literature review of organizational consequences for managing performance
Beau Zyad*
Master thesis Controlling – Faculty of Economics and Business, University of Groningen, Groningen, Netherlands
Abstract
While business analytics (BA) emerges across various research areas, the insights remain for the most part fragmented within different research domains. Through conducting a systematic literature review, this study examines the insights regarding organizational consequences for managing performance described in BA research. With the majority of BA studies considering the element of ‘organizational performance’, a performance management (PM)-centric review approach helped to clarify current knowledge gaps. The results suggest that a more comprehensive understanding of specific BA resources is needed to better understand BA’s role in managing performance. Additionally, current research has overlooked the interplay between BA resources. The results also indicate that there is insufficient empirical evidence to understand how organizations use operational insights obtained through BA for strategic purposes. Finally, the review highlights the importance of managements’ role in adopting BA. These contributions aim to guide future researchers into overcoming the literature’s fragmentation and ultimately establish a unified BA research agenda.
Key words: business analytics; performance management; knowledge gaps; organizational consequences; systematic literature review
Student number: s3524574 Supervisor: dr. Andrea Bellisario Co-assessor: dr. Kristina Linke
Submission date: 24th of August, 2020
Word count (excl. references and appendix): 10.704 *Contact: b.j.zyad@student.rug.nl; beauzyad@gmail.com
1. Introduction
Over the past decade, the field of business analytics (BA) has increasingly become a point of focus in the efforts to streamline organizational performance (Schläfke, Silvi, & Möller, 2013b). Research in business analytics (BA) has progressed from describing statistical and econometric methods of analyzing business data (Cosic, Shanks, & Maynard, 2015; Raffoni, Visani, Bartolini, & Silvi, 2018; Zhang, Yang, & Appelbaum, 2015) towards informing and supporting organizational decision making (Wang, Kung, & Byrd, 2018). A recent text analysis of practitioner and academic BA literature (Power, Heavin, McDermott, & Daly, 2018) asserts that BA is best referred to as, ‘a systematic thinking process that applies qualitive, quantitative, and statistical computational tools and methods to analyze data, gain insights, inform, and support decision making’. This operational definition serves to highlight that BA is a process that applies various analytical tools and methods to analyze available data and ultimately inform and support organizational decision making. As such, by influencing the way how decision are being made, BA inherently impacts many organizational disciplines. Consequently, BA has emerged across various research areas (e.g. Abbasi, Sarker, & Chiang, 2016; McAfee, Brynjolfsson, Davenport, Patil, & Barton, 2012; Mikalef, Pappas, Krogstie, & Giannakos, 2018; Mikalef, Pappas, Krogstie, & Pavlou, 2020; Nudurupati, Tebboune, & Hardman, 2016) and became a major topic of interest.
Since the year 2010, the Information Management (IM), Operations and Technology Management (OTM) and Operations Research and Management Science (ORMS) research domains have been leading the academic discussion of BA (Mikalef et al., 2020; Power et al., 2018) 1. However, although
knowledge of BA has advanced over the years, research remains for the most part fragmented within these research domains (Mikalef et al., 2020). Moreover, the growing overlap between them has not yet been systematically examined.
The IM research domain has focused mainly on how and to what extent analytical tools and methods can be leveraged to obtain competitive advantage and improve organizational performance (Bharadwaj, El Sawy, Pavlou, & Venkatraman, 2013; Henderson & Venkatraman, 1999). To broaden the understanding of how the analytical insights are actually used by organizations to achieve the proclaimed competitive advantage, the IM research domain has only recently developed a research agenda to understand the organizational consequences of BA (Mikalef et al., 2018; Mikalef et al., 2020). Likewise, within the OTM and ORMS domains, research has focused on BA’s potential in
1 Mikalef et al. (2020, p. 3 & 4) point out that ‘’information systems research’’ and ‘’management studies’’ have dominated
the academic interest of BA. Similarly, Power et al. (2018, p. 50) highlight ‘’operations research’’ as the main contributing domain to BA within management studies. However, the exact journal categorizations are not specified in these studies. Therefore, to provide a verifiable demarcation of the research domains and corresponding journals, this study adopts the research-field specifications recommended by the latest Academic Journal Guide (AJG) (2018). The AJG’s research-field specifications are based on a combination of peer review-, editorial- and expert judgement.
informing and supporting organizational decision making. As such, BA has been conceptualized in performance management (PM)2 (Raffoni et al., 2018) to better understand the role of BA within
managing organizational performance. Similar to the IM research agenda, several PM studies suggested the organizational consequences of BA as a future research area (e.g. Nudurupati et al., 2016; Raffoni et al., 2018). Thus, as these research domains and respective literature streams converge in the direction of BA induced organizational consequences, the need to understand how BA precipitates organizational resources and processes is growing more urgent.
Within the fast-expanding PM literature there has emerged a growing consensus that BA is of central importance in guiding and informing PM practices (e.g. Warren Jr, Moffitt, & Byrnes, 2015). In this overarching framework of thought, BA has demonstrated considerable potential in improving strategy formulation (Raffoni et al., 2018; Visani, 2017), validating causal relationships between performance variables (Silvestro, 2016) and enabling (Raffoni et al., 2018) a more holistic approach to PM (cf. Bititci et al., 2011). But as reports of information overload and the inability to effectively identify useful data (Raffoni et al., 2018; Visani, 2017) by managers (Vidgen, Shaw, & Grant, 2017) demonstrate, adopting the strategic potential of BA does not necessarily translate into successful BA implementations in enterprise resource planning (ERP) platforms (Ranjan, Jha, & Pal, 2017; Sun, Strang, & Firmin, 2017) and PM systems (Raffoni et al., 2018; Schläfke, Silvi, & Möller, 2013a). Moreover, although the insights into what types of BA applications can inform organizations about what happened (descriptive analytics), what will happen (predictive analytics) and what should happen in the future (prescriptive analytics) have progressed (Holsapple, Lee-Post, & Pakath, 2014), organizations still struggle to adopt BA applications in the appropriate contexts (Appelbaum, Kogan, & Vasarhelyi, 2017; Schläfke et al., 2013a; Unit, 2013).
To deepen the understanding of how organizations should deal with abundant data and hence effectively leverage BA to gain a competitive advantage, a substantial stream of IM literature focused on the technical aspects of BA, such as data collection, storage and analytical tooling (Chugh & Grandhi, 2013; Constantiou & Kallinikos, 2015; Zhang et al., 2015). With respect to the issue of organizational consequences, this work showed that these technical aspects play an enabling role within organizational decision making (C. P. Chen & Zhang, 2014; Rouhani, Ashrafi, Ravasan, & Afshari, 2016). To further understand this enabling role and the practical application of BA within PM, Raffoni et al.’s (2018) conceptual ‘business performance analytics (BPA)’ framework has become particularly
2 Several definitions of performance management are represented in the OTM and ORMS research domains.
Commonly used terms are ‘performance management’ and ‘performance measurement’ or a combination of both, i.e. ‘performance management and measurement’. In this study, the use of the terms ‘managing performance’, ‘performance management’ or ‘PM’ refers to the most recent holistic conceptualization provided by Ferreira and Otley (2009).
important. BPA refers to the management and control of organizational performance via the use of data analytics (Schläfke et al., 2013a; Silvi, Bartolini, Raffoni, & Visani, 2012) and this work has yielded insights into the aspects that may directly inform and support PM systems, such as formulating strategic questions and the identification of relevant data and analytical methods applicable to the diagnostic- and interactive control dimensions of PM (Raffoni et al., 2018; Visani, 2017).
Despite the above contributions to the field of BA, none of the above research domains has provided a comprehensive overview of BA induced organizational consequences. As a result, the understanding of how BA influences organizational performance is incomplete and two major issues remain.
Firstly, as the different research domains cumulate fragmented insights into various aspects of BA, it remains unclear whether there is actually a knowledge gap about how and to what extent organizations use BA or whether it concerns a lack of empirical evidence. Consistent with Bharadwaj et al.’s (2013) argument that ‘a more refined and detailed understanding of digital resources and their role in impacting value creation,’ would be necessary to advance knowledge about the organizational transformations that follow from BA implementations, past reviews identified a wide variety of overlapping terminology used to address BA (Mikalef et al., 2018; Power et al., 2018) and proposed a unified research agenda. Although responding to this challenge is made difficult by the extensively fragmented literature, it is crucial when seeking to further develop the understanding of BA.
Secondly, although the PM literature regarding BPA sheds light on how BA influences the process of managing organizational performance, the existing work remains largely silent about how organizations actually use BA to manage performance. Several IM studies (Ashrafi, Ravasan, Trkman, & Afshari, 2019; Wamba et al., 2017) implicitly address this issue by, for example showing that BA can provide a competitive advantage through improved information quality (and hence influence strategic decision making), but they stop short in identifying other possible ways in which BA can increase organizational performance. Hence, the mechanisms through which BA can increase performance are not explicitly examined. As a result, it remains unknown how adopting BA affects the allocation of organizational resources and how it changes organizational processes.
These major considerations led to the review of the documented evidence of organizational consequences for managing performance induced by adopting BA practices. To clarify the above knowledge gaps, the literature is critically evaluated against a holistic PM framework. The reason for this is because the majority of BA research is grounded in established theories that consider the element of ‘organizational performance’. Therefore, a framework dedicated specifically to PM allows for a thorough investigation of relevant BA research. Moreover, considering that the focus on decision making in PM research is consistent with Power et al.’s (2018) definitional highlight of BA’s decision
support function, a PM centric review approach allows for a synthesis of BA research that represents the operational definition of BA. This, in turn, contributes to formulating more specific future research areas within BA research. Thus, through systematic literature review this research will provide an answer to the following research question: ‘’What are the organizational consequences for managing performance described in business analytics research?’’.
The remainder of this paper is structured as follows. The next section describes the research methodology and outlines the steps to extract the literature. Next, in the third section, the analytical lens is explained through which the literature is interpreted and reviewed. In the fourth section the findings are presented. Finally, in the fifth section, the findings are discussed with specific attention to clarifying the knowledge gaps and future research areas.
2. Methods
Based on the premises that ‘a synthesis of evidence from multiple studies is better than evidence from a single study’ (Briner, Denyer, & Rousseau, 2009), systematic reviews of evidence are considered as a fundamental tool of advancing knowledge in a field. Following the established methodology of a systematic literature review (Thomé, Scavarda, & Scavarda, 2016; Tranfield, Denyer, & Smart, 2003), this review was conducted in distinct stages. A review protocol (cf. Moher, Liberati, Tetzlaff, & Altman, 2009) was developed to structure the review’s workflow and ensure replicability by providing a transparent audit trail of the reviewers research steps (Tranfield et al., 2003). Note that the steps from the protocol were iteratively refined during the course of the study to ensure the quality and reliability of the data extraction and analysis. The review protocol as implemented in this research is shown in Figure 1 and described below.
2.1 Identification and screening of records
To perform a literature search that corresponds to the core concepts of the research question presented in this review, a set of keywords was developed to identify the relevant studies. The selection of keywords for the concept of BA reflected the contemporary view of BA (Davenport, 2006; Power et al., 2018). Because the terminology of BA is mostly similar across different research domains, a general search was sufficient to identify BA related studies. To ensure inclusion of BA’s decision making function and the research’s shared element of organizational performance, several combined inquiries were made. The term ‘decision making’ reflected Power et al.’s (2018) definition of BA and ensured inclusion of BA studies relevant to the decision support function. The terms ‘organizational impact’ and ‘performance’ reflected the common terminology used in BA research to address
organizational performance and ensured inclusion of insights relevant to the concept of managing performance.
The concept of organizational consequences, however, presented particular difficulty as there is no agreed-upon definition in the literature. Therefore, to capture the full range of BA induced organizational consequences, Gupta and George’s (2016) IT business value framework was used to drive the choice of appropriate keywords. Grounded in the resource-based view of the firm (RBV), which is consistent with most IM studies of BA (Akter, Wamba, Gunasekaran, Dubey, & Childe, 2016; Mikalef et al., 2018; Petter, Rai, & Straub, 2012), the IT business value framework conceptualizes the organizational constructs (tangible, intangible and human skills) associated with BA. These organizational constructs are then further specified into so-called sub-dimensions of organizational resources (e.g. basic resources, technology, data, managerial skills, technical skills, data-driven culture and organizational learning). In other words, the IT business value framework provides a comprehensive set a keywords that can be used to identify the resources that organizations use when applying BA. Thus, for the purpose of this review, Gupta and George’s (2016) framework helped define the concept of ‘organizational consequences’ in a way that allowed a general as well as a specific search for relevant organizational consequences in BA research. The IT business value framework is shown in Figure 2.
Figure 2: The IT Business Value Framework used to define organizational consequences (adapted from Gupta and George (2016)).
All searches were performed in the collective database of Business Source Complete (EBSCO) as this provided the greatest coverage in all disciplines of business studies and the largest number of full text articles. To ensure that the results were similar in other major collective databases, a search in the database of Web of Science (WoS) was performed as a secondary check.
Several inquiries were made using combinations of the terms ‘business analytics’ AND ‘resources’ , ‘business analytics’ AND ‘performance’, ‘business analytics’ AND ‘skills’ as well as ‘business analytics’ AND ‘decision making’. Joined up by the OR operator, these search combinations yielded 916 results and separately the searches yielded respectively 295, 423, 247 and 359 results. The individual search strings were then expanded and refined using the selected keywords listed in Table 1.
Additionally, an expert panel of academic researchers (consisting of one expert in accounting and one expert in performance management) was consulted to ensure identification of relevant research domains. Considering the review’s concept of managing performance, they pointed out that evidence from the accounting domain could not remain unrecognized. As accounting journals may not directly surface in the deployed EBSCO and WoS database searches, the expert panel advised to pay specific attention to publications from accounting journals. Therefore, to include evidence from the accounting research domain, several accounting journals were directly searched through their respective database. This search yielded 4 eligible (see section 2.2) publications that were included in the evidence base (see Table 2).
Finally, all searches were cross-checked to avoid duplicates. Altogether, the identification and screening of records yielded 782 unique records.
Table 1: The keywords employed in the systematic search
Category Keywords Rationale
Business Analytics (BA)
business analytic*; business intelligence; big data analytic*; decision making
Reflects the contemporary view of BA (e.g. Davenport, 2006; Power et al., 2018).
Organizational consequences
resource*; skill*; basic resource*; technolog*; data; managerial skill*; technical skill*; data driven; organizational learning
Captures Gupta and George’s (2016) conceptualization of organizational constructs and resources.
Managing performance
organizational impact; performance Reflects the common terminology used to address organizational performance in BA studies (Power et al., 2018).
Table 2: The extracted publications from accounting journals.
Accounting Journals Extracted publications
Management Accounting Review (Quattrone, 2016)
Accounting, Organizations and Society (Melnyk, Bititci, Platts, Tobias, & Andersen, 2014)
The British Accounting Review (Moll & Yigitbasioglu, 2019)
Accounting, Auditing and & Accountability Journal (Arnaboldi, Busco, & Cuganesan, 2017)
2.2 Eligibility and inclusion of records
Both the search in EBSCO and the search in specific accounting journals were limited to peer-reviewed academic papers written in English. Significant contributions in the field of BA can be traced back to 2006 (Davenport), 2010 (Trkman, McCormack, De Oliveira, & Ladeira), 2012 (H. Chen, Chiang, & Storey; McAfee et al.) and onwards. However, Power et al. (2018) and Mikalef et al. (2020) show that the publication frequency of BA research has gained momentum since 2012 and since then has experienced a steady growth (see Figure 3). Therefore, the timeframe for this review’s literature search was set between 01 January 2012 and 01 June 2020. This timeframe reflected the greatest proportion of BA research available at the time of the database searches. Next, InCites’ Journal Citation Report (JCR) (Clarivate Analytics) was used to assess the quality of the search results. As such, only journal-categories relevant to this review’s concepts were selected and journals that ranked in the Journal Impact Factor’s (JIF) top quartile (Q1) were included (see Appendix 1). At this stage, 176 papers were identified and their full texts were subjected to a three-part selection filter. First, studies that were not relevant to the research question (i.e. not discussing organizational consequences for managing performance in relation to BA or organizational resources associated with BA) were excluded. Second, studies that were aimed at a specific BA application were excluded if their insights would be too specific too fit in the analytical lens (see section 3) of this review. Third, as the research question aimed at identifying existing organizational consequences, studies that employed only mathematical models or simulations were excluded. Overall, 40 papers passed all stages of the protocol and formed the final
evidence base for the synthesis. An overview of the studies that formed the evidence base is presented in Appendix 2 along with the authors’ names, year and journal of publication, the employed research method, the sample size, empirical and non-empirical categorization, and the aim of the study.
In order to maintain feasibility of reviewing the evidence base, some rigor was applied during the filtering of the search results. The author apologizes for the oversight of any relevant studies that possibly followed from this rigor.
Figure 3: BA publications per year for the time period 2010-2018 (adopted from Mikalef et al. (2020)).
3. Synthesis and interpretation of the findings: analytical lens
A narrative3 PM-centric review approach guided the synthesis and interpretation of the findings. In
order to identify organizational consequences for managing performance described in BA research, Ferreira and Otley’s (2009) Performance Management Systems Framework was used as conceptual guidance for analyzing the literature (see Figure 4). The choice for this framework was based on several reasons. Firstly, unlike other frameworks as for example Raffoni et al.’s (2018) BPA framework, which provides a specific perception of how analytical methods fit within PM, Ferreira and Otley’s (2009) framework provided a comprehensive understanding of multiple elements of PM. This made sure that both broad and specific elements of managing performance were taken into consideration. Additionally, Ferreira and Otley’s (2009) framework has been more extensively validated and used in academic research compared to Raffoni et al.’s (2018) BPA framework. Secondly, the use of Ferreira and Otley’s (2009) definition of managing performance is consistent with both contemporary PM research on organizational performance and on PM oriented BA research (resp. Grafton, Lillis, & Widener, 2010; Schläfke et al., 2013a). Finally, the Ferreira and Otley (2009) framework was designed
3 In systematic literature reviews, a narrative approach is often used for topics that are studied within diverse
research disciplines and where reviewing every single article is simply not possible (Wong, Greenhalgh, Westhorp, Buckingham, & Pawson, 2013). To synthesize all potentially relevant insights that have implications on the studied phenomenon, a ‘narrative’ thematic or content analysis is a commonly used qualitative technique to identify, analyze and present the findings (Snyder, 2019).
as a research tool for describing the structure and functioning of PM. Hence, it was considered as a valid reference for interpreting organizational consequences for managing performance.
The synthesis and interpretation of the findings was as follows. The organizational consequences for managing performance were manually extracted from the evidence base by using Ferreira and Otley’s (2009) framework as conceptual guidance. During this process, the definitions of Ferreira and Otley’s (2009) specific PM elements (see Table 3) served to iteratively check the extracted findings against. This ensured that the extractions were justifiable and hence related to PM. Next, the findings from each paper were coded according to their respective organizational consequence for managing performance. The coding process consisted of briefly summarizing the extraction from the paper(s) followed by an explanation of the rationale (e.g. the underlying information presented in the source(s) and, if needed, brief conclusions to link the information to the findings) that lead to the extraction of the finding.
During the synthesis of the findings, the author was aware of several biases that could potentially influence the outcome of the findings. The decision support bias that followed from adopting the definition of BA from Power et al. (2018, p. 49) was taken into consideration. As this decision support bias stemmed from both the IM research domain and from PM literature, it was considered primarily a publication bias4. The potential effect of this publication bias on the outcome of the findings was
mitigated by deploying the aforementioned review methodology (e.g. performing a broad search to prevent collecting evidence from limited research domains). Furthermore, the qualitative nature of this systematic literature review presented the possibility of an outcome reporting bias5 to occur.
Although the narrative (Snyder, 2019) PM-centric review approach allowed for selective identification of relevant evidence, a thorough coding process was deployed to include as much relevant information from the evidence base as possible.
4 A publication bias can occur when publications of significant findings are favored above non-significant findings
(Drucker, Fleming, & Chan, 2016). In this review, the focus in prior research on the decision support function of BA was likely to result in a biased availability of publications.
5 An outcome reporting bias can occur when relevant information is selectively presented and other relevant
information is omitted (Drucker et al., 2016). However, to some extent, a narrative review approach allows for selective presentation of relevant information as long as the review strategy is transparently described and the final synthesis serves the overall purpose of the review (Snyder, 2019).
Figure 4: The conceptual framework used to guide the analysis of the literature (adapted from Ferreira and Otley (2009)). Table 3: Definitions of the PM elements (adapted from Ferreira and Otley (2009)).
Element Definition
Vision and Mission The organizational purpose and commitment to meeting stakeholder expectations.
Key Success Factors Essential elements for a successful pursuit of the organizational objectives.
Organization structure Formal definition and configuration of roles and tasks.
Strategies and Plans Generation, adaption and communications of strategies and plans throughout the organization.
Key Performance Measures Financial and non-financial measures of performance for key objectives.
Target Setting Rationale and methods for setting performance targets. Performance Evaluation Evaluation of individual, team and organizational performance. Reward System Financial and non-financial rewards (and penalties) for meeting
(failing to meet) performance targets. Information Flows, Systems, and
Networks
Feedback and feed-forward mechanisms that support performance management.
PM Systems Use Use of performance information for various control purposes. PM Systems Change Ability to change the structure and operation of controls in
response to organizational changes.
Strength and Coherence Integration between different control mechanisms and their alignment with key objectives.
Contextual factors and culture External influences and cultural factors which impact performance management.
4. Findings
4.1 Descriptive findings
The descriptive analysis of the 40 sources confirmed this review’s introductory statement (as adopted from Mikalef et al., 2020; Power et al., 2018) that the majority of BA research is published in the IM, OTM and ORMS research domains. By representing respectively 37.5%, 20% and 10% of the evidence base, together, these research domains provided the main contributions to the current state of knowledge regarding BA. The remainder of the sources came from the Accounting (ACC), General Management, Ethics, Gender and Social Responsibility (GMEGSR) and Organizational Studies (OS) research domains. Surprisingly, the GMEGSR research domain provided a substantial amount of empirical studies to the evidence base. Due to the focus on managing performance and the expert panel’s advice to pay specific attention to accounting publications, the proportion of accounting sources in the evidence base was not unexpected and perhaps slightly skewed. The exact distributions of sources by research domain and corresponding journal are presented respectively in Figure 5 & 6.
The evidence base included 25 empirical studies and 15 non-empirical studies which included conceptual papers and literature reviews. The distribution of empirical and non-empirical studies are presented per year of publication in Figure 7 and per research domain in Figure 8. Out of the 15 non-empirical studies there were 3 studies that adopted additional non-empirical methods (case-based, secondary data analysis and survey) to validate their conceptual work (see Appendix 2). The 25 empirical studies were case-based (N = 10), survey-based (N = 7), Delphi studies (N = 2) or relied on secondary data (N = 6). The empirical strength of the evidence base was underpinned by 39 cases, 2154 surveyed responses, 124 Delphi observations and 2252 empirical observations studied through secondary data analysis.
No specific trends can be observed from the distribution of empirical and non-empirical studies. However, the OTM and GMEGSR research domains provided predominantly empirical evidence whereas the included OS study was strictly non-empirical. The rest of the domains provided a more or less even distribution of empirical and non-empirical studies.
A majority (n = 13 out of N = 28) of the empirical studies (including the non-empirical studies that adopted empirical methods) drew from various6 industries for their data collection. The remainder of
the studies either drew from service industries (n = 7) or manufacturing industries (n = 8) 7.
6 ‘Various’ refers to a mix between service- and manufacturing oriented industries.
7 Industry classification was based on information presented in the studies. In case of ambiguity between
service- and manufacturing classification, the primary business activities of the mentioned sub-industry or company was decisive.
Figure 5: Distribution of sources by research domain (classification adopted from Academic Journal Guide 2018).
Figure 6: Distribution of sources by journal (with corresponding research domain abbreviation between brackets).
Figure 7: Distribution of empirical and non-empirical studies by year of publication.
0 2 4 6 8 10 12 14 16
Accounting (ACC) General Management, Ethics, Gender and Social…
Information Management (IM) Operations and Technology Management (OTM) Operations Research and Management Science (ORMS) Organization Studies (OS)
0 1 2 3 4 5 6 7 8
Accounting, Auditing and & Accountability Journal (ACC) Auditing: A Journal of Practice & Theory (ACC) California Management Review (GMEGSR) Journal of Applied Accounting Research (ACC) Journal of Knowledge Management (OS) Management Science (ORMS) MIS Quarterly (IM) The British Accounting Review (ACC) International Journal of Production Economics (OTM) Management Accounting Research (ACC) European Journal of Operational Research (ORMS) International Journal of Production Research (OTM) Production Planning & Control (OTM) Journal of Business Research (GMEGSR) Decision Support Systems (IM) Information & Management (IM)
0 1 2 3 4 5 6 7 N o n -e mp iri cal Emp iri cal N o n -e mp iri cal Emp iri cal Emp iri cal N o n -e mp iri cal Emp iri cal N o n -e mp iri cal Emp iri cal N o n -e mp iri cal Emp iri cal N o n -e mp iri cal Emp iri cal N o n -e mp iri cal 2013 2014 2015 2016 2017 2018 2019 2020
Figure 8: Distribution of empirical and non-empirical studies by research domain.
4.2 Thematic findings
The thematic findings are organized around the key organizational consequences for managing performance that were extracted from the evidence base. These key organizational consequences are related to several elements of Ferreira and Otley’s (2009) framework and are presented accordingly. To this extent, the ‘enabling mechanisms of PM systems’ and ‘the use of PM systems’ are used as the overarching themes as they each contain multiple elements of the framework that related to the extracted organizational consequences. The ‘enabling mechanisms of PM systems’ refers to elements that enable PM within organizations (i.e. Information Flow, Systems and Networks, PM Systems use, PM Systems Change, and the Overall Strength and Coherence of PM systems). The ‘use of PM systems’ refers to elements that drive the strategic and operational functioning of PM within organization (i.e. Vision and Mission, Key Success Factors, Organizational Structure, Strategies and Plans, Key Performance Measures, Target Setting, Performance Evaluation and Reward Systems).
The specific extractions, rationale and the sources of which the thematic findings are based on are presented in Table 4 & 5. Note that some sources are coded as a whole and some are coded with specific page references. The extractions and rationale of the sources that are coded as a whole relate to information discussed throughout the source. The extractions and rationale of the sources that are coded with specific page references relate to the respective (sub-)sections described in the referred pages. 0 1 2 3 4 5 6 7 8 9 N o n -e mp iri cal Emp iri cal Emp iri cal N o n -e mp iri cal Emp iri cal N o n -e mp ir ic al Emp iri cal N o n -e mp iri cal N o n -e mp iri cal Emp iri cal
4.2.1 Consequences for the enabling mechanisms of PM systems
As information is critical input for any decision making, the way how information is obtained and used through BA is also critical for enabling effective PM. Therefore, the information flow within organizations is an important element of the enabling mechanisms that influence PM systems. Regarding the information flow within organizations, the literature characterizes BA as a transformation process of information (Holsapple et al., 2014). The organizational culture, organizational resources and capabilities as well as decision making practices are generally seen as the main antecedents of this transformation process (Delen & Demirkan, 2013; Gunasekaran, Yusuf, Adeleye, & Papadopoulos, 2018; Holsapple et al., 2014). Expanding on the view of BA as a transformation process of information, Chae (2014) argues that BA is neither an IT product nor a collection of non-IT resources. Instead, BA must be seen as a configuration of both non-IT resources and IT systems. As a consequence for the enabling mechanisms of PM systems, this indicates that in order to establish effective information flows, the combination of various resources need to be considered when adopting BA in organizations.
Building on the resource based view of the firm (RBV), several studies have outlined a wide spectrum of organizational resources and capabilities associated with BA (Akter et al., 2016; Liu, Lee, & Chen, 2020). Although the literature does not provide consensus on the specific resources, the organizational constructs (tangible, intangible and human skills) described by Gupta and George (2016) are represented across many resource oriented BA studies (Akter et al., 2016; Angalakudati et al., 2014; Krishnamoorthi & Mathew, 2018; Liu et al., 2020; Mikalef et al., 2020). Generally, the resource oriented BA studies make a distinction between analytical technology assets (i.e. IT systems and applications) and business analytics capabilities (i.e. IT- and qualitative skills) (Krishnamoorthi & Mathew, 2018).
The analysis revealed that the resource oriented BA studies often examine the associated resources and capabilities in isolation of each other. In other words, while the multitude of specific resources and capabilities in relation to BA value creation and organizational performance has been examined (Akter et al., 2016; Chae, Olson, & Sheu, 2014; Mikalef et al., 2020), the interplay between the resources and the resulting effect(s) on organizational performance remains underrepresented in this review’s literature. Although Liu et al. (2020) empirically tested and confirmed a direct positive relationship between so called ‘meta-IT capabilities’ (i.e. IT capabilities that enable other IT capabilities) and firm performance, there is little explicit evidence supporting the notion of combining resources to establish effective information flows.
Although less explicitly linked to enabling effective information flows, some evidence supporting a combination of resources existed within the studies discussing the human skills resource. Kowalczyk and Buxmann (2015) address this issue by first distinguishing between analytical experts (who provide technical BA knowledge and skills) and decision makers (who provide domain specific knowledge). They find that combining the two is difficult and advocate the implementation of a multidisciplinary analytics team as this would increase decision quality. In support of the value of an analytics team, Sanders (2016) describes several successful BA cases in which decision makers are either part of an analytics team or where analytical capabilities are successfully deployed on a large scale throughout the organization. Hence, the use of an analytics team is arguably a configuration of different (human) resources that, consequently, organizations should emulate in order to effectively manage performance.
However, the literature mentions several cognitive constraints that affect BA enabled information flows. Regardless of resource configuration in the form of analytics teams, these cognitive constraints have the potential to negatively influence the decision quality. Kowalczyk and Buxmann (2015) found intuition to negatively impact the overall decision quality and imply that a more diverse information flow would mitigate this problem. Similarly, Mikalef et al. (2020) point out that the presence of a confirmation bias (i.e. developing analytical insights that confirm a preexisting viewpoint while disregarding contradicting data) is problematic and argue that further research is needed in the area of BA implementation approaches. But although Vidgen et al. (2020) highlight the need to include ethical dimensions in organizations’ analytical development, the literature mainly stops short in specifying how to establish a diverse information flow and hence effectively enable PM through BA.
Another important issue that came to light regarding the mechanisms that enable PM is the issue of strategic alignment. In order to leverage the BA resources and improve organizational performance, organizations need to align their BA activities with their regular business activities (Holsapple et al., 2014; Raffoni et al., 2018). Unfortunately, not much of the reviewed studies explicitly addressed how strategic leverage of BA resources can be established. Rather, the literature tends to be limited to showing the impact of BA on strategy alignment (Akter et al., 2016) and emphasizing its importance for achieving improved organizational performance (Bordeleau, Mosconi, & de Santa-Eulalia, 2020; Chae, Yang, Olson, & Sheu, 2014; Liu et al., 2020; Ramanathan, Philpott, Duan, & Cao, 2017). The blend in technical- and business resources and capabilities displayed in the resource oriented studies suggests that strategic alignment is required to enable PM. However, there is not enough evidence to draw conclusions on whether and how strategic alignment enables PM and hence what the implied organizational consequences are.
Table 4: Findings from section 4.2.1 coded by extractions.
Extraction Rationale Source
BA is characterized as transformation process that yields both insights and actionable decisions.
- BA enables transformation of data into information and hence facilitates better informed decision making.
- (Holsapple et al., 2014, p. 135 & 136)
Organizational culture, resources and capabilities, and decision making practices are the main antecedents of BA as transformation process.
- Cultivating an analytics culture drives evidence based problem recognition and -solving. - Complex integration of organizational resources
and capabilities is needed to transform business data into information and knowledge.
- Decision making practices both affect and are affected by BA. I.e. decisions guide the transformation process, and the acquired knowledge leads to improved decision making.
- (Holsapple et al., 2014, p. 134 & 135) - (Chae, 2014; Delen & Demirkan, 2013) - (Gunasekaran et al., 2018; Holsapple et al., 2014) Outline of organizational resources and capabilities
associated with BA.
- BA management-, technology- and talent capabilities have a significant positive impact on firm performance. These capabilities relate to Gupta and George’s (2016) organizational constructs.
- Data attributes, -sources and analytical methods reflect Gupta and George’s (2016) tangible organizational constructs. BA capabilities, resource orchestration and data governance reflect Gupta and George’s (2016) intangible constructs.
- Resource allocation is shown to benefit from BA supported decision making in operational planning activities.
- An organizations’ ability to deploy various IT- and non-IT resources and adapt them to changing market- and business environments is defined as an important organizational capability.
- Analytical tools and techniques, and IT infrastructure are defined as ‘analytics technology assets’ and relate to tangible constructs.;
Prioritization of analytics’ usage, alignment with business, analytics culture, organizational structure, skills and people management, and the extent to which evidence-based decision making is embedded in the organizational values and processes are defined as ‘analytics capabilities’ and relate to intangible and human skills constructs. - (Akter et al., 2016, pp. 120-125) - (Mikalef et al., 2020) - (Angalakudati et al., 2014) - (Liu et al., 2020, pp. 2-4) - (Krishnamoorthi & Mathew, 2018)
Isolated examination of BA resources and capabilities.; Underrepresentation of research regarding the interplay of BA resources and the resulting effect(s) on organizational performance.
- ‘…the interdependencies or restrictions that data create for analysis or interpretation are seldom talked about and even more sparsely empirically researched.’
- Both the deployed conceptual- and research model used to test several hypothesis regarding supply chain specific BA resources, assume an exclusively individual relationship between the examined resources and operational performance.
- The impact of individual BA capabilities on firm performance are examined to illuminate the differences between capabilities in terms of value creation. Examination of interplay between capabilities remains lacking.
- (Mikalef et al., 2020, p. 7) - (Chae, Olson, et al., 2014, pp. 4696-4699) - (Akter et al., 2016) Evidence supporting the notion of combining resources (i.e. deploying a multidisciplinary analytics team) to establish/enable effective information flow.
- Deploying multidisciplinary analytics teams facilitates ambidexterity in BA supported decision making processes to overcome managerial and cognitive challenges that are inherent to BA supported decision making. This ultimately increased decision quality. ‘…effective Business Intelligence and Analytics support requires collaboration between analysts and decision makers.’; ‘…decision processes that achieve higher levels of ambidexterity also realize higher decision quality.’
- A competent business-aware team and performant technical infrastructure are necessary to leverage operational BA capabilities.
- Several examples of how multinational organizations successfully deployed analytical resources. Deploying real-time supply chain data analytics across different divisions and functions. Use analytical applications that ‘talk’ to each other, thus enabling coordination of activities. Deploy advanced analytics to support operational decision making and uses analytical ‘surveillance systems’ to gauge costumer behaviors and adjust their business accordingly.
- (Kowalczyk & Buxmann, 2015) - (Bordeleau et al., 2020, p. 174 & 181) - (Sanders, 2016) Cognitive constraints negatively influence BA enabled information flows and decision quality
- Analytics teams that exhibited high levels of ambidexterity yielded higher decision quality and found to favor rationality above intuition. - Confirmation bias is inherently present in
decision making processes. Future research is needed regarding BA (implementation) approaches.
- (Kowalczyk & Buxmann, 2015) - (Mikalef et al., 2020, p. 6 & 7)
Strategic alignment enabling effective PM
- Showing impact of strategic alignment of BA- and business resources.
- Emphasize importance of BA strategic alignment for (improved) organizational performance.
- (Akter et al., 2016) - (Bordeleau et al., 2020; Chae, 2014; Liu et al., 2020; Ramanathan et al., 2017)
4.2.2 Consequences for the use PM systems
In contrast to the consequences for the enabling mechanisms, the literature’s link to the use of BA in PM system is more explicit. A recurring consequence of BA for managing performance is the fact that BA adoption changes the conventional way of strategy formulation and execution. Conventional PM imposes operational performance to result from executing strategic objectives and imply a top-down approach to establishing an organization’s PM system (Ferreira & Otley, 2009). This implied top-down approach (e.g. deriving key success factors from an organization’s vision and mission, basing the organizational structure on the key success factors and formulating the strategic objectives on the basis of the organizational structure) seems to be reversed by adopting BA (Raffoni et al., 2018). This is consistent with earlier evidence suggesting the need to recast the existing approach to PM and hence better synchronize strategic and operational measures (Melnyk et al., 2014). In other words, as BA provides organizations with more operational insights to support effective decision making, literature indicates the importance of strategic objectives that reflect the operational performance (Bordeleau et al., 2020; Raffoni et al., 2018).
However, based on the discussion within the accounting domain regarding the so called ‘digital revolution’, where both BA and PM systems are arguably part of, an important note can be placed at the implied reversed approach to PM. Some conceptual studies (e.g. Arnaboldi et al., 2017; Quattrone, 2016) argue that the digital revolution in management accounting can impossibly deliver perfectly rational decision making since politics, epistemological limitations and biases will always be intertwined with data. So although BA literature implies a reversed approach to PM, it remains uncertain whether and to what extent it is desirable to change the conventional approach to establishing PM systems.
Nevertheless, a recent study sheds light on the extent to which BA can add value and how this can be achieved. Bordeleau et al. (2020) refer to organizational learning as a way for organizations to use operational insights gained through BA. They argue that organizations that balance their organizational learning between exploiting the conventional way of doing business with exploring new ways are more likely to improve financial performance. Hence, what is dubbed as organizational learning
ambidexterity, this balance would help leverage the strategic business value of BA. Translated to a PM system, this means that by adopting BA in PM, a balance should be found between exploiting the conventional- and exploring new ways of managing performance. Given that, conventionally, strategic objectives are mainly driven by the key success factors (Ferreira & Otley, 2009), one of the organizational consequences for managing performance is that the formulation and execution of the strategic objectives should additionally be driven by operational insights obtained through BA.
In their supporting case study, Raffoni et al. (2018) identify resistance coming from managerial functions, such as project managers and controllers, towards the implications of adopting BA in their work. Heinzelmann (2018) reveals that along with extensive use of BA, more routine work and reporting (‘’dirty work’’) is assigned to the role of management accountants and controllers. As this dirty work would lessen their professional identity (Heinzelmann, 2018), this could explain the resistance to exploring new ways of managing performance. Another possible explanation for the resistance highlighted by the literature is the fact that management frameworks address BA-related strategic management issues predominantly on a highly abstract level (Pröllochs & Feuerriegel, 2020). While advanced analytics are frequently used for steering operational decision making, insights regarding individual business units, activities and processes are rarely used to define strategic management. Unfortunately, the strategic use of operational insights derived from BA appears to be underrepresented in current research (Pröllochs & Feuerriegel, 2020; Raffoni et al., 2018). Therefore, it remains difficult to fully understand the consequences of BA’s strategic value for managing performance.
However, the discussion displayed in IM literature about the managerial roles and behaviors regarding BA implementation does shed light on the functioning of the organizational structure defined in PM systems. Several studies use the term ‘IT reinvention’ to explain how BA applications can become a strategic resource (Nevo, Nevo, & Pinsonneault, 2016). By personalizing BA, users can not only better use the obtained insights for achieving their goals, but they also make the BA application more difficult to imitate by competitors thus creating a strategically valuable resource (Nevo et al., 2016). Extending this thinking, Wang and Hajli (2017) conceptualize the path-to-value chain of BA and map out the path through which BA leads to business value. Interestingly, their analysis shows that BA applications more often lead to operational benefits than to managerial or strategic benefits. Their definitions of these benefits reveal that, although BA results in improved operational performance, it does not often result in more efficient and effective strategic managerial decision making. Hence, the actual strategic use of BA as a competitive differentiator is held back by how managers use the insights (Wang & Byrd, 2017).
Supported by other studies testing the effect of BA on organizational performance (Song, Zheng, Zhang, & Yu, 2018; Wamba et al., 2017), the importance of BA skills among managers becomes eminent. Wamba et al. (2017) show that, what others call absorptive capacity (Wang & Byrd, 2017), the ability to change business processes according to obtained BA insights is positively associated with organizational performance. Similarly, management capabilities and personnel experience (Torres, Sidorova, & Jones, 2018) and data usage (Song et al., 2018) are shown to positively affect organizational performance.
So while it remains difficult to fully understand the organizational consequences of BA’s strategic value, the managerial skills to adapt the organizational structure and strategic objectives to the obtained operational insights seems to play a key role in adopting BA in a PM system.
Regarding the operational aspect of adopting BA in PM, the literature revealed various BA resources that are of interest for managers seeking to leverage the operational insights obtained through BA. Besides the aforementioned personalization (Nevo et al., 2016), the extent to which decisions are being based on BA, the quality of available data (Chae, Yang, et al., 2014; Pape, 2016) as well as the linkages between different data sources and degree of information sharing within organizations (Gunasekaran et al., 2017) appear to impact the use of BA within organizations through improving the operational performance. Furthermore, Gunasekaran et al. (2017) conceptualize a three-stage process (acceptance, routinization and assimilation) through which these BA resources are adopted and stipulate the mediating effect of management commitment. For managing performance, this means that BA applications in PM systems are subject to the a number of requirements and undergo a specific adoption process before they can deliver operational insights.
Table 5: Findings from section 4.2.2 coded by extractions.
Extraction Rationale Source
Adopting BA changes the conventional way of strategy formulation and execution
- An analytical breakdown of (operational) performance measures is explicitly put next to (strategic) macro factors as a starting point for assessing the business strategy and performance model.; Importance of linking analytical methods to organizational objectives is emphasized.
- There is a need to recast the relationship between strategy and PM.
- ‘Industry 4.0’ can benefit from BA by exploiting operational insights to improve process- and firm performance.; BA benefits are more explicit and easier to measure at process level than at firm level.
- (Raffoni et al., 2018, p. 62) - (Melnyk et al., 2014) - (Bordeleau et al., 2020, p. 174 & 175) Digital revolution in management accounting and decision making.
- Reasonable, and not rational choices are what organizations seek through developing data driven decision making.
- (Arnaboldi et al., 2017; Quattrone, 2016)
- ‘…decision making does not deal only with ‘data’ to be used neutrally in decision-making.’ Instead, other factors such as politics, pressure, egos and other factors inevitably move decision making away from rationality.
- As a characteristics of technology enabled systems and networks, governance of data as well as the alteration of information and decision making processes are highlighted as important future research areas. - (Quattrone, 2016, p. 120 & 121) - (Arnaboldi et al., 2017) Organizational learning ambidexterity as a way to use operational insights obtained through BA.
- ‘Ambidexterity requires reaching an equilibrium between operational excellence and exploitation in one hand, and innovation and exploration in the other.’; A link is shown between organizational learning ambidexterity and BA capabilities. However, the strength of this link remains undecided and needs future research. - (Bordeleau et al., 2020, p. 183) Managerial resistance towards adopting BA and possible explanations.
- Project managers resisted to the idea of statistically analyzing and reporting on new operational performance information that was embedded in the available data. Specific resistance towards the implied increased visibility of individual outcomes and potential redistribution of responsibilities; Controllers resisted to analytically investigating the determinants of revenues (as pointed out by the adopted BA approach) instead of cost drivers. Resistance was based on controllers’ reliance on previous experience and limited perception of own responsibility (e.g. analyzing revenues was attributed to marketing functions). - (Raffoni et al., 2018, pp. 60-63) Possible explanations for managerial resistance: Dirty work lessening management accountants’ professional identity.; Abstract elaboration of strategic issues.
- IT system changes incline more ‘dirty-work’ (e.g. routine work) to management accountants. Dirty work evokes changes in their professional identity. Facing IT-inclined changes, management accountants prioritize their existing professional identity above the (implications of) IT changes.
- Despite operational performance measures being input for strategy and performance assessment, it seems as it does not directly translate to changes in data collection. Instead, only strategic considerations are included in the first three stages of the BPA framework. Hence, a strategic focus is present in the conceptualization of BA within PM systems.
- Strategic management issues are shown to benefit from adopting advanced analytics in the management frameworks. However, operational elaborations of using BA in strategic management remain underrepresented. Particularly, with regards to leveraging company data to create business value. ‘Management frameworks predominantly address overall strengths and weaknesses on a highly abstract level.’; ‘The use of firm related data appears
- (Heinzelmann, 2018) - (Raffoni et al., 2018, p. 62) - (Pröllochs & Feuerriegel, 2020, p. 10)
underrepresented in current research on big data analytics…’. Managerial roles and behaviors regarding BA implementation.
- IT can become a strategic resource through personalized ‘reinvention’ of the technological capabilities. Hence, managers need to personalize BA applications to make it more valuable to them. ‘By reinventing IT, users are making it more unique, less common, and more difficult to imitate by competitors.’
- Path-to-value chain of BA (simplified): BA → BA components → BA capabilities → Benefit dimensions → Business value.;
Out of 238 causal-chain connections between BA capabilities and benefit dimensions, 37 were coded as managerial benefits and only 7 as strategic benefits compared to 128 operational benefits. This shows that BA leads to improved operational performance, but strategic implementation remains lacking. - BA is shown to improve an organizations’ absorptive
capacity.; Effective use of BA tools indirectly influences decision making effectiveness.; Absorptive capacity mediates the relationship between effective use of BA and decision making effectiveness. Thus, the capability to effectively use BA tools as well as an organizations’ absorptive capacity (e.g. the ability to change business processes according to obtained insights) are crucial when implementing BA.
- BA technological infrastructure, BA management capabilities and BA personnel expertise are (significantly) positively associated with firm performance. Note that management capabilities offsets the significance of personnel expertise suggesting that management capabilities are more relevant for leveraging BA than personnel expertise. - ‘...both demand-side and supply-side data analytics
usage has a positive effect on the performance…’
- (Nevo et al., 2016, p. 180 & 181)
- (Wang & Hajli, 2017, pp. 293-296)
- (Torres et al., 2018; Wamba et al., 2017; Wang & Byrd, 2017, pp. 524-526, 532 & 533) - (Torres et al., 2018, p. 830 & 831) - (Song et al., 2018) BA resources impacting the leverage of operational insights. - Personalization of BA applications.
- The extent to which advanced fact-based analytics are being used in operational decision making and data accuracy, impact the operational performance.; The extent to which analytics are being used also mediates the impact that data accuracy has on operational performance. This suggests a certain degree of complementarity between BA resources. This suggests that the mediating role of complementary BA resources could be an alternative explanation to BA’s value creation.
- Data connectivity and degree of information sharing positively impacts supply chain performance and organizational performance.; Top management commitment plays a significant mediating role in BA adoption. Managers must not only acquire BA resources, but also commit to orchestrating and investing on resource bundling.
- (Nevo et al., 2016) - (Chae, Yang, et al., 2014, pp. 121-124) - (Gunasekaran et al., 2017, p. 314)
5. Discussion
5.1 Contributions to BA research
By reviewing the current state of knowledge regarding BA induced organizational consequences for managing performance, this systematic literature review uncovered prominent areas of research that are of high relevance for developing the field of BA. Through focusing on organizational consequences for managing performance, the understanding of how BA influences organizational performance improved in several ways. Firstly, by highlighting the organizational consequences of BA for managing performance, the current implications for management research and practice are addressed more specifically. And secondly, the PM centric review approach clarified the knowledge gap regarding the use of BA in managing organizational performance. The remainder of the discussion is structured around these two main contributions.
5.2 Organizational consequences for managing performance
The findings highlight several organizational consequences for managing performance that are described in BA research. While the literature holds clear implications as to how BA impacts the use of PM systems, the understanding of how BA enables PM systems is found to be less explicit.
Although strong empirical evidence is somewhat lacking, there are roughly three organizational consequences that can be distinguished regarding the enabling mechanisms of PM systems. First, the literature points out that organizations need to combine tangible, intangible and human skills resources to effectively use information obtained through BA. Through configuring both IT systems and non-IT resources, BA can be used to establish effective information flows and hence enable PM (Chae, 2014; Holsapple et al., 2014). Secondly, to configure both IT systems and non-IT resources, organizations should emulate analytical teams comprised of both analytical experts and decision makers (Kowalczyk & Buxmann, 2015; Sanders, 2016). By combining technical BA knowledge and skills with domain specific knowledge about the business activities, BA can contribute to effective performance management. Note that the negative impact of cognitive constraints such as biases (Mikalef et al., 2020) and managerial intuition (Kowalczyk & Buxmann, 2015) on decision quality need to be taken into consideration as this can hamper the effective enabling of PM through BA. Lastly, the findings suggest that strategic alignment of technical- and business resources is required to enable PM. In other words, the information obtained through BA should be aligned with the overall organizational strategy.
In contrast to the organizational consequences for the enabling mechanism, the BA induced organizational consequences for the use of PM systems are more explicitly described in the literature. There is both a conceptual understanding of how BA precipitates managerial processes as well as
convincing empirical evidence supporting this understanding present in the literature. First and foremost, adopting BA in PM systems implies that the obtained operational insights need to be used as input for strategy formulation and execution. Consequently, this asks for a change of the conventional top-down approach towards a more bottom-up approach to establishing PM systems. The need to recast the approach to establishing PM systems is empirically supported by studies showing organizational performance to benefit from: using operational insights obtained through BA (Bordeleau et al., 2020), effective use of combined BA resources (Torres et al., 2018) and a broad usage of available data (Song et al., 2018). However, some studies (e.g. Arnaboldi et al., 2017; Quattrone, 2016) implicitly question the extent to which changing the approach to establishing PM systems is desirable by referring to fundamental limitations to decision making that prevent perfectly rational decision making. With respect to organizational consequences for the use of PM systems, a valuable takeaway from this criticism is to consider that irrational factors such as politics, pressure, egos and biases will inevitably move decision making away from rationality. Hence, completely basing strategy formulation and execution on operational insights is unlikely to provide optimal decision quality.
Another BA induced organizational consequence for the use of PM systems arises from the literature’s elaboration of how operational insights can be used as strategic input. The literature suggests that organizations that combine new ways- (e.g. BA supported decision making) with the conventional way of managing performance are more likely to improve financial performance (Bordeleau et al., 2020). Moreover, an organization’s ability to change business processes according to BA obtained operational insights is found to positively affect organizational performance (Wamba et al., 2017; Wang & Byrd, 2017). Hence, using operational insights as additional input for strategy formulation and execution is arguably desirable for improving PM. However, it is shown that, in practice, BA obtained operational insights rarely improve strategic decision making (Wang & Hajli, 2017). By examining managements’ role in implementing BA in (PM related-) decision making, the literature points out the importance of managerial commitment (Gunasekaran et al., 2017) and managements’ BA skills (Nevo et al., 2016; Raffoni et al., 2018).
Although it remains difficult to comprehensively understand the organizational consequences of BA’s strategic value, the literature highlights some specific BA resources that can improve the strategic usage of BA obtained operational insights. Personalizing BA applications by managers to better support their decision making (Nevo et al., 2016), the linkages between (Gunasekaran et al., 2017) and quality of the available data (Chae, Yang, et al., 2014; Pape, 2016) are explicitly shown to improve the use of BA within organizations. Hence, regarding the BA induced organizational consequences, this means that these specific BA resources need specific consideration when adopting BA in PM systems as they can help managers to better leverage BA’s strategic value.
5.3 The knowledge gap and future research directions
To advance the understanding of how and to what extent organizations can apply BA and hence improve organizational performance, it is necessary to overcome the literature’s fragmentation. While examining the research’s overlapping insights regarding BA induced organizational consequences for managing performance, some important focal points and oversights of current research became clear.
Overall, the reviewed studies displayed a disproportional focus on BA resources which added limited value to the understanding of BA’s role in managing organizational performance. Although Gupta and George’s (2016) organizational constructs (tangible, intangible and human skills) are represented in different studies, the lack of consensus on the specific resources remains a troublesome knowledge gap. In other words, current literature lacks an elaboration of what kind of tangible-, intangible- and human skills BA resources play a role in managing performance. The examples of specific resources, e.g. analytics teams (Kowalczyk & Buxmann, 2015), BA skills of managers (Torres et al., 2018; Wang & Byrd, 2017) and ease of information sharing (Gunasekaran et al., 2017), hold valuable implications for PM systems. However, they hardly provide a comprehensive elaboration of all BA resources involved in managing performance. Given the support for Gupta and George’s (2016) organizational constructs, using their definitions of specific organizational resources as guidance for future research can help overcome the literature’s lack of consensus.
Additionally, the findings suggest that the interplay between BA resources is an oversight that prevents empirical validation of BA’s role in information flows. While conceptual work highlights the importance of combining BA resources (Chae, 2014; Holsapple et al., 2014) for processing information (Wamba et al., 2017), empirical evidence remains scarce. And although substantial evidence emerged in support of BA’s impact on decision making processes (Liu et al., 2020; Nevo et al., 2016), it remains unknown how combinations of resources affect the information flows within these decision making processes.
To better understand BA as a transformation process of information it is therefore important to include the combination of specific resources in future research. Relevant research questions would be, for example, ‘How do combinations of BA resources affect organizational performance?’ or ‘What combinations of resources are deployed in successful BA implementations?’.
Also, the lack of evidence for BA’s role in strategic alignment of organizational resources presents a knowledge gap in the understanding of the strategic value of BA for managing performance. The upcoming research on BPA extended the knowledge of how BA can support management by quantifying strategic performance dimensions (Raffoni et al., 2018). However, the reviewed literature provided very little empirical evidence for BA’s strategic value. Particularly, the evidence regarding how and to what extent organizations use operational insights obtained through BA as input for
