11), 1222–1245.
Kosko, B. (1986). Fuzzy cognitive maps. International Journal of Man-Machine Studies,
24, 65–75.
Kosko, B. (1998). Hidden patterns in combined and adaptive knowledge networks.
International Journal of Approximate Reasoning, 2, 377–393.
Lynch, R. L., & Cross, K. F. (1991). Measure up—the essential guide to measuring business
performance. London: Mandarin.
Maskell, B. H. (1992). Performance measurement for world class manufacturing: A
model for American companies. Cambridge: Productivity Press.
Morecroft, J. (1994). Executive knowledge, models and learning. Portland: Productivity
Press.
Morris, H. (2002). Balanced scorecard report: Insight, experience and ideas for strategy focused organisations. Harvard Business School Publishing, 4(1), 1–17.
Neely, A. (1998). Measuring business performance—why what and how. London:
Economist Books.
Neely, A. (1999). The performance measurement revolution: Why now and what
next? International Journal of Operations & Production Management, 19(2), 205.
Neely, A. D., Mills, J. F., Bourne, M. C. S., Kennerley, M., Platts, K. W., Richards, H.,
& Gregory, M. J. (2000). Performance measurement system design: Developing and testing a process-based approach. International Journal of Operations &
Production Management, 20(10), 1119–1145.
Senge, P. N. (1992). The fifth discipline: The art and practice of the learning organization.
London: Century Business Press.
Takikonda, L., & Takikonda, R. (1998). We need dynamic performance measures.
Management Accounting, 80(3), 49–51.
Xirogiannis, G., Chytas, P., Glykas, M., & Valiris, G. (2008). Intelligent impact assessment of HRM to the shareholder value. Expert Systems with Applications, 35(4),
2017–2031.
Zhang, W. R., & Chen, S. S. (1988). A logical architecture for cognitive maps. In Proceedings of the second IEEE international conference on neural networks San Diego,
CA.
Zhang, W., Chen, S., & Bezdek, J. C. (1989). Pool 2: A generic system for cognitive
map development and decision analysis. IEEE Transactions on Systems, Man and
Cybernetics, 19, 31–39.
Zadeh, L. A. (1997). The roles of fuzzy logic and soft computing in the conception
design and deployment of intelligent systems Software agents and soft computing: Towards enhancing machine intelligence concepts and applications. Lecture
Notes in Computer Science, 1198, 83–190.
Michael Glykas is Assistant Professor at the department of Financial and Managemen Engineering of the University of the Aegean. His research interests include:
business process management, human resource management, performance measurement and fuzzy cognitive mapping.
Panagiotis Chytas is a researcher at the deparment of Business Administration at
the University of the Aegean in Greece. His research interests include: performance
measurement, information technology, fuzzy cognitive mapping.
George Valiris is an Associate Professor at the department of Business Administration at the University of the Aegean in Greece. His research interests include:
information technology, human resource management, business process management.