A Case Study at ASSA ABLOY

(1)

Business Intelligence: Competencies and Cross-Functional Integration -

A Case Study at ASSA ABLOY

Master’s Thesis 30 credits

Programme: Master’s Programme in Accounting and Financial Management Specialisation: Management and

Control

Department of Business Studies Uppsala University

Spring Semester of 2021

Date of Submission: 2021-06-02

Jon Ariel Borgsø

Maxim Svensson

Supervisor: Jan Lindvall

(2)

Acknowledgements

Firstly, we would like to express our deep gratitude to our supervisor, Jan Lindvall, for the inspiration and support, constantly providing challenging and valuable insights. We also wish to thank our fellow students for the engagement and constructive criticism that helped drive the study forward. Lastly, we would like to pay our special regards to ASSA ABLOY and the respondents in this study. This research would not have been possible without their

informative and collaborative participation.

__________________________________ __________________________________

Jon Ariel Borgsø Maxim Svensson

(3)

Abstract

Business Intelligence (BI) and data analytics has grown to become one of the most prioritized technological investments for organizations today. For BI systems to be valuable for

organizations’ decision making and support of end-users, research argues that competencies of multiple areas need to be represented in the work with BI. This includes knowledge of both IT and business domains, where challenges such as lack of domain competencies have been identified in the Swedish industry sector. The purpose of this study is therefore to investigate the representation of BI competencies, with focus on IT, business domain, data analytics and their integration. The research is conducted through a qualitative case study at ASSA ABLOY, a leading company in the Swedish industry sector, where interviews are made with respondents involved with five BI tools from different functions of the company.

The empirical findings show that competencies of IT and business domains are represented to a higher degree than data analytics. In addition, the findings show that while integration between these areas is being promoted, there is potential for further involvement with in- house IT and a need for cross-border knowledge to bridge the gap between functions involved with BI.

Keywords: Business Intelligence, Business Domains, Competencies, Integration, Data Analytics, Roles, Functions, Assessment Framework.

(4)

List of Tables

Table 1. Summarization of the completed interviews ... 20

Table 2. Operationalization of the four competency areas ... 22

Table 3. Aggregated assessment ... 47

(6)

1

1. Introduction

___________________________________________________________________________

This chapter presents a brief background to Business Intelligence and related competencies.

It continues with a problematization of why this is considered a relevant research topic for the Swedish industry sector, followed by the research question and limitations.

___________________________________________________________________________

1.1 Background

Companies are today described as analytical competitors that aim to stay relevant by searching for tools and strategies for enhancing their automation, digitization and data analytical capabilities (Gudfinnsson, Strand & Berndtsson, 2015; Powell & Dent-Micallef, 1997). Business Intelligence (BI) was introduced as a way to manage data analytics and has become increasingly popular in academia, as well as being ranked the highest of

technological priorities among businesses (Chen, Chiang & Storey, 2012; Elbashir, Collier &

Davern, 2008).

BI contains a wide selection of applications and functions used for collecting, storing, analyzing and processing data with the purpose of improving decision making (Gürdür, El- khoury & Törngren, 2019). However, implementing BI solutions into a company does not automatically add value to its business. In fact, a reappearing challenge found in the BI research is not having a clear understanding of the competencies required for dealing with BI (Williams, 2016; Viaene, 2008). According to Debortoli, Müller and Brocke (2014), the key to successful integration is to include competencies of the multiple facets of BI, something that is argued to be critical when assembling a BI team.

Researchers further state that IT competencies, such as knowledge of system technologies, have shown to be relevant for successful BI implementation in collaboration with

competencies regarding business domains (Salmasi, Talebpour & Homayounvala, 2016;

Viaene, 2008). Employees with competencies of the business, such as operations or supply chain, can typically cope with the complexity of providing high-quality information and analysis about the domains which BI serves. Knowledge regarding plants, customers, KPIs, products or business units in the development stage is argued to increase relevance and value for the BI users (Williams, 2016).

(7)

2

1.2 Problematization

Many times, organizations consider BI tools solely as an IT initiative, putting the

responsibility of developing BI on the IT function and those with technical competencies (Williams, 2016). This is argued to increase the risk of failed integration with the rest of the business (ibid.). Volvo is one example of this. In the early 2000s, the company implemented BI systems with the ambition of developing analytics for decision support and value creation for the business (Viaene, 2008). Nonetheless, the initiative led to dissatisfaction for the end- users due to the BI team’s difficulties in aligning the tool with the rest of the business.

Arguments are drawn that organizations require more knowledge regarding the competencies, roles and functions that should be involved in the stages of BI implementation, maintenance and development. According to Williams (2016), the absence of a clear strategy regarding competencies is not rare for companies moving towards BI and digitalization. Similarly, lack of urgency related to digitalization and BI has been evident in the Swedish industry sector.

Studies have also revealed that only eight percent of the Swedish senior leaders want to invest more resources in digitalization and automation (Azet, 2018; Industrinyheter, 2018). In addition, Gürdür, El-khoury and Törgngren (2019) identified lack of management

involvement as a challenge hindering companies in the Swedish industry sector from seizing opportunities for BI.

These results might come as a surprise considering that Sweden is described as a leader in digitalization compared to other European countries (Gürdür, El-khoury & Törgngren, 2019;

Björkdahl, 2020; Larsson & Wallin, 2020). Nevertheless, the challenges identified indicate that more knowledge regarding BI competencies might be needed. There seems to be limited studies offering an assessment on BI competencies currently represented in the Swedish industry sector. Hence, more research is required for understanding what competencies are seen as important when dealing with BI in practice and how the integration of these

competencies is promoted. Collecting insights from the workplace could facilitate continuous BI developments and help practitioners, such as companies, BI providers and consultants, to better support BI initiatives.

(8)

3

1.3 Research Question

The purpose of this study is to investigate the competencies of the multiple functions involved with BI by focusing on one organization in the Swedish industry sector. The selected organization of this study is ASSA ABLOY which is further presented in Chapter 3.2.1. This organization will be examined by studying the competencies represented for managing and maintaining BI, and how the integration between the involved competencies is promoted. This has led to the following research question:

● What competencies are represented in relation to Business Intelligence and how is the integration between competencies being promoted in ASSA ABLOY?

1.4 Delimitations

The organization will be examined by studying the representation of IT, business domain and data analytics competencies for managing and maintaining BI, as well as the integration between these competencies. The focus will be on the internal perspective of the

organization, from the roles and functions involved in the BI work. Hence, for this study, the term competencies refer to the job roles and functions involved in BI rather than individual skills, education or training. Similarly, the term integration is used for focusing on the collaboration and communication between roles and functions, rather than the technical integration between BI systems. These delimitations are further elaborated in the literature review.

(9)

4

2. Literature Review

___________________________________________________________________________

In this section, a literature review is presented with focus on BI research, competencies and integration, and a summarization of the literature review with an included assessment framework.

___________________________________________________________________________

2.1 Business Intelligence 2.1.1 Development of BI

The growth of data analytics has grown to become significant for businesses over the recent decades (Davenport, 2006). Because of the increasing access to large amounts of real time information and Big Data, the world is experiencing a paradigm shift of digitalization, leading to the so-called Fourth Industrial Revolution (Chen, Chiang & Storey, 2012; Lasi et al., 2014; Schwab, 2016). Learning to use data analytics in a skillful way and retrieve useful information from data is described as the secret to increasing organizational performance (Weill & Woerner, 2015). Therefore, new solutions are required for organizations to seek opportunities for optimizing data analytics and remain competitive.

BI was introduced to offer a way of managing data and has emerged to become one of the most important decision making systems among organizations (Debortoli, Müller & Brocke, 2014; Elbashir, Collier & Davern, 2008). BI refers to a systematical and technical

infrastructure that collects, stores, visualizes and analyzes the data created by a company’s activities. Researchers describe BI as a tool for transforming data into insights, decisions and actions for increasing organizational performance (McAfee et al., 2012). The introduction of BI has shaped businesses and generated new requirements on investments and resources, such as human capital and required competencies (Oesterreich & Teuteberg, 2019).

2.1.2 Clarification of Business Intelligence system

Research suggests different definitions for BI. These definitions emphasize various approaches which categorizes BI as a product, process, system or paradigm (Talaoui,

Kohtamäki & Rajala, 2020). For this study, BI is distinguished from the general information system movement of data analytics and Big Data to the actual IT-artifacts used in

organizations (Arnott, Lizama & Song, 2017). This distinction focuses on BI as a system or a

(10)

5

tool, which can be considered common among BI research. In addition, Arnott, Lizama and Song (2017) argue that vendors, consultants and researchers usually perceive BI as an enterprise system. This also goes in line with Gartner’s (2021a) definition of BI, which emphasizes enterprise platforms and applications for describing BI. Therefore, when using the term BI forward, the focus will be on the IT artifact that is BI system, citing from

Davenports (2006, p. 106) definition as: “(...) a wide array of processes and software used to collect, analyze, and disseminate data, all in the interests of better decision making.”

Davenport’s (2006) studies are frequently referred to in the BI literature. The author’s definition emcompasses BI as a complex system with the aim of supporting management decision making. BI systems take into account both larger enterprise BI systems, generally managed by IT functions and connected to enterprise-wide data warehouses or a federation of data marts, as well as smaller functional BI tools (Arnott, Lizama & Song, 2017). The latter is often restricted to one division, department or function, where the responsibility of

maintaining the BI system also lies on the business unit, rather than on the IT department alone (ibid.).

In addition, nowadays it is essential to observe that BI systems include more modern use of BI platforms and applications (Gartner, 2021a). Since traditional BI systems, in general, inhabit numerous layers of software-defined by IT-functions, BI platforms are characterized by more easy-to-use tools. Such BI platforms commonly consist of well-defined bases, regularly provided by vendors or consultants supporting the full analytical workflow from data preparation to business insights (Howson et al., 2019). Examples of modern BI platforms are Tableau, Qlik or Microsoft Power BI.

2.1.3 Successful value creation of Business Intelligence

There seems to be conformity among researchers that BI and data analytics generate higher performance for businesses. McAfee et al. (2012) showed that organizations at the top of their industry in the use of data-driven decision making were on average 5 percent more productive and 6 percent more profitable than their competitors. Similarly, Howson et al.

(2019) evaluated the most common BI platforms available, ranking them according to their advantages and disadvantages. Despite that the platforms showed different results,

organizations that offered users any platform with access to a curated catalog of internal and

(11)

6

external data, were argued to derive more business value from analytics investments than those who do not (ibid.).

Researchers and studies offer various explanations for what are considered critical success factors when implementing BI. Examples of such factors are BI governance (Watson &

Wixom, 2007), portfolio and project management (Williams & Williams, 2004; Eckerson, 2007; Geiger, 2009), or strategic alignment and vision (Eckerson, 2007; Williams &

Williams, 2004). In addition, factors for determining the strengths of various BI systems have been proposed by analytics, which include dimensions such as agile IT-enabled workflow, decentralized self-service analytics, governed data discovery and embedded analysis

(Howson et al., 2019). Such factors are considered important to take into consideration when implementing BI, depending on the industry, organization or the business domains the BI system serves.

According to Salmasi, Talebpour and Homayounvala (2016), BI can be considered successful and value-creating when organizations obtain tangible benefits from the BI investments.

However, tangible benefits might differ among organizations and can be difficult to measure (Sabanovic & Søilen; 2012). Therefore, it is argued that BI value depends on what benefits organizations expect from BI which can be anything from improved profitability, reduced costs, to improved efficiency (Framnes et al., 2017; Grover et al., 2018). This is determined by what is considered relevant for the business domains that the BI system serves.

2.1.4 Pitfalls with Business Intelligence

Studies also indicate that implementing BI can be difficult and might lead to failure of providing the type of information that users need or not supporting organizations’ processes adequately (e.g. Viaene, 2008; Frisk & Bannister, 2017). Researchers argue that BI generates higher performance for businesses but only if it is implemented and used correctly (Brooks, El-Gayar & Sarnikar, 2015).

In fact, adaptability to the business domains and the possibilities that comes with connecting BI to its users is a recurrent topic in the BI literature (Viaene, 2008; Geiger, 2009; Eckerson, 2007; Williams & Williams, 2004). Properly understanding and managing the technologies, functionalities and business aspects related to BI are seen as important for creating value. BI

(12)

7

requires knowledge and experience regarding IT processes such as storage, data mining, process analysis, Online Analytical Processing (OLAP), but it also needs to be partnered with other business workflows for creating substantial information and contribute to actionable decision making in the business (Rouhani et al., 2016). Integration, communication and collaboration between roles and functions in organizations are therefore vital (Viaene, 2008;

Eckerson, 2007). Different areas of BI need to be represented and aligned for the

contemporary organization wanting to gain benefits from BI. Business domains and IT need to engage by: “(...) negotiating, influencing, educating, socializing and interacting in other ways across organizational levels and functional boundaries to develop greater alignment and coordination throughout the company” (Viaene, 2008, p. 33).

Unfortunately, organizations seem to fail with their BI initiatives due to a lack of integration between departments (Salmasi, Talebpour & Homayounvala, 2016). Arguably, this is because organizations require more collaboration between the various aspects of BI, often overstating the responsibility of IT (Viaene, 2008; Williams, 2016). Research offers different approaches to the facets of BI and related competencies, which is presented below.

2.2 Competencies Related to Business Intelligence 2.2.1 Definitions of BI competencies

The concept of competencies found its way into BI literature in relation to supporting BI goals and an organization’s ability to derive benefits from their investments in BI (Chasalow, 2009). Researchers adopted different approaches for studying BI competencies. One way to categorize these approaches is to divide BI competencies into individual or organizational level competencies (Salmasi, Talebpour & Homayounvala, 2016). Individual level

competencies focus specifically on knowledge, experiences and skills influencing the motivation and behaviors of individuals, argued to affect performance and outcomes in the organization (ibid.). On the other hand, organizational level competencies are defined as assets, skills or resources belonging to the company which allows organizational activities to be performed (EnEscrig-Tena & Bou-Llusar, 2005).

Taking these two approaches into consideration, the approach of organizational level competencies is arguably a better fit for this study. While not neglecting the importance of understanding individual motivations and behaviors, the aim of this research is to understand

(13)

8

the professional roles and functions involved with BI, rather than individuals. Similar arguments have been drawn in other fields that apply organizational competency as a term, such as in performance management, human resources and organizational change (Nienaber

& Sewdass, 2016). As well as the BI field, where the terms job roles and role descriptions have been used for identifying vital competencies for BI success (Miller, Bräutigam &

Gerlach, 2006). It might also be important to clarify that one role does not equal one individual, rather, one role might require several individuals or one individual might fulfill several roles (Laursen & Thorlund, 2016).

2.2.2 Categorization of BI related competency areas

Studies have offered different takes on what are considered important competencies for BI.

These are often categorized into groups to cover several areas. Chasalow (2009) categorized five competency components that impact BI success: the ability to learn, participative leadership style, clearly defined business goals, technological resource availability and financial resource availability. Another example is the 35 competencies identified by Salmasi, Talebpour and Homayounvala (2016), later divided into groups such as data management, IS/IT development or human capital resources. These studies offer a

categorization of a broad range of competencies, but they do not attend to the challenge of integrating these competencies. Furthermore, neither of these studies offers a framework for the related roles or functions that need to be involved in the different stages of BI.

Salmasi, Talebpour and Homayounvala’s (2016) categorization of BI competencies was inspired by the three competency areas of Miller, Bräutigam and Gerlach (2006). These areas are IT, data analytics and business skills, argued to be primarily technical in nature, on the organizational level and identified through studying job roles in organizations. The

competency areas have also been emphasized in a study by Laursen and Thorlund (2016) that focuses on business analytics. The study presents a model where IT, analytics and business skills make up an information wheel that connects all areas into one workflow (Lausen &

Thorlund, 2016). The model is argued to show how domain-specific information needs to be provided by business representatives, delivered to the IT function for data sourcing and derivation, for lastly being analyzed, interpreted and applied by professional analytics (ibid.)

(14)

9

Given that Laursen and Thorlund’s (2016) model is created for the broader concept of business analytics, it might also be considered applicable to BI, as BI represents the data- driven aspect of business analytics and decision making. The model is in accordance with Miller, Bräutigam and Gerlach (2006) statements that the areas of IT, data analytics and business domain offer a comprehensive definition of the competency areas required for working with BI. As these areas have guided other BI research, such as Popovič et al. (2012) and Salmasi, Talebpour and Homayounvala (2016), it will be used to further frame the literature review on BI competencies below.

2.2.3 Information technology competencies

IT competencies are described to be more vital than ever since organizations continuously search for new technological possibilities, tools and strategies for improving their competitive advantage on the market (Gudfinnsson, Strand & Berndtsson, 2015; Powell & Dent-Micallef, 1997; Chen, Chiang & Storey, 2012). This applies to BI, where companies that want to be up to date with the latest BI technologies need to look for competencies on par with the rapidly changing technological environment.

IT competencies are important for various reasons when dealing with BI. As stated by Arnott, Lizama and Song (2017), the essence of the BI system is the IT artifact, where the

responsibility of design, support, maintenance and development often lies on the IT

department. The IT aspects of BI includes various areas. One example is the data warehouse (DW) where data is collected, stored and retrieved for use (Inmon, 1995; Gartner, 2021b).

When introduced, the DW was created as a large and centralized analytic repository for historical data, mainly for the purpose of analyzing historical performance (Inmon, 1995).

Today, the data derived from the DW can be combined into an aggregated and dynamic form, suitable for advanced enterprise data analytics, as well as predictive analytics and reporting (Gartner, 2021b). Such functions require relevant knowledge for supporting data analytics and overseeing the data flows through the DW such as input, conversion, administration and extraction.

IT competencies also include management of the integration of information systems within the organization. BI systems can, for example, be connected to Enterprise Resource Planning (ERP) systems used by larger companies for consolidating financial information from the

(15)

10

divisions to group level. Having knowledge regarding the design, structure and functions of such systems, can facilitate the technical integration of BI, as well as the implementation and recommendation of appropriate BI tools (Oesterreich & Teuteberg, 2019; Zouine & Fenies, 2015).

Furthermore, BI requires technical competencies for designing and implementing data governance systems that ensure the privacy, integrity and security of the data (Oesterreich &

Teuteberg, 2019). Data governance systems need to be designed with routine maintenance of stakeholder data to make sure it is compliant with privacy laws and policies. Such systems are also described to ensure data utility, where managing the data flow through its life cycle can eliminate the risk of stale and incomplete emerging data, improve processes for

preventing data-related issues, as well as identifying and correcting inaccurate data (IMA, 2018). In this way, the IT department serves as a support function for optimizing the way that data is collected by users analyzing the data or the users for whom the data is reported. This can be done by identifying and improving weaknesses in the data flow, as well as atomizing the data collection through data models or other software tools (IMA, 2018).

As Viaene (2018) explains, the responsibility of BI often lies on the IT department alone, which is argued to create an imbalance in the representation of other aspects such as business domains. However, researchers also argue that the involvement of the IT function varies depending on the type of BI system used in the organization. According to Howson et al.

(2019), modern BI platforms do not require significant involvement of in-house IT departments to the same extent as traditional BI systems. This is due to the modern BI platforms being predefined and better adapted for self-serve use, where the IT support and maintenance are represented by vendors or external technical consultants providing the platforms (ibid.). Arguably, this does not downgrade the importance of IT competencies involved in BI, rather it shows a shift from where these competencies might be represented.

In addition, having many of the IT aspects of BI outsourced, such as the design,

implementation and development of the BI tool, does not eliminate the importance of in- house IT departments in relation to BI. The IT department can still be responsible for aspects such as ensuring information systems integration, for example, ERP systems or protecting the security of the data (ibid.).

(16)

11

2.2.4 Business and domain specific competencies

According to several researchers, BI competencies cannot be separated from business competencies (Chen, Chiang & Storey, 2012; Provost & Fawcett, 2013; Waller & Fawcett, 2013). Business competencies in relation to BI refer to the skills, experiences and knowledge of the business domain which the BI system serves (Debortoli, Müller & Brocke, 2014).

These competencies provide guidance for assessing, developing and tailoring the BI tools to fit the strategy of the specific business domain. Researchers argue that for BI to be relevant to the organization, substantial industry and business knowledge must be represented in the development of the tools (Debortoli, Müller & Brocke, 2014). Having industry knowledge is considered important for keeping pace with industry trends, identifying risks related to specific vendors or customers and formulating ways to increase competitive advantage and identifying new sources of value (IMA, 2018).

Moreover, a considerable part of developing a BI strategy is to understand how BI can leverage core business functions such as marketing, sales, operations, financial management and more (Williams, 2016). For example, for performance scorecards and dashboards to present relevant financial and operational data, it is important to first have an understanding of the business domains. Dashboards can then be better prepared by selecting the right information and creating custom-designed templates or models for how to present the data.

Knowledge of areas such as inventory, operations or supply-chain is argued to be essential for optimizing key figures such as stockouts, shipments, volume, timekeeping or logistics (Debortoli, Müller & Brocke, 2014; Davenport & Patil, 2012)

Furthermore, Laursen and Thorlund (2016) argue that business competencies are mainly important for creating relevant and actionable BI content. Business competencies revolve around the perspective of the end-users and understanding what creates value for the users.

BI needs engagement from employees closely aligned to products and processes in the organization, in a way that co-locates BI with the business units (Davenport & Patil, 2012).

Hence, business competencies can be argued to connect BI to its very purpose of being a tool for business related decision making. The information provided from BI needs to be precise and present actionable solutions for users to base their decisions on it (Laursen & Thorlund, 2016).

(17)

12

2.2.5 Data analytics and data visualization competencies

While the model of Laursen and Thorlund (2016) emphasizes integration between all the competency areas, data analytics is placed in the middle of the data flow wheel. In a way, the competences of data analytics can be seen as the mediator between IT and business domains.

This is due to the critical role of data analytics to make sense of data and appropriately communicate it to end-users. The purpose of data analytics is to extract, transform, and analyze data for gathering valuable insights, improvements and supporting decision making (IMA, 2018).

Researchers define data analytics using three levels of analytics to enhance business performance (Delen & Demirkan, 2013; Lustig et al., 2010). These levels consist of descriptive, predictive and prescriptive analytics. Descriptive analytics employ summary statistics, such as mean, mode or median, to generate insights and characterize data for describing the current stage of a company (Lustig et al., 2010). Predictive analysis uses tools such as SAS Enterprise Miner or SPSS Modeler that focuses on data mining and machine learning techniques to understand the future state of a company based on historical data (ibid.). Lastly, prescriptive analysis aims to generate decisions of what should be done in the future, using simulation or optimization tools for data analytics (Delen & Demirkan, 2013;

Lustig et al., 2010).

The level of analytics used depends on the organization’s purpose of BI, what level of analytics the BI tool offers and what competencies employees are responsible for in the data analytics. More advanced data analytics usually include analytics of big data such as

discovering hidden patterns and unknown correlations, market trends, customer preferences, and other useful analytics of unstructured data. Unstructured data needs to be transformed into appropriate forms of analysis, for example, mapping out raw data to a more structured form of data, also known as data wrangling (IMA, 2018). In addition, advanced analytics includes the development of predictive models and making use of quantitative and qualitative techniques for exploratory analysis (Oesterreich & Teuteberg, 2019).

The development of data analytics has led to studies emphasizing new important roles for organizations when striving to capture the opportunities that appear with raw data and

advanced data analytics. For example, the emergent role of data scientists has been described

(18)

13

to create strong prominence in organizations (Davenport & Patil, 2012). Data scientists are defined as the people with competence to find answers important for its business from today’s tsunami of unstructured information. This level of competency is argued to differ from the traditional data analyst who might not be able to subdue a mass of unstructured data and change it into a form in which it can be analyzed.

Whether it is the data scientist or traditional data analyst, researchers argue that both roles need to have the competence of data visualization (IMA, 2018; Davenport & Patil, 2012).

Data visualization refers to the ability to communicate data through presentations and knowing how to explain key patterns, trends and correlations that are effective and suitable for the users (Oesterreich & Teuteberg, 2019). This is completed through, for example, utilizing best practices of table and graphic designs to avoid distortion in the communication of complex information (IMA, 2018). Data visualization is shown to promote common meanings and address the risk of interpretive differences when communicating information to various actors within a business (Berinato, 2016). Understanding how to best utilize data visualization is to provide a story by evaluating visualization options and selecting the best presentation approach for the intended audience (Beranito, 2016; Davenport & Patil, 2012).

2.3 Integration of Business Intelligence Competencies 2.3.1 Integration challenges and competency gap

Having presented the three competency areas of IT, business domains and data analytics, the remaining part of the literature review will discuss research promoting integration between these, specifically the collaboration and communication between involved roles and

functions. As earlier stated, despite the importance of involving all areas when dealing with BI, organizations usually tend to perceive BI as an IT-investment, putting the responsibility of BI on the IT-function (Williams, 2016). Similarly, Salmasi, Talebpour and Homayounvala (2016) argue that BI teams with overrepresentation of IT often lack understanding of the importance of effective partnership with business areas.

Further, the overrepresentation of IT is elaborated in the study by Viaene (2008), where the representation of business domains showed to be strikingly limited. The contact between the IT-function and the rest of the business was through technical user support or once a year through a user satisfaction survey. Consequently, this caused a disparity between IT and

(19)

14

business domains, where the focus from IT was putting in place enterprise-class

technological capabilities rather than aligning the BI tools with what was considered relevant to the businesses (ibid.). Research argues that the absence of input from business domains causes a problem from the user perspective, where the business domain often represents the demand of the end-users which BI serves. An imbalance between IT and the businesses is therefore often paralleled with an imbalance of the BI suppliers and the users (Williams, 2016).

While Viaene’s (2008) study shows issues related to the imbalance between IT and business domains, it does not offer perspective into the representation of data analytics. Arguably, promoting integration of BI competencies also includes promoting the involvement of data analytics. People with knowledge of how to tackle business problems need to be brought together with the right data, but also with those who can effectively exploit the data (McAfee et al., 2012). According to researchers, there seems to be a competency gap in organizations today and a need for bridging the areas of IT, business and data analytics (Laursen &

Thorlund, 2016; Oesterreich & Teuteberg, 2019). For example, data interpretation was included among the top trends of most needed competencies in the finance sector, arguing to affect the organizational ability of data-driven decision making (van der Meulen, 2020).

Research has offered different solutions for bridging the competency gap, as well as promoting integration among the discussed areas when dealing with BI. Examples of these solutions will be discussed below.

2.3.2 Business Intelligence competency center

Business Intelligence competency center (BICC) is one way of promoting integration (Foster et al., 2015; Strange & Hostmann, 2003; Viaene; 2008). BICC is described as consisting of cross-functional teams that have defined roles, tasks, responsibilities and processes regarding supporting and promoting the effective use of BI across the organization (Miller, Bräutigam

& Gerlach, 2006). The purpose of BICC is to act as a center of expertise for BI systems, as well as to drive and support its use throughout the organization. BICC draws on the

fundamental recognition of partnering technological and business aspects for making optimal use of BI capabilities, while also defining requirements for data quality and governance (Viaene, 2008; Gartner, 2021c).

(20)

15

However, research on BICC has been criticized for not offering a model suited for all organizations and their specific needs (Safeer & Zafar, 2011). Businesses look different and must develop their own frameworks for finding the right fit. Furthermore, BICC was declared

“dead” by Gartner Research Group, arguing that although BICC has been important in the past, BI and data analytics are evolving to something more (Duncan, 2016). Instead of investing in competency centers, BI and data analytics should be integrated into the very culture and processes of an organization, creating a so-called analytics community (Duncan, 2016). This has led to more recent research focusing on new competency areas and roles, with the purpose of further integrating BI and data-driven decision making into the whole organization.

2.3.3 Competencies of data literacy and analytics translation

Data literacy and analytics translation are examples of competency areas that have received recent attention (Panetta, 2019; Henke, Levine & McInerney, 2018). These examples could be argued to fall under the competency area of data analytics, earlier presented in the literature review. However, since their very purpose is to promote knowledge of data across functions, which encompasses all areas of IT, business and data analytics, they are

categorized as a measure for integration in this literature review. These roles can also be considered more recent in the BI literature, where no common framework among researchers seems to be used.

Data literacy is described as a language that makes it possible for individuals to access, handle, critically evaluate and ethically manage data. It has been argued to become more important for organizations to learn in today's data-driven world (Panetta, 2019).

Incorporating the language of data is described as vital for organizations who want to

optimize decision making based on data (Hippold, 2019; Herring et al., 2019; Panetta, 2019;

Wolff et al., 2016). Promoting a cross-functional data language goes in line with creating a common culture and mindset of data-driven decision making (Hippold, 2019).

Similarly, the role of analytics translators meets the challenges of integrating BI related competencies. Analytics translators are described as not necessarily having advanced technical knowledge in programming or modeling yet being vital for bridging the gap between data analytics and business domains (Henke, Levine & McInerney, 2018). In line

(21)

16

with the competencies of data visualization, the emphasis for the analytics translator is on communication and interpretation of data. The purpose of analytics translators is to have an understanding of analyzed data and translate it into insights understood by the rest of the business (ibid.).

2.4 Summarization of the Literature Review 2.4.1 The Competency Assessment Framework

The literature review has presented BI research in regard to competencies and the integration between competencies. To summarize, a Competency Assessment Framework is presented in Appendix 1. The framework is primarily based on the Management Accounting Competency Framework (MACF) developed by the Institute of Management Accountants (IMA, 2018).

The categorization of competencies in the framework are based on studies from Miller, Bräutigam and Gerlach (2006) and Laursen and Thorlund (2016), and the integration of competencies from the studies by Viaene (2008) and Williams (2016).

The assessment framework uses MACF’s assessment levels, ranging from limited knowledge to expert, for describing the representation of competencies (IMA, 2018). The description examples in the assessment levels are primarily from MACF (IMA, 2018), but also from previously presented sources (Miller, Bräutigam & Gerlach, 2006; Laursen & Thorlund, 2016; Viaene, 2008; Williams, 2016; Delen & Demirkan, 2013; Osterreich & Teuteberg, 2019). It might be important to note that MACF mainly assesses the domain of accounting and finance, leaving out other critical business domains supported by BI. However, the descriptions of competencies related to accounting and finance are not included in this framework, instead, the framework is applying MACF’s categories of technology and analytics, while being more general to the business domains. For this reason, MACF is considered applicable for this study.

2.4.2 Expected representation of BI competencies

As stated in the literature review, IT competencies are described to be represented to a higher extent, where companies tend to view BI as an IT-investment and not emphasizing the importance of partnership with other functions (Viaene, 2008; Williams, 2016). Hence, despite the importance of domain competencies for increasing the user relevancy of BI, business domain competencies are argued to be less represented due to the dominance of IT.

(22)

17

Furthermore, when it comes to the level of data analytics, descriptive and predictive analytics might be considered more common in BI practice. This is mainly because advanced

prescriptive analytics and the use of unstructured data are described as recent developments not yet utilized by the majority (IMA, 2018; Delen & Demirkan, 2013; Lustig et al., 2010).

Lastly, research indicates that there is a competency gap in organizations today and a need for stronger integration and collaboration between the areas of IT, business domains and data analytics in relation to BI. However, by emphasising new fields and roles, such as data literacy and analytics translators, further measures for promoting integration between functions seems to be emerging (Viaene, 2008; Oesterreich & Teuteberg, 2019).

This expected representation of the competency areas, as well as the Competency Assessment Framework (Appendix 1), will guide the research further.

(23)

18

3. Research Design

___________________________________________________________________________

This chapter presents the choice of research design, data collection, the studied case

organization and respondents. Furthermore, the operationalization and method for analyzing the empirical material is discussed. Lastly, the chapter includes an evaluation of the research design and ethical considerations.

___________________________________________________________________________

3.1 Choice of Method

The purpose of this study is to examine the competencies represented in relation to Business Intelligence and how the integration between the competencies is being promoted in ASSA ABLOY.

For studying the representation of BI competencies, an abductive approach with a qualitative research design is considered appropriate. This is due to the abductive approach combining both the theories presented in literature review and the empirical findings for creating

conclusions to the study’s research question (Bell, Bryman & Harley, 2019). Moreover, using a qualitative research design goes in line with the purpose of creating a deeper understanding of the competencies involved with the current use of BI. Open qualitative studies are argued to be appropriate when entering new fields and for laying a foundation for future studies (Bell, Bryman & Harley, 2019). Hence, it is considered fit for this study due to limited research focusing on BI competencies, particularly in the Swedish industry sector.

Furthermore, for applying a more in-depth research approach, the design of a case study has been used. According to Bell, Bryman and Harley (2019) a case study differs from other research designs since it focuses on a situation or phenomena in one or few organizations. For this research, the focus is on one organization due to the purpose, time and scope of the study.

Making a comprehensive assessment of competencies involved with BI requires inclusion of multiple roles and functions. Studying more than one organization might increase risk of missing important aspects when making an assessment of represented competencies.

Therefore, to focus on one organization with the aim of making an assessment that is representative of the organization’s work with BI is considered appropriate.

(24)

19

Lastly, in line with the studies of Miller, Bräutigam and Gerlach (2006), job roles and role descriptions will be used for studying the representation of competencies. Interviews are used as a method for collecting data from employees and descriptions of their work in relation to specific BI systems. In this research, the interviews will be semi-structured to allow for open- ended answers while still remaining within the research topic (Bell, Bryman & Harley, 2019).

Semi-structured interviews are eligible to use when having a predefined set-up of interview questions, but still leaving room for the respondents to add additional information, formulate answers freely, as well as replying to predefined or spontaneous follow-up questions (ibid.).

3.2 Data Collection

3.2.1 Selection of organization

The organization selected for this study is ASSA ABLOY. The organization could be

described as one of the larger companies in the Swedish industry sector and a global leader in opening solutions and entrance automation - more specifically locks and keys (ASSA

ABLOY, 2021). The company has a turnover on par with larger Swedish industrial

manufacturers such as Atlas Copco and Sandvik (Statista, 2021), consisting of six divisions where two are based in Sweden.

For this study, ASSA ABLOY is considered to fulfill the requirements of the representative case (Bell, Bryman & Harley, 2019). A representative organization belongs to a certain group or category that is of research focus, in this case the Swedish industry sector. Therefore, the aim of this study is to make an assessment on BI systems in ASSA ABLOY Group and one of its divisions for understanding the competencies involved in the work with BI. The results could be used as a benchmark for other organizations in the Swedish industry sector for evaluating their BI competencies, perhaps particularly for organizations of similar size.

3.2.2 Selection of respondents

For the selection of respondents, it is vital that roles or functions from the areas of IT, business domains and data analytics in relation to BI are included. Since a part of the assessment is focused on the integration of competencies, respondents with different roles were needed to get an understanding of the collaboration and communication between functions.

(25)

20

Furthermore, one of the authors had an internship and worked part-time at ASSA ABLOY, hence, the first two respondents were selected by knowing of their involvement with BI in advance. The rest of the respondents were selected by using a snowball sampling with the help of employees in the organization, which is argued to be useful when locating relevant people for a study (Bell, Bryman & Harley, 2019). The business domain mainly represented was the finance function, which was due to respondents recommending roles involved with similar or the same BI tools. Because of the size of ASSA ABLOY, not all recommended employees were approached for interviews, instead people involved with the same BI tools, but belonging to different functions, were used as a guideline when selecting respondents.

This was for including multiple perspectives surrounding a small number of tools, rather than including too many tools with the risk of leaving out relevant input from involved roles.

In this research, seven interviews were held. The number of interviews deemed fit to the time and scope of the study, mainly because of the extensive work required for transcribing, analyzing and making assessments based on the interviews. A lower number of interviews might make it difficult to create an assessment that is as accurate and representative as possible. On the other hand, a larger number of interviews can be too time-consuming with the risk of taking focus from properly discerning and analyzing the empirical data (Bell, Bryman & Harley, 2019).

Table 1. Summarization of the completed interviews

Respondent Role Function Time Duration

Respondent A Head of Group Internal Audit

Internal Audit 48:33

Respondent B Financial Controller Group Finance 45:00 Respondent C Financial Analyst Divisional Finance 48:58 Respondent D Program Controller IT-function 56:00 Respondent E MKS Service

Improvement and Automation Manager

IT-function 46:30

Respondent F Head of Group Financial Control

Top Management and Group Accounting

57:19

Respondent G Business Analyst Supply Chain 1:01:03

(26)

21

3.2.3 Interview process

To complete the interviews, an interview guide was created (Appendix 2). The interviews lasted for approximately 45 to 60 minutes, which is argued to be within the proposed time span of semi-structured interviews (Bell, Bryman & Harley 2019; Lantz, 2013; Jamshed, 2014). The interviews consisted of 30 to 40 questions, including follow-up questions. The interviews could only be held virtually due to the current pandemic, COVID-19. According to Bell, Bryman and Harley (2019), it is preferable to conduct interviews in person since it is easier to understand the body language or expression of feelings towards certain questions.

Therefore, video-interviews were used when possible depending on respondents' preference.

In addition, the interviews were recorded to facilitate transcription, which is argued to be critical for decreasing researchers' interpretation to intervene with the analysis of the respondent’s answers (Bell, Bryman & Harley, 2019: Lantz, 2013).

3.3 Operationalization of Competency Areas

The concepts of IT, data analytics, business domains and integration were used as a basis for formulating the interview guide. The questions in the interview guide were formulated according to the descriptions presented in the Competency Assessment Framework

(Appendix 1). It was considered important that the formulation of questions represented the descriptions of the competency areas provided by the literature review, as well as the assessment levels of the Competency Assessment Framework, to properly analyze the material. Table 2 below presents examples of formulation of questions and follow-up

questions according to the description of the four concepts presented in the literature review.

Furthermore, the beginning part of the interview guide consists of introductional and open questions allowing the respondents to formulate their answers in a free manner. The second part contains more framed questions in relation to the descriptions presented under the categories in the Competency Assessment Framework. Here, the respondents were asked to describe their role and their function’s role in relation to specific aspects of BI. The purpose of asking open questions in the beginning of the interview is to limit the risk of the more specific questions to frame respondents' answers to later questions (Bell, Bryman & Harley, 2019; Lantz, 2013). Furthermore, categorizing the questions into two groups, more open or more structured, as well as under the areas of IT, data analytics and business domains, was

(27)

22

considered to generate a better flow of the interview than skipping from one topic to another (Bell, Bryman & Harley, 2019).

Lastly, as shown in Table 2, follow-up questions were included in the interview guide for further elaboration if necessary. These questions were formulated and written down in the guide to allow for standardized probes if the respondent's first answers required any further elaboration. These follow-up questions were also used if the respondent asked for additional information or struggled to understand the primary question (Bell, Bryman & Harley, 2019).

Table 2. Operationalization of the four competency areas for the interview guide

Concepts & Description Examples from the Competency Assessment Framework (Appendix 1)

Examples of Questions Example of Follow-up Questions

Information Technology (IT)

Knowledge of information systems and software, design, support and maintenance of data storage, etc.

Does your function work with information systems connected to BI?

How is the data storage handled?

What ERP systems are used in the organizations?

Do you store the data on servers or in a data warehouse?

Who is responsible for maintaining stored data?

Business Domains

Knowledge of user perspective, BI systems alignment to products and business processes, industry knowledge, etc.

What is your view on user experience in relation to BI?

Do you know what BI tools are used among your competitors?

What do you do to optimize relevancy for the users?

What do you do to benchmark or match the tool to competitors?

Data Analytics

Extracting and transforming data to insights, the three levels of descriptive, predictive and prescriptive analytics, industry knowledge, etc.

Can you describe the process of data analytics you are involved in?

Do you use analytics tools when working with data?

Do you work with transformation of data?

Do you map out data or evaluate data?

Do you use, for example, SQL, Python or #C?

Integration

Measures to bridge the competency gap and promote integration of competencies such as involvement and communication between multiple functions and roles, etc.

According to you, does the level of integration and collaboration between functions in relation to BI work well today?

What does the communication between involved roles look like?

Why? Why not?

Do you see any improvements that could be made for increasing the integration?

How often do you communicate with each other?

What form does it take? Email, chat, surveys?

(28)

23

3.4 Analysis of Empirical Data

Several steps were conducted for analyzing the empirical material. Firstly, the interviews were transcribed to facilitate categorization, assessment and analysis of the respondents’

answers. Transcription has been described as one of the more extensive and time-consuming processes of a qualitative study (Bell, Bryman & Harley, 2019). For this study, the

transcription of seven respondents resulted in 83 pages.

Secondly, since the interviews were held in Swedish the transcriptions had to be translated.

According to Xian (2008), translation processes have a risk of creating inaccuracies due to respondents using words or phrases that do not exist or can be difficult to translate to another language. However, since Swedish was the most comfortable language for all respondents and both researchers, it was considered the best suited language for the interviews. It was therefore important that the researchers sought to make the translation more accurate by translating the transcriptions together. Moreover, the interviews completed for this research were transcribed the same day or the day after the interview was conducted. This was due to Bell, Bryman and Harley’s (2019) arguments that it is important to not distort the received information from the respondents that is in accordance with creating a strong analyzing process.

Lastly, the analysis was guided through Bell, Bryman and Harley’s (2019) description of a thematic analysis, consisting of three steps: first-order themes, second-order themes and one aggregated dimension. According to Ryan and Bernard (2003) a theme can be described as a category that is identified in the empirical data and relates to the research focus. For this study, the first-order themes were IT, business domain, data analytics and integration, based on the Competency Assessment Framework (Appendix 1). The second-order themes

consisted of the assessment scale included in the Competency Assessment Framework, ranging from limited knowledge to expert. The purpose was to fit the descriptions of the respondents and additional BI related material with the descriptions of each category in the assessment framework. Finally, the last stage of the analysis, the aggregated dimension, consisted of combining the assessment made from the different BI tools into one final

assessment. This was done by placing all the assessments into one assessment framework and selecting the assessment levels that were most represented by the studied BI tools.

(29)

24

3.5 Evaluation of the Research Design

3.5.1 Credibility, dependability and data triangulation

Researchers argue that for a qualitative study, the internal validity corresponds to the

credibility of the findings and whether the results can be considered reliable and valid (Bell, Bryman & Harley, 2019). Using multiple observers or researchers is argued to be a method for increasing credibility (Eisenhardt & Graebner, 2007). This is considered particularly important for this study where one of the researchers worked at ASSA ABLOY. The researcher had insight into the BI work for the finance function before conducting the research. Hence, it was important that both researchers were present during all stages of the research, from formulation of the interview guide to analyzing respondent’s answers, for decreasing the subjective interpretation of one individual researcher to intervene with the research.

In addition, the credibility of the study relied on the respondent’s answers and whether they are true representations of their experiences working with BI. According to Berg and Frost (2005), interviewing employees creates a risk of subjectiveness such as overstating the positive aspects of their jobs. Hence, for this study, it was important to include respondents from multiple aspects in relation to BI such as management, maintenance and user

perspective, as well as the areas of IT, business domain and data analytics. Another way of improving the credibility of a case study is to triangulate the sources from the collected data (Eisenhardt & Graebner, 2007). Data triangulation involves more than one source of data, where additional relevant documents and materials regarding the studied BI tools were used.

This was provided by the organization and consisted of user guides, screenshots of the BI tools, documentation from consultants or ASSA ABLOY’s website. The extra-material was used as a check against the respondent’s answers for confirmation of the descriptions of the BI tools or level of analysis, which was analyzed and presented with data from the

interviews.

Lastly, reliability corresponds to dependability in a qualitative study, ensuring that the findings are likely to apply at other times if they are being replicated (Bell, Bryman &

Harley, 2019). To increase the dependability, the Competency Assessment Framework (Appendix 1) was used for formulating the research questions and analysing the empirical data. The Competency Assessment Framework offers guidance based on the literature review

(30)

25

that can facilitate replication of the study and decrease researcher's values or interpretation to intervene with the findings.

3.5.2 Ethical considerations

The ethical considerations for this study were based on the four areas of Diener and Crandall (1978): (1) Whether there is harm to participants, (2) lack of informed consent, (3) an

invasion of privacy or (4) any kind of deception involved. The nature of this research and topic was not considered likely to harm any participants, neither physical nor psychological.

The respondents were also informed of the research topic and the purpose of their

involvement before accepting whether to participate or not. In addition, the respondents were informed of their anonymity and asked if their job titles could be documented for the study.

Lastly, respondents were informed of the estimated length of the interview, that they did not have to answer all the questions and could withdraw at any time.

Since the questions were formulated to revolve around the respondent’s professional roles in relation to BI, invasion of privacy was not considered to be a risk. Furthermore, the ethical considerations of doing individual assessments for creating an aggregated assessment was regarded. Rather than focusing on the individual respondents, the focus of the empirical material is on the BI tools discussed by the respondents. As earlier stated, representation of competencies refers to the representation of roles and functions involved with the BI, not individual skills and knowledge.

Lastly, ethical considerations increasingly emphasize the importance of openness and honesty in communicating information about the research to all interested parties (Bell, Bryman &

Harley, 2019; Bell & Wray-Bliss, 2009). This arguably goes further than informed consent, ensuring benefits of both parties and making the interviews a collaborative effort where participants are seen as moral beings rather than research subjects (Bell, Bryman & Harley, 2019). Therefore, it was important to communicate that the aim of the assessment was also to be of value for the respondents of ASSA ABLOY, by offering a compilation of the strengths, weaknesses, opportunities and challenges of their BI work from the study’s findings.

(31)

26

4. Empirical Findings

___________________________________________________________________________

This chapter presents the empirical findings starting with a description of the BI tools discussed by the respondents, followed by the areas of IT, business domains, data analytics and integration. The chapter ends with an aggregated assessment.

___________________________________________________________________________

4.1 The BI Tools

All respondents were asked to describe the BI tools which they were involved in. Most of the respondents were involved with multiple tools, where their roles would differ between being users of one or multiple tools or being service owners responsible for maintenance and development of a tool. The main tools discussed were two tools from Qlik, for Financial Reporting and Supply Chain, and three tools from Microsoft Power BI, Internal Audit, Operations and IT. The five tools are presented below and are based on the reponsondents descriptions including screenshots, user guides and an IT architecture document.

Qlik Sense Financial Reporting

The Finance tool is in the late development phase and early implementation phase. The purpose of the tool is to digitize and automize the monthly financial reporting used for performance reviews and analysis by the group and its divisions. Currently, the tool has separate applications for all divisions with possibilities to view financial data for levels ranging from base level (production units) to group level, depending on the users’ accesses.

The tool retrieves monthly data from an ERP system and visualizes key figures such as organic growth, cash flow, and EBIT. Group Finance are the service owners, represented by Respondent B.

QlikView BI Supply Management (BISM)

The BI Supply Management (BISM) tool is currently using QlikView but has plans to migrate to Qlik Sense. BISM includes two parts, where the first gathers data of direct material, components and other articles for supporting spend analysis. The second part includes analysis for indirect material such as travels, hotels and IT-services. These parts were separate applications to start with but are since three years back included into one tool.

(32)

27

It is maintained by Group Supply Chain and described to be well established with users from the divisions and their regions. Respondent G is the service owner.

Power BI Internal Audit

The Internal Audit tool was developed and implemented two years ago and is currently used by internal audits in the group and the divisions. The purpose of the tool is to visualize and support follow-up of internal audit reporting and control. It provides an overview of internal audit issues and their progress of completion. The tool also includes a dashboard used as a basis for internal audit reports presented to Group Management every quarter. Respondent A is the service owner.

Power BI Operations Division

The Operations tool is used by one of the divisions at ASSA ABLOY and its market regions.

The purpose of the tool is to make it possible for all units to compile KPIs from different data streams that are reported to the division every month. Respondent E implemented the tool on request of the Operations manager. Recently, the tool has been accepted by users and is now entering its next step with development of a new generation of reports. These reports are created to be more detailed and allow visualization of KPIs for base units and factories.

Power BI IT Dashboard

The IT tool is a dashboard presentation of the IT-landscape used for decision support

regarding costs, IT-security and operational KPIs. The purpose is follow-up and oversight, to make sure that Global IT keeps its promises towards the business. Respondent D is the service owner, responsible for maintenance and development of the tool, together with two external consultant parties. The tool is described to be well established and used by users in the divisions.

4.2 IT and Technical Aspects

The area of IT surrounds data quality, governance, related information systems, software tools, storage and data flows. All tools, apart from the Operations tools, were described to involve external consultants in connection to the technical aspects of either development or maintenance. The service owners of the Operations tool and the IT tool belonged to or were closer connected to the IT functions, while the service owners for the remaining tools

A Case Study at ASSA ABLOY

Business Intelligence: Competencies and Cross-Functional Integration -