doi: 10.1111/isj.

12071

Info Systems J (2016) 26, 339–367 339

Towards an understanding of the role of

business intelligence systems in
organisational knowing
Arisa Shollo* & Robert D. Galliers†,‡

*Department of Operations Management, Copenhagen Business School, Frederiksberg,

Denmark, email: [email protected], †Information & Process Management and Sociology
Departments, Bentley University, Waltham, MA USA, and ‡School of Business and
Economics, Loughborough University, Loughborough, UK, email: [email protected]

Abstract. Recent advances in information technology (IT), such as the advent of

business intelligence (BI) systems, have increased the ability of organisations to
collect and analyse data to support decisions. There is little focus to date, however,
on how BI systems might play a role in organisational knowledge creation – in
organisational knowing. We develop a conceptual framework of organisational
knowing based on a synthesis of the literature, and use this as a framework to in-
vestigate how BI systems facilitate knowing in a case organisation. We identify two
practices triggered by BI systems that distinguish them from prior applications of
IT: the ability to initiate problem articulation and dialogue, and that of data selection
(e.g. to address information needs of organisational decision makers at different
managerial levels). This study provides empirical evidence of the performative out-
come of BI systems in relation to organisational knowing through the practices of
articulation and data selection. It provides a practice perspective on BI and focuses
on the role of BI systems in organisational knowing thereby opening up a new de-
parture for BI research that considers the implications of BI systems in organisa-
tions with actual practice in mind.

Keywords: business intelligence systems, organisational knowing, data selection,

articulation, practice studies, case study research

T H E A DV E N T A N D I M PAC T O F B I S Y S T E M S

Recent advances in information technology (IT) in providing business intelligence (BI), com-
bined with the recent financial and economic crisis, magnify the criticality of organisational
decision-making in an increasingly competitive and sometimes adverse business environment.
As a consequence, BI systems have gained prominence over recent years (Elbashir & Williams,
2007). For several years now, BI spending has increased compared with the IT budget overall
(Gartner Press Release, 2014). Luftman & Zadeh (2011) identify BI as the most influential

© 2015 Wiley Publishing Ltd

 340 A Shollo & R D Galliers

technology in organisations, while McAfee & Brynjolfsson (2012) provide evidence that the adop-
tion of BI technologies leads to organisational productivity increases between 4% and 6%. Thus,
BI systems are acknowledged to play – or at least have the potential to play – an increasingly
important role in organisational decision-making. The extant literature recognises the contribution
of BI systems to decision-making (e.g. Popovic et al., 2012) and identifies time saving and better
information to support decision-making as the main direct benefits (e.g. Watson et al., 2002).
Research on BI has primarily focused on the decision-supporting role of BI but without
explaining how new knowledge may be obtained as a result (Arnott & Pervan, 2014; Davenport,
2010). Organisations are in a continuous struggle to make sense of their dynamic business en-
vironments with a view to gaining new insights and creating new knowledge. Processes and
practices that lead to improved knowledge constitute what has been called a ‘knowing organi-
sation’ (Choo, 1998) and are fundamental to knowledge work (Newell et al., 2009; Newell,
2014). Positioning BI systems solely as supporting decision tools thus limits our understanding
of the impact BI might have on other, explorative, organisational processes, and even more so
on how they impact knowledge work.
Two views on the role that BI systems play in organisations can be gleaned from the extant
literature. First, there is what might be termed a ‘traditional’, technologically oriented, view that
sees BI systems’ primary role as the transformation of raw data into information, and of informa-
tion into knowledge, in support of decision-making in organisations. A common premise of this
view is that decision-making (and learning more generally) is a cognitive activity where BI sys-
tems and related technologies (e.g. enterprise and knowledge management systems) identify,
capture and produce new insights and knowledge that ultimately lead to better decisions being
made (Golfarelli et al., 2004; Cheng et al., 2006; Clark et al., 2007). This view, because it fo-
cuses on technology and views knowledge as inert ‘material’, which is fixed and can be con-
trolled, transferred and used in decision-making processes (Gherardi, 2000; Petrini &
Pozzebon, 2009), is limited and does not take into consideration the social, participative and
contested nature of knowledge creation and of decision-making activities in organisations
(Cook & Brown, 1999).
Second, a ‘practice’ view has been proposed to overcome the limitations of this ‘traditional’
perspective. The ‘traditional’ view has been criticised for focusing mainly on technology at
‘the expense of people’ (Swan et al., 1999; Galliers & Newell, 2001; Shollo & Kautz, 2010), in
a similar fashion to criticisms of business process reengineering (Davenport, 1995; Galliers &
Swan, 1999). Cook and Brown (1999) argue that the concept of knowing helps researchers
and practitioners in understanding people’s practices in interacting with the social and physical
world. This practice perspective, thus, focuses on the actual practices of organisational actors,
emphasising the tacit, situated and dynamic nature of performing knowledge work.
The previously presented views are complementary, and thus, we argue that only by adding
knowing – what we do – to knowledge – what we know – can we build a more complete picture
of how BI systems impact organisational activities. However, notwithstanding the considerable
research interest in BI (e.g. Chen et al., 2012), there has to date been little attention given to
improving our understanding of the role of BI in the process of organisational knowing and
the underlying mechanisms of BI systems that facilitate this process. We seek to fill this gap
by examining how BI facilitates knowing in dynamic organisational settings. We do so, as a

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 341

case in point, by considering the use of BI in a financial institution in which we interviewed em-
ployees who use BI systems in their everyday work.
Our contribution to the literatures on BI and knowledge work is twofold. First, we develop a
conceptual framework of organisational knowing, based on the extant literature. We use this
as sensemaking device in analysing the case study data. Second, and what is the main contri-
bution of this study, we highlight the cyclical nature of the ‘data selection’ and ‘articulation’ prac-
tices triggered by BI systems in the cycle of organisational knowing. With this, we provide
empirical evidence of the agency of BI systems in the active process of knowing.
In the following sections, first, we provide a brief history of the evolution towards the BI sys-
tems of the present day. Then, we unpack current perspectives on BI and organisational know-
ing and continue, in the section on Organisational Knowing as an ‘Active Process’, to
synthesise ‘knowing’ as an active process based on a contemporary literature review on the
concept. As a result, we develop our conceptual model and clarify our research questions. In
the Research Design and Presentation of the Case Study section, we detail the case study
design and methodology, and then go on to present the results of the fieldwork. A discussion
of our findings follows, in which we identify implications for theory and practice and propose a
future research agenda.

C O N T E X T – I T A N D M A N AG E M E N T D E C I S I O N - M A K I N G

Business intelligence systems are the latest in a long line of technologies that have been devel-
oped to support organisational activities and decision-making (e.g. Somogyi & Galliers, 1987).
The introduction of the mainframe computer in the 1950s led to the development of data pro-
cessing systems, and, in turn, management information systems (MIS) – systems designed
to support managerial decision-making. Figure 1 illustrates the evolution of systems designed
to support managerial decision-making, for example, from static mass-produced reports to

Figure 1. The evolution of information systems designed to support managerial decision-making.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 342 A Shollo & R D Galliers

interactive personalised reports, dashboards and datasets. From the mid-1960s, mini com-
puters marked the birth of decision support systems (DSS), group decision support systems
(GDSS) and subsequently executive information systems (EIS) (cf. Nunamaker, 1989; Pervan,
1998). MIS output was in the form of large printouts – periodic reports that provided summary
data but did not provide an interactive capability to support managers in their decision-making.
Thus, DSS (and GDSS subsequently) were developed – designed to support semi-structured
and unstructured decisions in a more interactive manner (Gorry & Scott-Morton, 1971; Arnott
& Pervan, 2005; 2014). During the 1970s, DSS incorporated search functionality in documents,
giving birth to document-based systems (Power, 2007). With the development of relational da-
tabases – in the late 1970s – the potential of capturing increasing quantities of data and im-
proved modelling capability led to the arrival of EIS (Arnott & Pervan, 2005). These systems
used pre-defined information screens that facilitated executive enquiry. However, it was only
with the arrival of data warehousing and online analytical processing that EIS began to become
more widely used (Inmon, 1992). At roughly the same time, partly as a result of downsizing be-
cause of business process engineering (Davenport, 1995), the concept of knowledge manage-
ment and knowledge management systems (KMS) (e.g. Leidner, 2000a, 2000b) emerged. This
was due, to some extent, to the adverse impacts on the retention of organisational knowledge,
which resulted from managerial redundancies (Galliers & Newell, 2001). Thus, we see a pro-
gression in the development of information systems (IS) in support of managerial decision-
making from those that relied on standardised, mass produced, period reports that did not allow
for online queries on the part of managers (traditional MIS) to those that were increasingly inter-
active and available on demand to individual managers (EIS and KMS), as illustrated in
Figure 1.
The latest in this line of technologies are BI systems (BIS in Figure 1). BI systems are inte-
grated systems that are linked to a data warehouse and other applications and are designed
to facilitate the analysis of stored (real-time and historical) data in support of ad hoc managerial
decision-making (Davenport, 2006). BI systems differ from their predecessors in that the provi-
sion of information is enabled by the integration of databases from different business domains.
The idea is that the availability and combination of data from different sources may provide new
insights that should lead to better decisions being made. The increased granularity of the data
in BI analysis is also worth mentioning as it makes the BI artefact much more traceable, trans-
parent and easy customizable. Luhn (1958; p. 315) was the first to use the term BI in a scientific
context. He defines BI as, ‘the ability to apprehend the interrelationships of presented facts in
such a way as to guide action towards a desired goal’. BI systems consist of processes, tech-
nologies and applications that are meant to enable organisations to gather, store, analyse and
transform data into information that is relevant for decision-making (Wixom & Watson, 2010;
Davenport et al., 2012).
This evolution of MIS and the historical context of the onset of BI systems have contributed to
much of the BI research that views BI systems as the latest in the line of decision support sys-
tems (Arnott & Pervan, 2014). Decision-making, from this perspective, is the process of evalu-
ating several alternative courses of action and choosing the best of these alternatives to
proceed (March, 1994). In this context, decision-support activities address a specific problem,
which has been identified and acknowledged within an organisation (Skyrius et al., 2013). As

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 343

presaged previously, however, this paper takes a different view. Instead of focusing on BI
systems in terms of decision support, we aim to demonstrate the role that they play in
broader organisational knowing processes and practices. These include, but are not con-
fined to, decision-making. As Skyrius and colleagues argue, ‘business intelligence can
be considered an activity which, apart from encompassing decision support, has a perma-
nent nature and allows the discovery of problems and general awareness about the state
of activities’ (Skyrius et al., 2013; p. 33).
Thus, inspired by the recent strategy-as-practice literature (Henfridsson & Lind, 2014; Huang
et al., 2014; Peppard et al., 2014; Whittington, 2014), our focus in this paper relates to the point
that BI systems affect not only the processes of organisational decision-making but also the
practices of organisational actors – how they make sense of, create and share knowledge. In
the following section, we consider the current literature on the role of BI systems in organisa-
tions, before going to empirically consider the role of BI systems in organisational settings.

U N PAC K I N G P E R S P E C T I V E S O N B I A N D O R G A N I S AT I O N A L K N OW I N G

According to Davenport (2010), the premise and goal of BI is to support decision- making and
ultimately to lead to better decisions that have a real impact on organisational performance
(Chen et al., 2012; Sharma et al., 2014). During the last decade, the concept of BI has been
defined in many different ways (Shollo & Kautz, 2010) with some authors referring to BI as a
process (Golfarelli et al., 2004), a product (Lönnqvist & Pirttimäki, 2006), a mere technology
(Jermol et al., 2003) or all of these combined (Wixom and Watson, 2010). The latter view ap-
pears to have become the most accepted at the moment. For example, recent work by
Holsapple et al. (2014), conceptualises BI and analytics as an overall evidence/data-based par-
adigm for making decisions where specific BI activities are carried out by the system (e.g. gath-
ering and storing) as well as BI users (e.g. analysing), which are implemented via selected
practices (e.g. data standards) and technologies (e.g. query analyser) through a process of
transformation where data are transformed into insights or actions. Along similar lines, Chen
et al. (2012) also highlight the data driven property of BI.
From this perspective, BI systems are seen as supporting a continous process where data
are gathered and stored, then transformed into information by anaylysis. This information is
then transformed into knowledge to support decisions (Golfarelli et al., 2004; Lönnqvist &
Pirttimäki, 2006; Clark et al. 2007). Negash (2004; p. 6) considers BI as a form of knowledge,
and the role of BI systems as being to ‘convert data into useful information and, through human
analysis, into knowledge’. Thus, BI systems are viewed as being a means to create knowledge
useful for decision-making.
This view is similar to the ‘traditional’ technologically oriented perspective that regards the
role of IS (enterprise systems and KMS, for example) in knowledge creation, in that these sys-
tems have been assumed to be solutions that enable the translation of data into information
and, ultimately, knowledge (Newell et al., 2002). Studies have shown, however, that BI systems
have not delivered on their promises of fully supporting decision makers (Brydon & Gemino,
2008; Davenport 2010) given that more information does not necessarily lead to better

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 344 A Shollo & R D Galliers

decisions being made. But there is more to this argument than simply questioning whether or
not improved decision-making does occur.
This ‘traditional’ perspective views knowledge creation as a mental process and knowledge
as an ‘objectified commodity’ that can be transferred ‘as is’ from one point to another (Gherardi,
2000). It has been criticised for overemphasising technology and devaluing the human pro-
cesses of ‘sensemaking’ (cf. Weick, 1995) and knowing in organisations (Davenport, 1995;
Swan et al., 1999; Davison et al., 2012). Galliers & Newell (2003) draw attention to the distinc-
tion between data, information and knowledge, where data are context free and often located in
technological systems, but made sense of by the application of personal knowledge to become
informative, in a particular context, and for a specific purpose. As a result, the same data can
mean different things to different people. As Bateson succinctly puts it: information is a differ-
ence that makes a difference (i.e. as in generating an event, a sign, a symbol or an utterance)
(as cited in Mingers, 1995, p.289).
The knowing perspective, on the other hand, is rooted in actions and practices. It claims that
BI systems may facilitate these mental skills because knowledge creating and learning is pri-
marily a social and participative activity (Cook & Brown, 1999). Thus, instead of focusing on
supporting decision-making, where BI systems are treated as ‘a passive container’ that enables
knowledge to be stored, shared and analysed (Newell, 2014), a knowing perspective would
rather focus on how BI systems facilitate action, where BI systems are considered as an active
participant in knowledge work.
While the BI literature has not directly addressed the role of BI in organisational knowing per
se, BI systems have been shown to support selected aspects of organisational knowing. The
first traces of aspects of knowing incorporated in the literature are found, for example, in Steiger
(2010) and Wixom & Watson (2010). In particular, Steiger (2010; p. 29) has proposed a pre-
scriptive BI design theory as knowledge creation, based on Nonaka’s (1991, 1994) spiral
model, asking ‘how BI can be focused internally on the decision maker to discover and enhance
his/her mental model’. Wixom & Watson (2010) discuss the increasingly critical role of BI in or-
ganisations from being a decision support tool to becoming a prerequisite for organisational
success, highlighting the ‘transformative – sometimes unanticipated – power of BI’ in organisa-
tions (Wixom & Watson, 2010; p. 14). They call these organisations BI-based organisations,
which may be viewed as ‘knowing’ organisations where knowing is facilitated (or driven) by BI
systems. Although notions of knowing are camouflaged in their papers, neither Steiger (2010)
nor Wixom & Watson (2010) make any clear distinction between knowing and decision support.
Several studies have reported that BI systems facilitate organisational decision-making by
identifying opportunities and problems (Truxillo et al., 2012) and by detecting trends or patterns
in customer behaviours (Chau & Xu, 2012), operations (VanDiver et al., 2009) and business
processes (Elbashir & Williams, 2007). In a recent special issue on BI in MIS Quarterly, many
of the articles see BI systems as a technology that focuses on extracting intelligence or new in-
sights from implementing BI methods and techniques (Abbasi et al., 2012; Hu et al., 2012; Lau
et al., 2012; Park et al., 2012; Sahoo et al., 2012). The process of extracting insights and using
these to drive action is directly linked to processes of organisational knowing (Choo, 1998).
While we acknowledge such extensions to the decision-support literature, we go a step fur-
ther in this paper and explain how BI systems impact practice. In particular, we investigate

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 345

how knowing is mediated by technology, and more specifically, by BI systems. In the next sec-
tion, therefore, we provide an overview of the literature on organisational knowing as an ‘active
process’

O R G A N I S AT I O N A L K N OW I N G A S A N " AC T I V E P R O C E S S "

According to Choo (2002), the cycle of organisational knowing comprises three interwoven pro-
cesses – sensemaking, knowledge creating and decision-making – through which information
flows in organisations. Sensemaking processes are triggered from environmental stimuli and
have, as an outcome, less equivocal information, shared meanings and purpose. Knowledge
seeking and creating commences ‘when the organisation experiences gaps in its existing
knowledge or limitations in its current capabilities’ (Choo, 2002; p. 79). Such experience is
based on shared meanings and agendas as they emerge from the sensemaking process. Thus,
new knowledge is produced by organisational actors who share, link and synthesise their
existing knowledge – both tacit and explicit – through their (inter)actions – and their contesta-
tions. This knowledge-creating process generates new capabilities and improves existing ones,
making ‘possible new alternatives and outcomes, expanding the range of available
organisational responses’ (Choo, 2002; p. 80).
Choo’s description of organisational knowing is in line with the views of Polanyi (1958, 1966),
Tsoukas (2009) and Kolb (1984) in that knowledge creation in organisations starts from an in-
dividual’s immediate or concrete experiences and actions (Kolb & Kolb, 2005). These form the
basis for observations and reflections that result in new distinctions being made. These new
distinctions are assimilated and distilled into abstract concepts from which new implications
for action can be drawn through negotiation. These implications can be actively tested and
can serve as guides in creating new learning (Kolb & Kolb, 2005). ‘When new distinctions are
made and accepted, new organisational knowledge emerges and when the new distinctions
are developed into new products or processes, or are embodied in new actions, innovation
and learning respectively occur’ (Tsoukas, 2009; p. 2). Thus, knowledge is applied in the minds
of individuals, enabling new distinctions to be made1 but it is only when it is communicated to
others, transformed (based on discussions, negotiations and contestations), and accepted by
the community, are these new distinctions converted into organisational knowledge.2
We do not, however, claim that knowing consists of two linear phases and sequential steps,
but we do want to emphasise the deep interplay that exists between the cognitive and the social
approaches. In this paper, we adopt the mutual constitution view (Marabelli & Newell; 2012).
According to this view, knowledge is seen as a thing that can be possessed by individuals or
groups while knowing is seen as the working knowledge practices that socially construct our
world. These two dimensions can be perceived and used at different organisational levels
(e.g. individual, group and organisation). Boland & Tenkasi (1995) describe the process of per-
spective making and perspective taking through which knowledge work is accomplished and

1
This phase of knowing is approximate to the cognitive approach.
2
This phase is approximate to the social aspects of knowing.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 346 A Shollo & R D Galliers

new knowledge is created. This organisational knowledge drives actions through which
organisational learning occurs – the ‘generative dance’ of knowledge and knowing, to which
Cook & Brown (1999) refer.
Thus, we summarise the concept of knowing as ‘an active process’ (see Figure 2(a)) of
making new distinctions accepted in organisational settings and embodied in organisational
changes from which learning occurs. Individual and group knowledge, as well as explicit and
tacit knowledge are brought into play by knowing when interacting with the social and physical
world (Cook & Brown, 1999). Following Orlikowski (2002; p. 249), we take the view that ‘know-
ing in practice … highlights the essential role of human action in knowing how to get things done
in complex organizational work … [a] perspective [that] suggests that knowing is not a static
embedded capability or stable disposition of actors, but rather an ongoing social accomplish-
ment, constituted and reconstituted as actors engage the world in practice’.
We acknowledge that any such depiction as given in Figure 2(a) does not and cannot capture
the full extent of the complexities of knowing described previously as it shows knowing as a pro-
cess of sequential phases. The process of knowing is much more complex and messy. An
attempt to capture this messiness is provided in Figure 2(b), which illustrates the contestation
and negotiation that form a critical component of knowing in organisations (Kuhn & Jackson,
2008). While a simplification, the Figures nonetheless capture the essence of the previously
menitoned arguments, and were a helpful sensemaking (cf. Weick, 1995) device in terms of
organising and sorting through the data we collected in the field and its subsequent analysis.
To study knowing as something that people do means to analyse the dynamics of the socio-
technical systems through which knowing is accomplished (Blackler, 1995). In this way, know-
ing is analysed as a phenomenon that is mediated by language, technology, collaboration and
control, including the ‘material forms, artifacts, spaces, and infrastructures’ of which Orlikowski
(2006; p. 460) speaks. It is situated in particular contexts (time and space), provisional
(because it is constantly developing) and pragmatic (in terms of serving a specific purpose)
(Blackler, 1995).
Mapping the BI research into the active process of organisational knowing in Figure 2, it is
evident that the literature has focused mainly on providing evidence that BI systems facilitate

Figure 2. (a) Knowing in organisations as an ‘active process’, (b) Knowing in organisations as an ‘active messy process’.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 347

the emergence of new distinctions. However, the BI literature is scant regarding the role BI
might play in the transformation of personal knowledge into organisational knowledge, and
the latter into actions and learning. Thus, for the most part at least, the social dimension of
knowing is currently missing from the BI literature, although new conceptual studies are begin-
ning to appear, underlining this social dimension and the role BI has beyond decision-making.
Imhoff and White (2010), for example, although still focusing on decisions, point to the need for
collaboration between BI producers and consumers in order to understand, interpret and
contextualise the results. They bring to BI system designers’ attention the fact that not all infor-
mation workers in organisations are necessarily decision makers; they may perform different
roles such as BI producer, BI consumer or collaborator – ‘the “commentators” of the business’
in their words (Imhoff and White, 2010; p. 46). In the same spirit, Stoodley (2012) provides in-
dications as to how greater collaboration and better communication can be a positive side effect
of using BI systems in multiple business units, with analysts closely collaborating with decision
makers to understand and solve abstract organisational problems.
In light of the above and based on the concept of knowing as an active process – one that is
mediated by IT – we conducted a case study to examine how BI systems mediate knowing in
organisations. We thus focus on how BI systems can assist other organisational actors in iden-
tifying differences that to them make a difference in their everyday working lives. More precisely,
our research question is How do BI systems facilitate knowing in organisational settings? In an-
swering this question, we aim to shed light on how new distinctions emerge and how BI facili-
tates actions and learning within organisations.

R E S E A R C H D E S I G N A N D P R E S E N TAT I O N O F T H E C A S E S T U DY

The empirical basis for this research is an illustrative interpretive case study that investigates
the role and the use of BI in an organisational setting. Case studies are particularly valuable
for exploratory research where a thorough understanding of a phenomenon in its context is pre-
ferred (Conboy et al., 2012; Walsham, 1993). Another reason for the choice of case study as a
research method is the nature of the question that is being investigated. Case studies are best
suited for investigating ‘how’ questions (George & Bennett, 2004) and thus, in this instance, the
mechanisms through which BI systems might impact the process of organisational knowing.
Our empirical setting is a Scandinavian financial institution. The organisation is recognised
as a successful financial company with a high-market share, employing more than 20,000 em-
ployees. The study was conducted in different units of the organisation (e.g. different business
units, the IT unit and units such as branch offices that had direct contact with clients). Our
choice was based on the fact that this is an organisation where state-of-the-art BI solutions
(such as performance management systems, online analytical processing and reporting and
querying software tools), are employed to support and improve decision-making, and where de-
cision makers have a wide repertoire of experiences and in-depth knowledge of the domain.
The branch advisors, their managers and the BI analysts use BI tools (such as Query
Analyser, performance management systems and Excel spread sheets) to perform a variety
of data analyses. The performance management system is used to monitor the performance

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 348 A Shollo & R D Galliers

of different branches throughout the country. For example, managers can see how many new
customers they may have each month, how many meetings advisors have with their clients,
how successful such meetings are (e.g. in concluding a loan) and the financial details of such
deals (e.g. the price or interest rate offered). All these data are stored in the corporate data
warehouse from whence they can be accessed either by the Query Analyser or through the per-
formance management system. The back-end technology used for data modelling is Microsoft
SQL Server using Data Transformation Services and SQL Server Integration Services. At the
front-end, the company uses Analysis Services 2008R2 for the BI cubes. The branch advisors
can view data about their own and their branch’s performance but not about the performance of
other branches. Branch managers can view their branch performance and the performance of
branches that belong to the same peer group, but not branches that are outside that group. The
BI analysts have access to the performance data of all branches.
These performance management activities are supported by a BI application called the
Group Management Information (GMI) system – based on an in-house tool called OLAPViewer.
The company is currently migrating the reports to a Sharepoint based site with Reporting Ser-
vices and Performance point. For self-service Microsoft products like PowerPivot and Tableau
are used. As one interviewee reported, GMI was held in high regard:

‘…[it] is a business intelligence system primarily targeted for the branch network to enable
them to make … decisions locally in the market, based on the best foundation.’ (Performance
Management Specialist)

Data collection

We conducted 10 interviews initially with key personnel in the organisation in 2010, and two
with external subject experts, to assist in triangulation (Denzin, 1978). These were followed
by a further four interviews roughly 2 years later in order to obtain insights into the use and im-
pact of the BI system over time. While limited, the interviews were sufficient for the purposes of
proof of concept. Background information (including organisation charts, reports, spreadsheets,
forms, PowerPoint presentations, memos and meeting minutes) was also collected and served
as complementary material to the interview data. Those interviewed used BI in their everyday
work and represented different managerial levels.
The form of the interviews was semi-structured (Wengraf, 2001) based on an interview guide.
The interviews started with demographic and open-ended questions followed by questions fo-
cusing on the interviewees’ daily work and use of BI systems. We also asked for specific exam-
ples in order to gain a better grasp of what was being said. Towards the end of the interviews, we
asked for further reflections on the examples given. Each interview was conducted in English in
the interviewee’s office and lasted an average of 60 minutes. Participants were informed that the
interview was about the use of BI in their daily activities, but were not shown the questions ahead
of time. All the interviews were recorded and transcribed with the consent of the interviewees.
Following analysis of the interviews, and as our data structure took shape, we conducted
follow-up interviews with key users. These assisted in obtaining some further understanding

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 349

of the system’s impact and use over time, particularly with respect to the examples given in the
previous round of interviews. As noted previously, four such interviews were conducted: two
with BI analysts, focusing primarily on use and two with an advisor and a branch manager to
help gauge the impact of these BI analyses on organisational knowing. Data collection, there-
fore, took place over a period of 2 years. While the number of interviews is limited, following cod-
ification, we assured ourselves that data saturation (Strauss & Corbin, 2008) had been reached
following the initial round of interviews – at least for the purposes of this explorative study. The
follow-up interviews were required to gauge how use and impact may have changed over time.

Data analysis

When reviewing the interview transcripts and background materials, together with the field
notes that had been taken to record our impressions at the time of each interview, we looked
specifically for indicators of how BI was used by individuals in their everyday practices. It is
these everyday practices that constitute our unit of analysis. We employed constant compara-
tive techniques and open coding (Strauss & Corbin, 2008) for analysing our data. Analysis of
the data was an iterative process during which we discussed codes until agreement was
reached (See Appendix B for an overview of the different steps followed during data analysis).
The codes captured concepts such as ‘BI fosters dialogue’ and ‘reflections on data’. Codes
were assigned to words, sentences or even paragraphs in the margins of the interview tran-
scripts. We captured 250 codes relevant to the role of BI in organisational knowing.
We organised these first-order codes into data tables in Excel spreadsheets that supported a
single theme or topic across data sources (Vendelo & Rerup, 2011). Every new relevant state-
ment was listed under its appropriate code. The interview-coding activity continued until it was
not possible to ascertain any more distinct, shared patterns among the data. In this manner,
theoretical saturation (Suddaby, 2006) was accomplished. During the next phase of the analy-
sis, codes that were recognised as similar were collated into the same first-order codes, using
informants’ language whenever possible (Strauss & Corbin, 2008). In parallel with the develop-
ment of the first-order codes, linkages among the categories started to surface and become
evident. These linkages were the seeds that initiated the development of second-order themes.
We sorted the data further by developing second-order themes, using the three research
subquestions to go through the first-order codes. We also analysed the data and first-order
codes by looking for overlap or linkages among first-order codes to assemble these into
higher-order themes. The process of developing the second-order themes involved many iter-
ative cycles. During these cycles, the first-order codes were revised, merged and sometimes
abandoned in order to reach a higher level of abstraction and to arrive at five second-order
themes (Vendelo & Rerup, 2011). Last, the five second-order themes were combined into two
aggregate dimensions that captured the overarching concepts relevant for understanding the
role of BI systems in organisational knowing. In this final step – through iterative analysis – ‘data
selection’ and ‘articulation’ (and their relationship to ‘organisational knowing’) emerged as
transparently observable phenomena (Eisenhardt, 1989). As a result, the data-structure
(Eisenhardt, 1989) presented in Appendix A was developed and this formed the foundation
for the analysis that is presented in the section that follows. We commence the Analysis

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 350 A Shollo & R D Galliers

section by recounting an illustrative story of organisational knowing, and the role of the BI sys-
tem in this process, as this serves as a foundation for the presentation of our results.

A N A LY S I S

An illustrative story

The story concerns branch advisors and the loan price quotations they give to their clients.
From 2000, the bank’s strategy focused to a large extent on lending growth. This meant that ad-
visors, branch managers and top management were all mainly concerned with monitoring and
controlling lending growth. The goal was to lend more no matter what the price was. Thus, pric-
ing was not perceived to be a priority and, as such, while prices were monitored, no action was
taken:

‘Nobody thought about how pricing was made or how we can save money on pricing.’
(Branch Performance Analyst, 2012)

Fees and loan rates were set by the bank centrally, being automatically calculated from the
loan system according to a defined set of criteria. However, advisors were not always complying
with the fees and rates that had been set, frequently giving discounts. Overall the following pic-
ture emerged:

‘… in every region you had some branches that were quite good at taking the prices and the
fees [for a loan, as the system suggested] but you had … a lot … that … gave some sort of
discount.’ (Branch Performance Analyst, 2010)

The pricing map across the country looked as reported in the succeeding sentences:

‘… a lot of the regions over here [Western part of the country] … gave a discount [for each
loan] pretty much; perhaps not all the time, but they gave a higher discount than they do in
and around [the capital]’ (Branch Performance Analyst, 2010)

In the period of 2008–2009, when the financial crisis hits the market, growth lending was not
perceived to be a viable solution anymore. Hence, top management began to search for other
ways to increase bank earnings. Middle managers from all departments were requested to
bring suggestions to the table. For example, in the pricing department:

‘…[the Senior Performance Analyst] given his experience and daily contact with the advisers
and the branches, knew that there were some differences in pricing between the advisers in
the branches. He knew something was wrong here, [but he] need[ed] to show it in a way so [that
he] could raise it with top management, saying we have an issue here.’ (Branch Performance
Analyst, 2012)

This ‘distinction’ combined with the personal knowledge of the Senior Performance Analyst
that this did not seem right, led to a dialogue between the branch performance analyst and

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 351

the branch managers, who were asked for their explanations. Local market conditions were
often cited by the branch managers:

‘So, some [branch managers] said, “Well it’s because the people [clients] living over here
[in the West], they won’t pay the full price” … for people in [the capital] price doesn’t matter.’
(Branch Performance Analyst, 2010)

This was a myth that had been around in the company for years. Every time that an explana-
tion was required on discounts given, the local market conditions would be cited or it was attrib-
uted to a special customer case. Although, the analysts in the headquarters had a hunch that
this was not the case but they could not prove the opposite. The same Branch Performance
Analyst, in the follow-up interview, described how they analysed the data in greater detail:

‘First of all we said, “Well, is there a difference between how advisers are pricing the loans?”
So, we could see that at the regional level, yes, there were differences. [The capital] and the
surrounding areas had higher prices than [in the West]… This made us look in more detail
… we could see there’s a difference between the branches in the same area. From that we
said, “OK, perhaps there’s a difference between the advisers within the same branch.” The
funny part was that we could draw the same curve on all levels.’ (Branch Performance
Analyst, 2012)

Developing the latter point, and because the BI system enabled comparison across
branches, the Branch Performance Analyst had previously made the following observation:

‘… no matter which branch you took … you could draw exactly the same curve. So you could
say, you have some advisors in the same branch … who take the price every time and …
some advisors who give a discount every time. When we looked at this [the BI analysis] …
we could see that, even in the Northern part of [the country], there were some branches that
actually could take the prices … even in the branches that performed badly … some advisors
could take the price. So what we actually learned … is that it’s not a matter of marketplace,
it’s more about the advisor’s behaviour and how he thinks about price.’ (Branch Performance
Analyst, 2010)

According to the Branch Performance Analyst, the BI system ‘made it possible to document
the [analysts’] thesis’. The new insights were communicated to all the branches, advisors and
top management. He indicated that, during the last 4–5 years, there had been a continuous
effort to change and improve the pricing procedures based on the new knowledge the organi-
sation was gaining from analysing performance data. When asked about the role of the BI
system in this reorientation, the Branch Performance Analyst reported that:

‘The analysis done on pricing using the BI tools has helped to develop the general strategy in
the bank, changing the way that we price products in general. [For example, we found out
that] we have a lot of customers who aren’t contributing to the bottom line of the bank.’
(Branch Performance Analyst, 2012)

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 352 A Shollo & R D Galliers

This realisation spurred new discussions between the analysts, the branch managers and
the advisors about sharing knowledge to improve price compliance. For example,

‘So, actually, it can be possible to change behaviour … what we did was we had … some
meetings … [As a result] what we actually achieved was to raise the bottom line [company’s
net earnings]. We’ve estimated that we made about €53 million on that… So that’s an area
where we took, you could say, best performance, and tried to make the others learn from
that.’ (Branch Performance Analyst, 2010)

Summarising then, and referring to Figure 1, we could see that branch performance analysts
used the BI system to raise awareness among senior management and the branches about
non-compliance with pricing guidelines. Old beliefs were changed, through discussion and
learning, based on reporting actual data, and new distinctions and beliefs were formed and
acted upon. We shall now consider our findings in greater depth.

The ongoing, cyclical practices of articulation and data selection triggered by BI systems
in organisational knowing

While investigating the role of BI systems in the process of organisational knowing and BI-
use-in-practice (cf. Whittington, 2006, 2014), we observed two practices that were initiated
and reinforced by the use of BI systems – the practices of articulation and of data selection.
Articulation is the coherent communication process of one’s beliefs, opinions and ideas
(Vocabulary, 2014). We found clear patterns of variations in how articulation was initiated
by BI. Articulation practices had three main aspects: articulating new distinctions; articulating
different perspectives and articulating organisational actions. The data selection practices
had two main aspects: data on demand and turning data into evidence.

Articulating new distinctions

New distinctions emerge from the interpretation of BI data that something requires further inves-
tigation and analysis. This notion is also captured in the interviews where the IT Finance
Business Analyst reports that, ‘… we really think that 90% of our BI use is to explain deviations;
explain something which looks odd’. From a BI systems perspective, what makes these distinc-
tions more visible is the increased analysis capability, because of the granularity, of BI systems.
However, it should be borne in mind that the BI data themselves do not guarantee the identifi-
cation of distinctions because these distinctions occur in the mind of the analyst. In our case,
the BI system user is only partially aware of the elements that contribute during interpretation:
the meaning of the data is their focal target. The following quote from a regional manager is
illustrative:

‘… it’s a little bit difficult to discuss all this because I think it’s the first time someone has asked
us such questions and, in a way, it’s a very good experience because now, suddenly, we’re
thinking a little bit more why we’re looking at all the figures …’ (Regional Manager, 2010)

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 353

Almost all the interviewees agreed that supplementing the data with personal knowledge is
key. Personal knowledge incorporates previous experience and expertise, common sense
and contextual knowledge. The following quote is representative:

‘I think we’ll always use our common sense because you have to see figures over a very long
period.’ (Business Analyst, 2010)

Up to this point, the articulation practice takes place between the BI system and the user’s
personal knowledge, thereby giving voice to new distinctions. These new distinctions are
framed by the interplay between the BI system and the user’s personal knowledge. In our story,
the analyst interpreted the results of the ‘higher discounts’ in the Western part of the country as
something ‘odd’, triggering contestation among various actors as to why this might be so. In an
attempt to make sense of this unexpected event, he contacted the branch managers whose ad-
visors were giving these higher discounts. The story that he was told was that the higher dis-
count was a result of the local market conditions.

Articulating different perspectives

In an attempt to make sense of, and find a plausible explanation about, the new distinctions that
emerge in the process of knowing, different perspectives come to the surface and are
articulated, contested and negotiated. For example, the advisors articulate the role of intangi-
bles in their daily work; analysts and middle managers articulate their target setting and em-
ployee management practices, and top management further articulate their strategic pricing
practices.
The BI analysis indicated that the local market conditions story did not hold. In the same
branch, there were advisors that gave discounts, while others did not. The new distinctions that
emerged in the mind of an employee are followed by an investigation as to why this distinction is
the case. As a result, discussions then commenced between the different stakeholders con-
cerned (branch managers, advisors, performance analysts and top management):

‘Showing the results to the advisors created a common awareness of the problems and is-
sues the branch and individual advisers have. This awareness about pricing has created
room to talk about it and has helped … the advisors to talk about their difficulties.’ (Branch
Performance Analyst, 2012)

The BI system users viewed BI systems as tools that could foster dialogue between decision-
makers, stimulating reflection and discussion around less tangible factors. Additionally, the BI
system enabled this analysis at different levels. For example, in our story, we observe that
the problem was discussed based on the data that were extracted from the BI system at differ-
ent organisational levels, with a more holistic perspective being taken as a result:

‘… it creates a dialogue … a conversation in the executive committee …“do we actually have

the right price?” … and that creates a sort of decision-making process …’ (Performance
Management Specialist, 2010)

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 354 A Shollo & R D Galliers

‘Some advisors told me that it was very difficult to get new customers if they couldn’t give a
discount, [but] then we discussed about giving 10% discount instead of 50%, and it worked
for them.’ (Branch Manager, 2010)

However, users reported that the BI data were not always self-explanatory; for example, ‘So,
even though we have measures that give us a good … idea how it’s going, it doesn’t tell the
whole truth.’ (Regional Manager, 2010) Thus, the BI data captured only part of the picture
and, as a result, there was a need to investigate the intangible elements that are not captured
in the BI data:

‘So, the scorecard is a foundation for measuring performance, no doubt about that, but other
factors … [like] … a local initiative performed by the branch manager … or it could be the way
that he’s handling the staff, or … other information that he’s … extracting from the market
[to provide] to the management.’ (Performance Management Specialist, 2010)

Thus, even if new knowledge is obtained from an understanding of associated facts, this
knowledge is based only on the data available in the BI system: ’soft facts’ – qualitative facts
that are not in the system – are missing. It is at this point that the process of knowing shifts from
being (socio-)technology oriented to being socially oriented.

Articulating organisational actions

The fact that BI systems enable comparisons across different units, integrating data from differ-
ent systems facilitates the surfacing of common patterns, as illustrated in our story by looking at
trends over time. In addition, because BI systems facilitate benchmarking and the crystallisation
of common patterns, they inherently facilitate organisational actions to be taken:

‘I think this [BI system] has taken the dialogue from one step to another; has made it more
concrete and action-oriented. Instead of discussing what these figures are … we’re talking
about what should we do to improve those figures.’ (Branch Performance Analyst, 2012)

The comparisons facilitated knowledge sharing and learning. First, the comparisons exposed
which units (branches, managers or advisors) needed to communicate and share knowledge.
Second, knowledge sharing discussions could focus on specific actions that had been taken
or should have been taken – past actions in other words or actions that they believed needed
to be taken in the future, whether on the part of an employee, a department or a particular busi-
ness area within the bank:

‘… I try to find out what is it that’s wrong in one place, what is it that’s good in another place,
so we can learn from each other … I call them and say, “What do you do since you’re so
good?” And “What’s the problem, ’cos you’re in the red?”…’ (Business Analyst, 2010)

‘The branch managers are like bookkeepers; they need facts to act.’ (Branch Performance
Director, 2012)

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 355

However, we should make the point that organisational attention and strategic goals appear
to play as significant a role in how much the BI system actually impacts or facilitates
organisational action. Although observations are made and documented with the help of the
BI system in instances when a particular observation concerns a topic or issue that is not seen,
at the time, as being particularly relevant or high priority, then actions are not taken:

‘They can’t stay focused on it because they [the organisation] change the measures [where
you have focus] all the time. You have to have meetings once in three months to keep them
on track, otherwise their performance decreases.’ (Branch Performance Director, 2012)

Data on demand

Our analysis has shown that articulation of new distinctions, different perspectives and
organisational actions were each associated with differences, both in the amount and aggrega-
tion of data selected from the BI system. Data selection practices involved filtering of specific
data fields, dimensions and measures, which are collected and integrated, in order to investi-
gate a phenomenon or to measure different indicators. The data selection practices had two
main themes: data on demand and turning data into evidence.
From a BI system perspective, a ‘drill down’ of the data is performed where very specific
accountable variables are identified for the problem in question. The availability of data (data
on demand) at different levels of aggregation makes it possible to address different stake-
holders, decision makers and action takers. The ‘drill down’ and the ‘roll up’ activities that users
are able to perform within the BI system provide visibility and transparency in terms of the
measures and how they are calculated. Branch managers can have an overview of their
branch, but they can also drill down to identify their weakest link or where exactly their weak-
nesses lie. The advisors can track their performance over time and are also able to see the
effect they have on a branch level through data roll up. It is precisely because of these drilling
down and rolling up capabilities that knowledge sharing discussions can focus on specifics.
This is illustrated by the following:

‘I think it’s very sophisticated that you can go down to advisor level and … into the branch and
ask why you didn’t take the price … maybe there’ll be a good reason, but I think it gives the
manager a very good overview.’ (Regional Manager, 2010)

Turning data into evidence

The role of the BI system and the power of the data analysis in the dialogue (and the contesta-
tions and negotiations that take place) vary considerably, depending on the assumptions on
which the analysis is undertaken and on participant views:

‘… you need the data because that’s the only way that you can convince people that it’s a
serious problem.’ (Head of IT Credit Processes, 2010)

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 356 A Shollo & R D Galliers

The more data made available, or analysis undertaken, at several levels that lead to the same
result, the more trustworthy the BI data and the analysis become. For example, in our illustrative
story, the presentation that was given to top management, the regional and branch managers
and the advisors included several – 17 in fact – charts3 that presented the argument at different
levels and with many different measures:

‘When I showed the first slide they said, “Well, where did you get that from?” And, “All the
figures in our system are wrong”, and so on. And then I had ten observations, and, as you
can see, they all show the same.’ (Branch Performance Director, 2012)

The availability of data over a long period contributes to the perceived accuracy of the anal-
ysis, which in turn contributes to the persuasive power of the analysis when presented to differ-
ent stakeholder groups. In order to create a strong consensus based on the data, enabling
actions to be taken, the time dimension plays an important role along with its perceived quality.
The following shows that trends observed in the data over time have a stronger impact:

‘…you have to see figures over a very long period. If one of our best branch managers has
very bad figures, we’ll look at why, of course … but if you, year after year, have bad figures
compared with similar branches [then you’re sure something’s wrong]. It’s important [to have]
those figures, not in the short run but in the long run.’ (Business Analyst, 2010)

The transparency of the analysis is another factor that appears to strengthen the persuasive
power of the BI analysis. As mentioned earlier, all the concerned parties can see how the anal-
ysis is created. They can drill down to the raw data and, as a result, can become aware of the
underlying assumption or limitations of the BI analysis.

‘It’s a very strong argument to use that I have not found the data from some special system [of
mine]. I found these data from systems that all of you [branch managers and advisors] have
access to.’ (Branch Performance Director, 2012)

It seems, therefore, that the combination of an analysis undertaken at several levels, which
includes data from a period of time and is transparent to all concerned, grants to BI data persua-
sive power and legitimises its use in discussions and negotiations in the process of knowing.
However, issues such as data quality in BI implementations are still present and inhibit organi-
sations in getting the most out of BI initiatives. For example,

‘We’ve also had a couple of instances [where] it turned out that that the numbers in the
[BI analysis] were wrong because we were double counting, or we were comparing with num-
bers that were called the same so we thought we could compare them.’ (Head of Strategic
Business Unit, 2011)

The importance of data quality in order to leverage BI in organisational knowing is illustrated

by the following:

3
Taken from the pricing presentation document.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 357

‘Because if you don’t have the right data you can make as many fancy things as you like:
rubbish and rubbish. You need to have the data; these have to be right.’ (Head of IT Credit
Processes, 2010)

Such issues like these continue to operate as limitations to the potential beneficial effects of
BI in organisational decision-making, and need to be taken into account in our conceptualisa-
tion of the role of BI systems in organisational knowing – a topic to which we now turn.

DISCUSSION

The purpose of this research has been to investigate how BI systems facilitate knowing in
organisational settings in the context of everyday practices. We argue that BI systems facilitate
organisational knowing through two main BI-induced practices: articulation and data selection.
BI systems make it possible to articulate hypotheses that might arise from intuition, gut feeling
or previous beliefs and experiences, based on a selection of data that may not have been avail-
able previously. In the following paragraphs, we consider in detail the questions that had previ-
ously been derived, based on the active processes of knowing, specifically in relation to how BI
facilitates the emergence of new distinctions, organisational knowledge and organisational
actions and learning.
Because of the integration of data across different networks and business domains, cross-
network analysis may reveal previously unknown patterns, providing of course, that like is being
compared with like. With the integration and selection of specific data from different domains,
reshuffling and comparing these data with prior data, previously unknown patterns emerge,
leading to new insights that can be shared and debated among individuals. This finding sup-
ports earlier conclusions drawn from the BI literature where new insights are said to be one
of the main benefits that arise from BI systems use (Elbashir and Williams, 2007; Chau & Xu,
2012; Truxillo et al., 2012). The current study extends this earlier work, however, by
emphasising how these new distinctions are made and the active role of BI systems in this pro-
cess. New distinctions are made through the underlining practices of articulation and data
selection, primarily between the individual user and the BI system. The cyclical nature of ‘data
selection’ and ‘articulation’ (cf. Figure 3), which takes place throughout BI system use, mani-
fests itself at an individual and an organisational level. The BI system acquires an active influ-
ential role during the contestations and negotiations that take place in the organisational
knowing process. Individuals using a BI system or analysis (the product of the system) ‘take’
sense from the data; they add meaning, in other words. The meaning that emerges from the
interaction of the individual with the BI analysis, originates from the data through practices of
selection and articulation. The following example illustrates the point when the Senior Perfor-
mance Analyst analysed the advisors’ performance data across the branch network, he
engaged in data selection, and as a result of this, a new distinction emerged. In an attempt to
make sense of this distinction, he engaged with the system in articulation practices by
analysing data for different levels (regional, similar branches and branch level). In this way he
could conclude that the pricing ‘leakage’ was not a matter of the market place but depended

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 358 A Shollo & R D Galliers

Figure 3. The role of BI systems in organisational knowing.

on individual advisors’ behaviours. He was then able to articulate this conclusion to colleagues
and persuade then to adopt his view.
The BI system has an active role in both practices. During data selection, knowledge is (re)
produced where specific elements are brought to the foreground, and during articulation, these
elements take on different meaning depending on the mix of ‘actors’ (users, BI tools and data)
that are present. Newell (2014) calls this the translation process, where people and objects are
mediators that actively translate knowledge. Thus, this study provides empirical evidence of the
performative outcome of BI systems in relation to knowledge translation through the practices of
articulation and data selection.
As regards to the facilitation of organisational knowledge, apart from the role of BI systems in
revealing new insights on the part of individuals, the BI data themselves play another important
role in transforming these insights into organisational knowledge that can then be utilised in tak-
ing action. To make sense of these previously unknown patterns, individuals engage in dialogue
with others in the organisation, and with the system itself. As noted previously, contestation and
negotiation may occur (Kuhn & Jackson, 2008), but BI analyses stimulate problem articulation
from different angles, interpretations and perspectives. These are the dialogical exchanges and
the perspective taking of which Tsoukas (2009) and Boland and Tenkasi (1995) speak. When
productive, dialogue leads to ‘self-distanciation’: to individuals distancing themselves from their
customary, perhaps less reflexive ways. Boland and Tenkasi (1995; p. 357) argue that knowl-
edge integration ‘is a problem of perspective taking’ – the process through which ‘distinctive
individual knowledge is exchanged, evaluated, and integrated with that of others in the
organisation’. Our data suggest that BI systems can be the catalyst that just stimulates such
a dialogical exchange.
Further, this problem articulation can lead to shared understanding at the organisational
level. As Brown (1981) observed, effective communicating requires that the point of view of
the ‘other’ be realistically imagined. Others, such as Rommetveit (1980; p. 126), have affirmed
this point: ‘An essential component of communicative competence in a pluralistic social world
… is our capacity to adopt the perspectives of different others’. The capability of BI systems

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 359

to enable people to drill down and roll up data, enables them to track the data at each step,
thereby facilitating discussions about the assumptions underpinning the analysis, which leads
in turn to adoption or better understanding of other perspectives.
According to Schultze (2000; p. 7), seen ‘from the perspective of the sociology of knowledge,
the difficulty in … knowledge creation lies in convincing others of the reliability and validity of
[their] … knowledge claims’. When applied to our analysis, this suggests that BI systems can
facilitate the articulation of tacit knowledge not only because of an individual’s ability to articu-
late knowledge, but also by minimising ‘the constraints that others in the social community im-
pose on the articulation’ (Schultze, 2000; p. 7).
Our data support Schultze’s argument that BI systems provide the appropriate ‘format that
knowledge needs to take in order to be acceptable to others and the evidence that knowledge
workers must provide to support their new knowledge claims’ (Schultze, 2000; p. 7). For example,
the data show how managers use data selection practices to turn data into evidence by using
many observations of the same phenomenon and, at the same time, providing transparency be-
cause every step in the analysis can be traced back to data in the system. Thus, BI systems con-
tribute to the balancing of subjectivity and objectivity discussed by Schultze, where subjective
insights and tacit knowledge are articulated in a way that, with the backing of BI data, become
acceptable and appreciable. This is so because of BI systems’ capability to provide data at differ-
ent levels of analysis, thereby enabling comparisons and cross analyses, fostering dialogue and
surfacing new distinctions and insights. It seems, therefore, that the combination of a BI analysis,
undertaken at several levels (which includes longitudinal data, and is transparent to all concerned
parties), grants to BI data persuasive power and legitimises its use in discussions and negotiations
in the ‘knowing’ process. This observation might also explain the increasing interest that industry
and academia are showing in BI and ‘big data’ currently (Davenport et al., 2012).4
At an organisational level, BI systems facilitate an interpretive process in which actors influence
each other through data-driven discussions. Such discussions can lead to the process of perspec-
tive making (Boland & Tenkasi, 1995), where the individual investigates and confirms the new
meaning that emerges from the data via a more in-depth consideration of the analysis. This new
meaning, once argued and resolved, is transformed into a narrative in the person’s community
(cf. Popovic et al., 2014), thereby reinforcing organisational knowing and strengthening perspec-
tive making by articulation practices. When this new knowledge concerns another community, it
also triggers the process of perspective taking in that community. Through this process, each type
of expertise present can take its perspective and assist other actors with different expertise to eas-
ily recognise and accept more the different ways of knowing others. Thus, BI systems can be an
active agent in the perspective taking process, which facilitates the utilisation of distinct knowledge
through dialogical exchanges – at different levels of analysis – by triggering the practices of data
selection and articulation. BI data, as artefacts, may thereby act as an active agent for a collective
process of discussing and negotiating articulated beliefs and practices.
As regards to the facilitation of organisational actions and learning, the study shows that ac-
tions are more likely to occur when a common understanding is achieved or when new

4
This finding also answers the recent call of the editors of an EJIS special issue (Sharma et al., 2014) on
further research on how business analytics can be used to improve the acceptance of decisions.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 360 A Shollo & R D Galliers

knowledge is accepted (albeit after negotiation) by the different stakeholders. This understand-
ing and acceptance is achieved through the practices of articulating required actions, based on
data comparisons or benchmarking practices.
However, as reported during the follow-up interviews, there is considerably less benefit to be
gained from BI systems and analyses when management focuses on topics other than those
that arise from the analysis since less actions are taken. That is, even if employees gain new
insights from analysing BI data, they do not necessarily act upon these new insights if their su-
periors are focusing on other organisational issues.
Two years from the initial round of interviews, the focus had shifted to other organisational
goals (e.g. from pricing, to credit, and later to customer satisfaction), with the result that branch
managers were reluctant to act on knowledge acquired 2 years earlier with the same allegiance
or persistence. Indeed, it is the case that the pricing curve is now reverting to the shape it
displayed before the BI analysis had taken place. As a result, the performance of advisors in
terms of pricing levels has started to decrease to the levels that existed before the knowing cy-
cle triggered as a result of the BI data analysis.
The role of BI as an active agent in the process of organizational knowing has, in this
instance, become less influential and other agents have taken a more important role. As such,
organizational attention plays an important role in how much, and for how long, the cycle of
organisational knowing will continue to occur through the practices of selection and articulation
that are facilitated by BI systems.

CONCLUSION

In summary, instead of simply focusing on the impact of BI systems on supporting decisions – as

tends to be the case in the ‘traditional’, mainstream IS/BI literature – we have focused on the concept
of knowing and have investigated how BI systems impact the knowledge work of individuals in their
daily practices. As a result, we believe the contribution of the study to be twofold. First, we have de-
veloped a conceptual framework of organisational knowing processes, based on the extant literature,
and used this as sensemaking (cf. Weick, 1995) device in this case study. Second, and what we see
as the main contribution, we have highlighted the cyclical nature of the ‘data selection’ and ‘articula-
tion’ practices triggered by BI systems in the cycle of organisational knowing. With this, we provide
empirical evidence of the agency of BI systems in the active process of knowing. Thus, we argue that
BI systems can more usefully be considered as an active participant in knowledge work.
With this new perspective as a background, we articulate the practices through which BI tech-
nologies can facilitate the knowing process as well as highlighting their limitations given that
other factors may come into play (institutional, political etc.). So, while the impact of BI in our
case was considerable at first, we note that the bank went back to its old ways over time, with
the appearance of different managerial imperatives.
This is not to say that there are no limitations or boundaries to the agency of BI systems and
to what analysts and users of BI systems can do. The interviews show that there are serious
issues involved, both with the implementation of BI systems in organisations and their use,
as noted for example by the Head of Strategic Business Unit and Head of IT Credit Processes.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 361

Our data show that data quality remains one of the concerns of BI users. The poor quality of
BI data – thus leading to a questioning of whether these data are indeed ‘facts’ – is another is-
sue with which organisations may have to deal. This is an issue that leads to a decrease in BI
systems use (DeLone & McLean, 2003) and which in turn decreases its effect on organisational
knowing. High data quality – preceived high quality data, at least – is a prerequisite for the BI
systems and data to have their intended effects on organisational knowing. Nonetheless, as
our interviewees reported, data quality becomes less of an issue when there are a large num-
ber of observations over a long period where an overall trend is clearly visible.
There are a number of limitations to this study. First, while we acknowledge contestation and ne-
gotiation, we have not accounted for power considerations and how they might affect the use of BI
data. Especially interesting would be to investigate how meaning power is exercised on BI data and
how BI data impacts or redistributes meaning power (Azad & Faraj, 2011). Further study, based on
such power considerations (e.g., Pfeffer 1981; Langley 1989) would be useful in this context.
A second limitation is the single case study method. We cannot conclude from this study that
the case organisation is in any way representative of the manner in which BI systems are used
in all organisational settings. We make a more limited and modest claim. The way BI is de-
scribed to mediate organisational knowing through the practices of data selection and articula-
tion, in this particular case, is a step on the way to understanding how BI systems may improve
organisational knowing. Further studies, utilising the model that was developed arising from this
case, would help to test its generalisability. Thus, our findings should be interpreted with caution
since they are based on a single, albeit longitudinal, case study.
A third limitation relates to our interview data. Since the interviews were conducted in English,
which is not the native language of the participants, there may well have been some linguistic con-
straints when those interviewed attempted to express themselves. While care was taken to have
the interviewees check our transcripts, and we held follow-up interviews at a later stage to reassure
ourselves on this score, this remains a limitation, which further case research could help obviate.
Another limitation is the somewhat linear representation of the organisational knowing pro-
cess as depicted in Figure 3. We acknowledge, as mentioned earlier, that the process of know-
ing is nonlinear nor sequential but rather messy, with contestation and negotiation often
present. However, for the sake of clarity, we opted to show this version of the figure instead
of a more messy version similar to Figure 2(b).
Further, it would be also imperative to conduct future studies that would support, extend or
reject our findings in other cultural settings (Hong et al., 2006; Leidner, 2010). Thus, and despite
these limitations, we trust that the use of a practice lens on BI opens up a new departure for BI
research that considers the implications of BI systems in organisations with actual practice in
mind, and these initial findings will serve as a basis for future research that can be undertaken
to confirm, extend and challenge our conclusions.

AC K N OW L E D G E M E N T S

The authors would like to thank Sue Newell, Ioanna Constantiou, Marco Marabelli, Morten Thanning
Vendelo, Mari-Klara Stein and Opal Leung for their valuable feedback on previous versions of this paper.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 362 A Shollo & R D Galliers

REFERENCES

Abbasi, A., Albrecht, C., Vance, A. & Hansen, J. (2012) development, research focus and direction. MIS
MetaFraud: a meta-learning framework for detecting fi- Quarterly, 31, 579–615.
nancial fraud. MIS Quarterly, 36, 1293–1327. Conboy, K., Fitzgerald, G. & Mathiassen, L. (2012) Qualita-
Arnott, D. & Pervan, G. (2005) A critical analysis of deci- tive methods research in information systems: motiva-
sion support systems research. Journal of Information tions, themes, and contributions. European Journal of
Technology, 20, 67–87. Information Systems, 21, 113–118.
Arnott, D. & Pervan, G. (2014) A critical analysis of deci- Cook, S.N. & Brown, S.J. (1999) Bridging epistemologies:
sion support systems research revisited: the rise of de- the generative dance between organizational knowledge
sign science. Journal of Information Technology, 29, and organizational knowing. Organization Science, 10,
269–293. DOI:10.1057/jit.2014.16. 382–400.
Articulation [Def. 2]. (2014) Vocabulary Online. In Vocabu- Davenport, T.H. (1995) Why reengineering failed: The fad
lary. Retrieved January 2, 2014, from http://www.vocabu- that forgot people. Fast company, Premier Issue, 70–74.
lary.com/dictionary/articulation Davenport, T.H. (2006) Competing on analytics. Harvard
Azad, B. & Faraj, S. (2011) Social power and information Business Review, 84, 98–107.
technology implementation: a contentious framing lens. Davenport, T.H. (2010) BI and organizational decisions.
Information Systems Journal, 21, 33–61. International Journal of Business Intelligence Research,
Blackler, F. (1995) Knowledge, knowledge work and orga- 1, 1–12.
nizations: an overview and interpretation. Organization Davenport, T.H., Barth, P. & Bean, R. (2012) How ‘Big Data’ is
Studies, 16, 1021–1046. different. MIT Sloan Management Review, 54, 22–24.
Boland, J.R. & Tenkasi, V.R. (1995) Perspective making Davison, R.M., Ou, C.X. & Martinsons, M.G. (2012) Infor-
and perspective taking in communities of knowing. mation technology to support informal knowledge shar-
Organization Science, 6, 350–372. ing. Information Systems Journal, 23, 89–109.
Brown, R. (1981) Social Psychology. The Free Press, New York. DeLone, W.H. & McLean, E.R. (2003) The Delone and
Brydon, M. & Gemino, A. (2008) You’ve data mined. Now Mclean model of information systems success: a ten-
what? Communications of the Association for Informa- year update. Journal of Management Information Sys-
tion Systems, 22, 33. tems, 19, 9–30.
Chau, M. & Xu, J. (2012) Business intelligence in blogs: Denzin, N.K. (1978) The Research Act: A Theoretical
understanding consumer interactions and communities. Introduction to Sociological Methods. McGraw-Hill,
MIS Quarterly, 36, 1189–1216. New York.
Chen, H., Chiang, R. & Storey, V. (2012) Business intelli- Eisenhardt, K.M. (1989) Building theories from case
gence and analytics: from big data to big impact. MIS study research. Academy of Management Review,
Quarterly, 36, 1165–1188. 14, 532–550.
Cheng, S., Dai, R., Xu, W. & Shi, Y. (2006) Research on Elbashir, M.Z. & Williams, S. (2007) BI impact: the assimi-
data mining and knowledge management and its appli- lation of business intelligence into core business pro-
cations in China’s economic development: significance cesses. Business Intelligence Journal, 12, 45–54.
and trend. International Journal of Information Technol- Galliers, R.D. & Newell, S. (2001) Back to the future: from
ogy & Decision Making, 5, 585–596. knowledge management to the management of informa-
Choo, W.C. (1998) The Knowing Organization: How Orga- tion and data. Information Systems and e-Business
nizations Use Information To Construct Meaning, Create Management, 1, 5–13.
Knowledge, and Make Decisions, 2nd edn. Oxford Galliers, R.D. & Newell, S. (2003) Information systems
University Press, New York. strategy as data plus sense-making. In Images of Strat-
Choo, W.C. (2002) Sensemaking, knowledge creation, and egy, Cummings, S. & Wilson, D.C. (eds.), pp. 164–196.
decision making: organizational knowing as emergent Blackwell, Oxford.
strategy. In Strategic Management of Intellectual Capital Galliers, R.D. & Swan, J.A. (1999) Information systems
and Organizational Knowledge, Choo C.W., Bontis N. and strategic change: a critical review of business
(eds), pp. 79–89. Oxford University Press, Oxford. process Re-engineering, In: Rethinking Management
Clark, T.D., Jones, M.C. & Armstrong, C. (2007) The dy- Information Systems, Currie, W.L. & Galliers, R.D.
namic structure of management support systems: theory (eds), pp. 361–387. Oxford, Oxford University Press.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 363

Gartner Press Release. (2014) Predicts 2014: Don’t Try to Langley, A. (1989) In search of rationality: the purposes be-
Prevent the Digital Revolution, Exploit IT Now https:// hind the use of formal analysis in organizations. Admin-
www.gartner.com/doc/2642621 istrative Science Quarterly, 34, 598–631.
George, A. & Bennett, A. (2004) Case Studies and Theory De- Lau, R., Liao, S., Wong, K.F. & Chiu, D. (2012) Web 2.0 en-
velopment in the Social Sciences. MIT Press, Cambridge. vironmental scanning and adaptive decision support for
Gherardi, S. (2000) Practice-based theorizing on learning business mergers and acquisitions. MIS Quarterly, 36,
and knowing in organizations. Organization, 7, 211–223. 1239–1268.
Golfarelli, M., Rizzi, S. & Cella, I. (2004) Beyond data Leidner, D.E. (2000a) Knowledge Management and Knowl-
warehousing: what’s next in business intelligence?. In edge Management Systems. Elsevier, Oxford.
Proceedings of the 7th ACM international workshop on Leidner, D.E. (Ed) (2000b) Knowledge management and
Data warehousing and OLAP (pp. 1-6). ACM. knowledge management systems. The Journal of
Gorry, G.A. & Scott-Morton, M.S. (1971) A framework for Strategic Information Systems, 9, 101–262.
management information systems. MIT Sloan Manage- Leidner, D.E. (2010) Globalization, culture, and informa-
ment Review, 13, 55–72. tion: towards global knowledge transparency. The Jour-
Henfridsson, O. & Lind, M. (2014) Information systems nal of Strategic Information Systems, 19, 69–77.
strategizing, organizational sub-communities, and the Lönnqvist, A. & Pirttimäki, V. (2006) The measurement of
emergence of a sustainability strategy. The Journal of business intelligence. Information Systems Management,
Strategic Information Systems, 23, 11–28. 23, 32–40.
Holsapple, C., Lee-Post, A. & Pakath, R. (2014) A unified Luftman, J. & Zadeh, H. (2011) Key information technology
foundation for business analytics. Decision Support Sys- and management issues 2010–11: an international
tems, 64, 130–141. study. Journal of Information Technology, 26, 193–204.
Hong, J.F.L., Easterby-Smith, M. & Snell, R.S. (2006) Luhn, H.P. (1958) A business intelligence system. IBM
Transferring organizational learning systems to Journal, 2, 314–324.
Japanese subsidiaries in China. Journal of Management Marabelli, M. & Newell, S.M. (2012) Knowledge risks in
Studies, 43, 1027–1058. organizational networks: the practice perspective.
Hu, D., Zhao, L., Hua, Z. & Wong, M.C.S. (2012) Network- The Journal of Strategic Information Systems, 21,
based modelling and analysis of systemic risk in banking 18–30.
systems. MIS Quarterly, 36, 1269–1291. March, J.G. (1994) Primer on decision making: How deci-
Huang, J., Newell, S., Huang, J. & Pan, S.-L. (2014) Site- sions happen. Free Press, New York.
shifting as the source of ambidexterity: empirical insights McAfee, A. & Brynjolfsson, E. (2012) Big data. The
from the field of ticketing. The Journal of Strategic Infor- management revolution. Harvard Business Review, 90,
mation Systems, 23, 29–44. 61–67.
Jermol, M., Lavrac, N. & Urbancic, T. (2003) Managing Mingers, J.C. (1995) Information and meaning: foundations
business intelligence in a virtual enterprise: a case study for an intersubjective account. Information Systems
and knowledge management lessons learned. Journal Journal, 5, 285–306.
of Intelligent & Fuzzy Systems, 14, 121–136. Negash, S. (2004) Business intelligence. Communications
Imhoff, C. & White, C. (2010) Business intelligence and col- of the Association for Information Systems, 13, 54.
laboration: a natural marriage. Business Intelligence Newell, S., Robertson, M., Scarbrough, H. & Swan, J. (2002)
Journal, 15, 44–49. Managing Knowledge Work. Palgrave, Basingstoke.
Inmon, B. (1992) Building the Data Warehouse. Wiley, Newell, S., Robertson, M., Scarbrough, H. & Swan, J.
Chichester. (2009) Managing Knowledge Work and Innovation.
Kolb, D.A. (1984) Experiential Learning Experience as a Palgrave Macmillan, Basingstoke.
Source of Learning and Development. Prentice Hall, Newell, S. (2014) Managing knowledge and managing
Englewood Cliffs, NJ. knowledge work: what we know and what the future
Kolb, A.Y. & Kolb, D.A. (2005) Learning styles and learning holds. Journal of Information Technology, advance
spaces: enhancing experiential learning in higher educa- online publication, June 24, 2014; DOI: 10.1057/
tion. Academy of Management Learning & Education, 4, jit.2014.12.
193–212. Nonaka, I. (1991) The knowledge creating company. Har-
Kuhn, T. & Jackson, M.H. (2008) Accomplishing knowledge vard Business Review, 69, 96–104.
a framework for investigating knowing in organizations. Nonaka, I. (1994) A dynamic theory of organizational
Management Communication Quarterly, 21, 454–485. knowledge creation. Organization Science, 5, 14–37.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 364 A Shollo & R D Galliers

Nunamaker, Jr., J.F. (1989) Group decision support sys- Schultze, U. (2000) A confessional account of an
tems (GDSS): present and future. Proceedings: Hawaii ethnography about knowledge work. MIS Quarterly,
International Conference on Systems Sciences, Vol II, 24, 3–41.
Kailua-Kona, Hawaii, 3-6 January, 6-16. Sharma, R., Mithas, S. & Kankanhalli, A. (2014)
Orlikowski, W.J. (2002) Knowing in practice: enacting a col- Transforming decision-making processes: a research
lective capability in distributed organizing. Organization agenda for understanding the impact of business analyt-
Science, 13, 249–273. ics on organisations. European Journal of Information
Orlikowski, W.J. (2006) Material knowing: the scaffolding of Systems, 23, 433–441.
human knowledgeability. European Journal of Informa- Shollo A., Kautz K., (2010). Towards an understanding of busi-
tion Systems, 15, 460–466. ness intelligence. Proceedings: Australasian Conference
Park, S., Huh, S., Oh, W. & Han, S.P. (2012) A social on Information Systems, Brisbane, Queensland.
network-based inference model for validating customer Skyrius, R., Kazakeviciene, G. & Bujauskas, V. (2013)
profile data. MIS Quarterly, 36, 1217–1238. From management information systems to business in-
Peppard, J., Galliers, R.D. & Thorogood, A. (2014) Informa- telligence: the development of management information
tion systems strategy as practice: micro strategy and needs. International Journal of Artificial Intelligence and
strategizing for IS. The Journal of Strategic Information Interactive Multimedia, 2, 31–37.
Systems, 23, 1–10. Somogyi, E.K. & Galliers, R.D. (1987) Towards Strategic
Pervan, G.P. (1998) A review of research in group support Information Systems. Abacus Press, Cambridge, MA.
systems: leaders, approaches and directions. Decision Steiger, D.M. (2010) Decision support as knowledge cre-
Support Systems, 23, 149–159. ation: a business intelligence design theory. Interna-
Pfeffer, J. (1981) Power in Organizations. Pitman, Marsh- tional Journal of Business Intelligence Research, 1,
field, MA. 29–47.
Polányi, M. (1966) The Tacit Dimension. Anchor Day Stoodley, N. (2012) Democratic analytics: a campaign to
Books, New York. bring business intelligence to the people. Business Intel-
Polányi, M. (1958) Personal Knowledge. Towards a Post- ligence Journal, 17, 7–12.
Critical Epistemology. Routledge & Kegan Paul, Strauss, A. & Corbin, J. (2008) Basics of Qualitative Re-
London. search: Grounded Theory Procedures and Techniques,
Popovic, A., Hackney, R., Coelho, P.S. & Jaklič, J. (2012) 3rd edn. Sage Publications, Newbury Park, CA.
Towards business intelligence systems success: effects Suddaby, R. (2006) From the editors: what grounded
of maturity and culture on analytical decision making. theory is not. Academy of Management Journal, 49,
Decision Support Systems, 54, 729–739. 633–642.
Popovic, A., Hackney, R., Coelho, P.S. & Jaklic, J. (2014) Swan, J, Scarbrough, H. Preston, J. (1999). Knowledge
How information-sharing values influence the use of in- management: The next fad to forget people? Proceed-
formation systems: an investigation in the business intel- ings: 7th European Conference on Information Systems,
ligence systems context. The Journal of Strategic Copenhagen, Denmark, 2, 668- 678.
Information Systems, 23, 270–283. Truxillo, C., Lucas, B., Patteta, M., Richardson, K., Wells,
Petrini, M. & Pozzebon, M. (2009) Managing sustainability C. (2012), Advanced Business Analytics Course Notes,
with the support of business intelligence: integrating socio- SAS Institute Inc. Cary, NC, USA, Vol. 1.
environmental indicators and organisational context. The Tsoukas, H. (2009) A dialogical approach to the creation of
Journal of Strategic Information Systems, 18, 178–191. new knowledge in organizations. Organization Science,
Power, D.J. (2007). A brief history of decision support 20, 941–957.
systems. DSS Resources. COM, World Wide Web, http:// VanDiver, B., Charen, S. & Algara, R. (2009) Know thy cus-
DSSResources.COM/history/dsshistory.html, version tomer: fostering measurable impact on customer rela-
4.0. tionships with enterprise BI. Business Intelligence
Rommetveit, R. (1980) On meanings of acts and what is Journal, 14, 43–51.
meant by what is said in a pluralistic social world, In: Vendelo M., Rerup C., (2011). Sensegiving, materiality and
The Structure of Action, Brenner, M. (ed), pp. 108–149. crowd safety after the Pearl Jam concert accident,
Blackwell & Mott, Oxford. Proceedings: The Academy of Management Annual
Sahoo, N., Singh, V.P. & Mukhopadhyay, T. (2012) A Meeting, San Antonio, TX.
hidden Markov model for collaborative filtering. MIS Walsham, G. (1993) Interpreting Information Systems in
Quarterly, 36, 1329–1356. Organizations. Wiley, Chichester, UK.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 365

Watson, H.J., Goodhue, D.L. & Wixom, B.H. (2002) The Professor of Information Systems in the School of Busi-
benefits of data warehousing: why some organizations ness and Economics at Loughborough University. Previ-
realize exceptional payoffs. Information & Management, ously, he served as Professor and Research Director in
39, 491–502. the Department of IS at the LSE, Lucas Professor of
Weick, K. (1995) Sensemaking in Organizations. Sage, Business Management Systems and Dean of Warwick
London. Business School, both in the UK, and earlier as Foundation
Wengraf, T. (2001). Qualitative Research Interviewing: Biographic Professor and Head of the School of IS at Curtin University
Narrative and Semi-Structured Methods. Sage, London. in Australia. He received the AIS LEO Award for excep-
Wixom, B. & Watson, H. (2010) The BI-based organisation. tional lifetime achievement in IS in 2012 and was awarded
International Journal of Business Intelligence Research, an Honorary Doctor of Science degree by Turku University,
1, 13–28. Finland, in 1995. He is a Fellow of The Royal Society of
Whittington, R. (2006) Completing the practice turn in strat- Arts, the British Computer Society and the AIS. He has
egy research. Organization Studies, 27, 613–634. been keynote speaker at more than 60 major international
Whittington, R. (2014) Information systems strategy and conferences and symposia including the Australasian,
strategy-as-practice: a joint agenda. The Journal of Stra- European, Mediterranean, Scandinavian and UKAIS Con-
tegic Information Systems, 23, 87–91. ferences on IS. He has over 300 publications on his name,
including 90 journal articles and 12 books – the most re-
cent of which being Critical Perspectives on Business and
Biographies
Management: Management Information Systems (Taylor
Arisa Shollo has a PhD from the Department of IT Man- & Francis, 2015) and The Oxford Handbook of Manage-
agement at Copenhagen Business School and is currently ment Information Systems (Oxford University Press,
an Assistant Professor at the Department of Operations 2011). His work has been cited over 7500 times according
Management at Copenhagen Business School. Her main to Google Scholar and is trans-disciplinary in nature, focus-
research interests combine Organisational decision mak- ing primarily on organizational innovation/transformation,
ing, Management of IT and Desision support. the processes and practices of IS strategizing, knowing in
Robert D. Galliers became Bentley University’s inau- organizations and the intra-organizational and extra-orga-
gural University Distinguished Professor in 2009, having nizational impacts of ICT. He is the editor-in-chief of The
served as Provost since 2002, heading up Bentley’s prog- Journal of Strategic Information Systems – one of eight
ress to university status and EQUIS and EDAMBA accred- journals in the AIS Senior Scholars’ ‘basket’ of eight lead-
itations. He also holds a fractional appointment as ing IS journals.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 366 A Shollo & R D Galliers

APPENDIX A

Below we show the data-structure that was created after several iterations of analysing the data
from first order codes to final dimensions. It should be noted that here, we provide only few ex-
amples of raw data but in reality there are plenty of examples for each second order theme. The
themes consist of theoretically distinctive concepts that emerged from the data when analysed
in a more abstract level.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367

 The role of BI systems in organisational knowing 367

APPENDIX B

The following figure depicts the process we followed during data analysis.

© 2015 Wiley Publishing Ltd, Information Systems Journal 26, 339–367