The dynamic management revolution of Big Data : A case study of Åhlen’s Big Data Analytics operation

The dynamic management revolution

of Big Data

A case study of Åhlen’s Big Data Analytics operation

BACHELOR DEGREE

THESIS WITHIN: Business Administration NUMBER OF CREDITS: 15

PROGRAMME OF STUDY: International Management &

Marketing Management

AUTHOR: Linus Mautner

Gustaf Rystadius David Monell

Acknowledgements

We would hereby express our greatest gratitude to everyone who has been involved and contributed to our thesis.

Firstly, we would like to thank our tutor, Katrine Sonnenschein who has guided and supported us through the research process. With her experience, tips and knowledge, she provided a lot of useful insight and feedback which helped us throughout this process.

Secondly, a big thank you to every participant who contributed with their time, knowledge and insights to our final product. Without their contributions, this research would never have been possible.

Thirdly, we would like to express our gratitude to the opposing groups which during the seminars has challenges us in valuable discussions and provided insightful feedback.

Lastly, we would like to thank Brian McCauley and Anders Melander who provided us with valuable information and guidance in our research process.

Bachelor thesis in Business Administration

Title: The dynamic management revolution of Big Data - A case study of Åhlen’s Big Data Analytics

operation

Authors: Linus Mautner, Gustaf Rystadius, David Monell Tutor: Katrine Sonnenschein

Date: 2020-05-16

Keywords: Big Data Analytics; Big Data Analytics Capabilities; Ambidexterity; Agility; Big Data

__________________________________________________________________________________

Abstract

Background: The implementation of Big Data Analytics (BDA) has drastically increased within

several sectors such as retailing. Due to its rapidly altering environment, companies have to adapt

and modify their business strategies and models accordingly. The concepts of ambidexterity and

agility are said to act as mediators to these changes in relation to a company’s capabilities within

BDA.

Problem: Research within the respective fields of dynamic mediators and BDAC have been

conducted, but the investigation of specific traits of these mediators, their interconnection and its

impact on BDAC is scant. This actuality is seen as a surprise from scholars, calling for further

empirical investigation.

Purpose: This paper sought to empirically investigate what specific traits of ambidexterity and

agility that emerged within the case company of Åhlen’s BDA-operation, and how these traits

are interconnected. It further studied how these traits and their interplay impacts the firm's talent

and managerial BDAC.

Method: A qualitative case study on the retail firm Åhlens was conducted with three participants

central to the firm's BDA-operation. Semi-structured interviews were conducted with questions

derived from the conceptual framework based upon reviewed literature and pilot interviews. The

data was then analyzed and matched to literature using a thematic analysis approach.

Results: Five ambidextrous traits and three agile traits were found within Åhlen’s

BDA-operation. Analysis of these traits showcased a clear positive impact on Åhlen’s BDAC, when

properly interconnected. Further, it was found that in absence of such interplay, the dynamic

mediators did not have as positive impact and occasionally even disruptive effects on the firm’s

BDAC. Hence it was concluded that proper connection between the mediators had to be present

in order to successfully impact and enhance the capabilities.

Table of Contents

1 Introduction ... 1

1.1 Background ... 1

1.2 Problem Discussion ... 4

1.3 Research Purpose and Question ... 4

1.4 Delimitations ... 5

1.5 Definitions ... 5

2.0 Literature review ... 7

2.1 Method for collection of literature ... 7

2.2 Big Data and Big Data Analytics ... 8

2.3 The History of Capabilities ... 9

2.4 Big Data Analytics Capabilities ... 9

2.4.1 Management Capabilities ... 10

2.4.2 Talent Capabilities ... 11

2.5 Ambidexterity ... 12

2.6 Agility ... 14

2.7 Conceptual Framework ... 15

3. Methodology and Method ... 17

3.1 Methodology ... 17

3.1.1 Research Philosophy ... 17

3.1.2 Research Approach ... 18

3.1.3 Research Strategy ... 18

3.1.4 Case Company- Åhlens ... 19

3.2 Method ... 20

3.2.1 Primary Data ... 20

3.2.3 Sampling Process ... 21

3.2.4 Semi-Structured Interviews ... 22

3.2.5 Interview Guide; Operationalized Template ... 22

3.2.6 Data Analysis ... 24

3.2.7 Ethical Considerations and Data Quality ... 24

4 Empirical Findings ... 27

4.3.2 Transparency and Orientation ... 32

4.3.3 Cross-Functionality ... 33

4.4 Big Data Analytic Capabilities ... 34

4.4.1 Managerial Capabilities ... 34

4.4.2 Talent Capabilities ... 36

5 Analysis ... 38

5.1.1 Exploitation ... 39

5.1.2 Exploration ... 40

5.2 Analysis of Agility Traits ... 41

5.2.1 Adaptability ... 41

5.2.2 Transparency and Orientation ... 41

5.2.3 Cross Functionality ... 42

5.3 Interconnection of Ambidextrous and Agile Traits ... 42

5.3.1 Interconnection of Ambidextrous Traits ... 42

5.3.2 Interconnection of Agility Traits ... 43

5.3.3 Interconnection between the dynamic mediators ... 44

5.4 Impact on BDAC ... 44

5.4.1 Impact on Managerial Capabilities ... 45

5.4.2 Impact on Talent Capabilities ... 46

5.5 Summary of Analysis ... 48

6. Conclusion ... 49

7.0 Discussion ... 51

References ... i

Appendices ... x

Appendix 1 BDAC Categorization ... x

Appendix 2 Follow-up interview ... xi

Appendix 3 Consent Form ... xii

Table 1. 5V’s Definitions ... 8

Table 2. Interview Overview ... 21

Table 3. Interview Guide ... 23

Table 4. Traits of Ambidexterity ... 28

Table 5. Traits of Agility ... 31

Table 6. BDAC at Åhlens ... 34

Figure 1. Conceptual Framework (Mautner, Monell & Rystadius) ... 16

1 Introduction

_________________________________________________________________________________

This chapter entails a background to the conducted research and introductory information of dynamic mediators and capabilities in a BDA-operations. This is followed by an outline of the problem discussion, the emerged purpose and research questions of the study. Finally, delimitations and important definitions to the research are presented.

–––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––

1.1 Background

The popularity of analyzing the abundant amount of information and data available as a consequenceof the globalization and digitalization of today’s business has increased enormously in recent years, becoming the top priority of many firms (Conboy, Mikalef, Dennehy & Krogstie, 2020). The general expression of this copious amount of information is referred to as Big Data, which is a term for all existing data that companies collect from various sources (Baesens, Bapna, Marsden, Vanthienen & Leon Zhao, 2016). This information can thereafter be measured to extract useful information to predict and indicate behavior, preferences or financial outcomes, hence improving decision-making and business performance (Baesens et al., 2016; McAfee & Brynjolfsson, 2012). As a consequence of its potential, Big Data has been called the “new oil” of the century, now becoming many companies' most important asset (Rotella, 2012; Kroes, 2015).

However, Big Data is useless to its user in its raw form and therefore, analytics of Big Data is required to harvest all of the data to come up with useful information that companies can utilize (Davenport, 2014). Big Data Analytics (BDA) can hence be defined as the process of extracting knowledge from vast amounts of data to generate improved value, performance and competitive advantage (Wamba, Akter, Edwards, Chopin, & Gnanzou, 2015) and is therefore a crucial component to create value from Big Data. BDA offers managers a deeper knowledge about their business which in hand can enhance their ability to decision-making. By asking “what do we know?” rather than “what do we think”, a data-driven organization can reshape the way decisions are made (McAfee & Brynjolfsson, 2012). A company that is successfully capitalizing on BDA, seen as a frontrunner within the field, is the leading retail firm Amazon (Wamba, Gunasekaran, Akter, Ren, Dubey & Childe, 2017). Approximately 35% of the total sales recorded by Amazon in 2016 came from the personalized purchase recommendations produced by BDA-technology. This goes to show the tremendous value that lies within BDA.

The rapid speed of development within Big Data technology, its increased adoption and its importance within various industries have also caused problems and questions regarding how to manage the change

of this expansion (Sivarajah, Kamal, Irani, & Weerakkody, 2017). Scholars in the field of BDA have agreed that this sort of information technology that enables structuring and analyzing enormous volumes of data will not only reshape the way of conducting business, but is even described as a management revolution (Mikalef, Krogstie, Pappas, & Pavlou, 2020; McAfee & Brynjolfsson, 2012). Moreover, the speed of Big Data growth has now surpassed Moore’s law, which is an indicator of rapid technological progress (University Press, 2009). This creates opportunities as well as great challenges of how to properly manage the large increase of available information (Chen, & Zhang 2014; Sivarajah, et al., 2017).

For decades researchers have contemplated upon theories to solve the problem of how organizations should tackle the incapability of adapting to an ever-changing business environment through continuous improvements and innovation (Raisch, Birkinshaw, Probst & Tushman, 2009). Innovation and continuous development are further considered fundamental to successfully cope with the increased change in dynamic environments and can act as a mediator in the implementation and altering of Big Data technology (Aloysius, Hoehle, Goodarzi, & Venkatesh, 2018). Sprung from this, companies’ have increasingly started to develop dynamic skills and capabilities to utilize resources efficiently, acting as mediators between a company’s action and outcome (Teece, Peteraf & Leih, 2016).

O’ Reilly and Tushman (2004) described an example of dynamic mediators as a way to make steady improvements while concurrently developing radical innovative advances. This dual process of exploring for future and nuanced opportunities while simultaneously looking backwards, exploiting existing capabilities of a firm, was told to enable a successful adaptation and evolvement in such an environment (O’Reilly & Tushman, 2004). A company possessing this dual focus of operations, was termed ambidextrous (O’Reilly & Tushman, 2004). Luger, Raisch, & Schimmei (2018, p.466) further described an ambidextrous organization as possessing “the ability to dynamically balance exploration and exploitation” of capabilities and resources.

Concurrent with research and practice of an ambidextrous organization, another concept to mediate between uncertainty, change and business operations has emerged, termed agility (Tsourveloudis, & Valavanis, 2002). Agility is focusing more on operational flexibility and adjustment rather than solely the integration of a dual operational focus on resources. Agility is defined as “The ability to rapidly implement an effective response to unforeseen opportunities and threats” (Li, Wu, Cao & Wang, 2019, p. 2) and hence refers to a firm's aptitude to thrive in such an environment (Lu & Ramamurthy, 2011). Ambidexterity is hence argued to feasibly influence, enable and encompass organizational agility in an organization’s altercation towards change (Zhou Bi, Liu, Fang, & Hua, 2018) and their alignment could potentially create competitive advantage (Vrontis, Thrassou, Santoro & Papa, 2017). Together, they can

successfully enable, improve and sustain a company’s ability to cope with uncertainty and change (Lu & Ramamurthy, 2011; Zhou et al., 2018).

To efficiently conduct operations within BDA, firms must develop, adapt and maintain their set of skills and competences, more specifically, their capabilities within the ever-changing field (Akter, Wamba, Gunasekaran, Dubey & Childe, 2016). This is also emphasized by Davenport (2014), stating, that incorporation of such technology requires executives to act and shape organizations to fit their business model. A business’s Big Data Analytics Capabilities (BDAC) is described as a combination of BDA related resources within an organization, and its ability to integrate and deploy these into operation (Wamba et al., 2017). BDAC has been divided into three major themes by Akter et al. (2016) which are, talent capabilities, management capabilities and technical capabilities, which are additionally said to enable enhancement of the performance of BDA within sectors such as retailing.

The retail sector has heavily adopted BDA and many are nowadays dependent on the technology to attain and sustain results and competitive advantage (Marr, 2016; Chiang & Yang, 2018). Challenges such as the erosion of customer loyalty and intensified price wars in the aftermath of digitalization and the introduction of BDA has troubled the sector (Chiang & Yang, 2018). However, the rapid revolution has created great opportunities for retailers to take advantage of the vast existent customer information to draw innovative yet fact-based actions on the behavior of the purchaser. This development, emerging from the potential outcomes of BDA, has been called a game-changer within the industry (Marr, 2015).

Åhlens, which is a part of the retail chain corporation Axel Johnson, is the leading physical retail firm in Sweden with over 60 department shops and an established e-business (Ahlens, 2020.). Holding one of Sweden’s largest customer databases with information on over 2 million members, the company several years ago began such a transformation towards digitalization, development of BDAC and a more data-driven organization, to cope with this business revolution. In such a sector of rapid change, heavily relied on BDA-technology, the dynamic concepts and mediators of ambidexterity, agility and their alignment are emphasized as enablers to cope with this development (Lu & Ramamurthy, 2011; Zhou et al., 2018; Mikalef et al., 2020). However, what effect the implementation of such philosophies have on the utilization of specific capabilities within BDA-operations, still remains unclear. No matter how much data experts analyze ones and zeros, the direction and conclusions of these analyses ultimately ends and depends on the talent and managerial qualities of the workers.

1.2 Problem Discussion

In recent years, the discourse and investigation of BDA have reached what some researchers denote as a hype (Ransbotham, Kiron & Prentice, 2016) and is characterized by constant change and development (Li, et al., 2019). High emphasis has recently been put on ambidexterity to thrive and attain long-term competitive advantage in a dynamic business environment (Luger et al., 2018). Also, the urge of agile structures and business processes has been emphasized (Li, et al., 2019). Rich contributions to the field of BDAC, and how to manage organizational capabilities in general, have been conducted (Wamba et al., 2017; Akter et al., 2016), and research have indicated that the use of BDAC can assist operations within retailing, specifically (Akter et al., 2016).

As research has been conducted on the dynamic mediators of ambidexterity and agility in general (Li, et al., 2019; Luger et al., 2018), research investigating the specific traits of these dynamic mediators emergent in a BDA-environment, their interconnections and their impact on BDAC, are scant (Wamba et al., 2017; Mikalef et al., 2020). Hence, scholars have reflected on this surprising fact given the surge and rate of adoption and implementation of BDA in business and urges additional studies of how dynamic mediators of capabilities, such as organizational ambidexterity and agility, affects the utilization of BDAC (Mikalef, et al., 2020).

More specifically, there is a lack of empirical research to understand how the management of dynamic mediators affect the constructs of BDAC and the achievement of competitive advantage (Gupta & George, 2016; Mikalef et al., 2020). In general, reports touching upon this issue originates from unreliable and biased sources such as popular media and consultancy firms, and as a result there is a lack of reliable theoretical insights and support regarding this field (Mikalef et al., 2020). Hence, the lack of a qualitative and empirical investigation of how specific traits of the mediators of ambidexterity, and agility and the combination of these impact the utilization of BDAC is recognized as a gap within the current literature. Finally, research has highlighted that realizing operational success does not lie in the technological aspects of BDAC (Mikalef et al., 2020). Instead, enhancements of managerial and talent compounds of a BDA-operation are argued to be the critical factors to improve.

1.3 Research Purpose and Question

The purpose of this paper is to fill the gap within the existing literature and add to the field of research by underlining what traits ambidexterity and agility that emerges in a retail BDA-operation and how their combination impacts the managerial and talent BDAC identified by Akter et al. (2016). The authors aspire to discover and outline key features of these mediators and their interrelation at the

BDA-operation of the retail firm Åhlens and how they impact their established BDACs, which then could function as direction and assistance in firms’ employment of BDA operating in similar industries. This ultimately led to the following research questions:

1. What are the most emergent traits of ambidexterity and agility, and how do they interconnect within a retail Big Data Analytics operation?

2. How does these traits and their combination impact the managerial and talent capabilities within Big Data Analytics?

1.4 Delimitations

The empirical research of this study is solely conducted on one firm within the retail industry. The authors believe that the obtained results are viable and conclusive given the established knowledge of the interviewees and the position of the company within the BDA-sector. Still, making comprehensive, global yet reliable conclusions would entail a larger multiple case study encompassing a wider spectrum of industries, firms and interview objects. Such investigations were deemed too broad given the scope of this thesis.

Further, the conceptual framework is selectively delimited to exclude technological capabilities of BDAC. This since extensive research already has been conducted within this subdivision of BDAC compared to the managerial and talent aspects (Akter et al., 2016). The authors further deemed the latter aspects of BDAC more related to the research team’s initial knowledge and interest as well as the field of business administration. Finally, by including technological sub-categories of BDAC, the scope of this thesis was deemed to become too broad, potentially jeopardizing the overall quality of the study and was therefore excluded.

1.5 Definitions

Agility in a business context refers to a firm's ability to cope with uncertain and rapid change and thrive in an environment of unpredictable and changing opportunities (Lu & Ramamurthy, 2011).

Ambidexterity

Ambidexterity in a business context refers to the capacity of effectively managing a business through the combination of exploitation of current resources, and exploration of new opportunities and trends to cope with changing environments (O’Reilly & Tushman, 1996; Luger et al., 2018).

Big Data is a term for all existing raw data, characterized by the compounds of the 5 v’s: Volume, Velocity, Variety and Value, collected from various sources to enable extraction to make information-based decisions (Baesens et al., 2016). Further elaborated on the respective compounds will be provided in the literature review.

Big Data analytics (BDA):

The universal process of extracting knowledge from Big Data to generate improved value, performance and competitive advantage (Wamba et al., 2015).

Big Data analytics capabilities (BDAC):

Big Data analytics capabilities are widely defined as the enablement of companies to attain useful insights from the management of data, its infrastructure and human talent in order to alter these resources to a competitive strength (Kiron, Prentice, & Ferguson, 2014; Akter et al. 2016).

Talent BDA-Capabilities

Talent capabilities refers to the ability of an employee (e.g., analytics skills or knowledge) to act upon and perform tasks within a BDA-environment (Akter et al., 2016).

Management BDA-Capabilities

Management capabilities are described as the ability of a BDA-entity to manage procedures and practices in a structured routine, rather than based on an ad-hoc procedure, to cope with IT-resources in relation to business demand and priority (Akter et al., 2016).

2.0 Literature review

__________________________________________________________________________________

In the following chapter, concepts, theories and definitions previously introduced will be further elaborated upon in order to construct a theoretical and conceptual context that will provide a foundation for analysis and discussion. The first section reviews literature of Big Data and BDA followed by a history of capabilities which lead up to the term known as BDAC. The concepts of ambidexterity and agility will then be elaborated upon. Lastly, the conceptual framework is outlined and described which provides a foundation for method, and analysis.

_____________________________________________________________________________

2.1 Method for collection of literature

To create a nuanced and relevant frame of reference with in-depth knowledge, sources from recent years were primarily used. However, searches were not exclusively limited from a certain year and onwards, as early and heavily cited work of great impact was carefully considered and included as well. These provided the authors with necessary knowledge of connotations and emergence of theories and concepts.

In the process of identifying literature, the academic databases of Primo and Google Scholar were primarily used to collect literature, mainly within the year span of 2012-2020. A systematic screening and analysis of abstracts and summaries of articles using the keywords Big Data Analytics*, Big Data

Analytics Capabilities*, Ambidexterity* and Agility* were made. Further, a combination of these

keywords was applied to receive additional and interconnected information. This helped the authors to identify and use key scholars and impactful articles. In this process, the ABS-list of journals was continually screened to see if the foundational articles were published in a journal with a higher factor than 3. If not, the articles had to have been widely cited and recently published to be selected. Articles with a lower factor were either discarded or used due to the nascent character of the investigated field of ambidexterity and agility towards the utilization of BDAC. Moreover, to assure the overall quality of selected journals, higher emphases were limitedly selected based on the criteria of maintaining high impact factors (1.2 - 5.572). Through this limitation, the authors made sure to have attained literature of importance in terms of theoretical contribution and utility for its users (Law & Leung, 2019; Bharathi, 2011).

2.2 Big Data and Big Data Analytics

90% of the world’s existing data have been generated during the last two years (Marr, 2018). Not surprisingly, research within the field of Big Data is rapidly emerging within academic literature. However, the academic research development regarding the term has not proceeded as rapidly as the adoption of practical applications. As a consequence, there is a current debate among scholars regarding a uniform and formal definition of Big Data (Sivarajah et al., 2017).

Initial research defining Big Data bases it upon the 3V’s, namely Volume, Velocity and Variety (Laney, 2001). The expanding amount of data and the evolution within the field led to the addition of Veracity (Goes, 2014) and further the fifth V: Value, as the data can create insights for the company and create a competitive advantage (Wamba et al., 2015). The perspective of 5V’s is further supported by Mikalef, Pappas, Krogstie, & Giannakos, (2018) which outlines a description from previous literature of the respective critical factor of Big Data, see Table 1.

In its raw form, Big Data is useless to the user (Gandomi & Haider, 2015). Because of Big Data features such as variety, size and rapid change, BDA is required to harvest the data. (Davenport, 2014). With the application of analytics to the field of Big Data, companies can interpret data to gain considerable insight about their business. BDA enables the user to make informed number-based decisions and data driven predictions to improve a vast range of company functions (Wamba et al., 2017).

The rapid adoption of BDA has caused issues regarding its management and utilization to draw valuable insights from Big Data (Ransbotham et al., 2016). Kim, Trimi & Chung (2014) argues a lack of skill and resources in many cases and McAfee & Brynjolfsson (2012) further identifies five key challenges

to successfully integrate BDA as leadership, talent management, technology, decision-making and company culture. To tackle these challenges, some researchers turn to the concept of capabilities as they are argued to create a sustainable competitive advantage in the Big Data environment (Wamba et al., 2017).

2.3 The History of Capabilities

To improve operational efficiency, Barney (1991) established a resource-based view explaining how firms can use and leverage their existing resources to gain and sustain a competitive advantage, acknowledged as one of the most prominent theories in this regard. The framework builds upon the distinction of resources and capabilities and these two components make up the core of the concept and have in past empirical studies gained a lot of attention (Akter et al., 2016).

Resources can be depicted as tangible and intangible assets of organizational, technology or human aspects. (Makadok, 1999). It is something that a firm can access or has access to rather than something that they can do (Größler & Grübner, 2006). Capabilities on the other hand are defined as a collection of routines, based on learned behaviors which are patterned and founded in indirect knowledge (Winter 2003). Capabilities can also be described as a set of the firm's existing resources that are non-transferable and enhances the effectiveness and gains of the other resources (Makadok, 1999)

The theory of resource-based theory paved the way for the dynamic capability theory, which aims to explain the shortcomings of resourced based view by not only identifying resources and capabilities, but also how they integrate within the firm with a capability process approach (Teece, Pisano, & Shuen, 1997). The theory attempts to identify how a firm can sustain their competitive advantage and firm performance in a changing business environment (Priem and Butler, 2001), highlighting the need for integration among resources to be agile and elastic enough to cope with such an environment (Shan, Luo, Zhou & Wei, 2019). To enable the adaptation to changing environments, dynamic capabilities allows companies to reconfigure, integrate and build their capability portfolio and resources to create lasting and sustainable competitive advantages (Teece et al., 1997).

2.4 Big Data Analytics Capabilities

The theory and concept of BDAC is derived from an IT-adapted view of the resource-based theory and dynamic capability theory, specifically tailored to a BDA-environment (Piccoli and Ives, 2005; Lu & Ramamurthy, 2011; Mikalef et al., 2018). BDAC are widely defined as the enablement of companies to attain useful insights from the management of data, its infrastructure and human talent in order to alter

these resources to a competitive strength (Kiron, Prentice, & Ferguson, 2014; Akter et al. 2016). It can also be described as the ensemble of BDA-based resources and the ability to mobilize and deploy these resources together with other assets and capabilities (Wamba et al., 2017). BDAC is identified as an important capability to leverage and create a sustainable competitive advantage within the Big Data environment (Mikalef et al., 2020; Goes, 2014, McAfee & Brynjolfsson, 2012).

Akter et al. (2016) identified and developed a framework, categorizing three dimensions of BDAC as management capabilities, talent capabilities and infrastructure capabilities. The terminology of the BDAC dimensions varies, but these three terms are often reoccurring (Davenport & Patil, 2012; Akter et al., 2016; Wamba et al., 2017). McAfee & Brynjolfsson (2012) additionally identifies the major challenges within BDAC as its infrastructure, decision-making capabilities and talent management capabilities that needs to be split across functions within the organization. Critique against the field argues that due to clashing theoretical lenses of BDAC, a uniform foundation is yet to be found (Mikalef et al., 2018). Furthermore, it is said that literature tends to overlook the organizational resources required to leverage the technology and to heavily focus upon the technical aspects of BDAC (Mikalef et al., 2020)

.

2.4.1 Management Capabilities

Akter et al. (2016, p.118) defines management capabilities as a capability that ensures “that solid business decisions are made applying proper management framework”. Kim, Shin & Kwon (2012, p.336) further defines management capabilities as “a BDA unit's ability to handle routines in a structured (rather than ad hoc) manner to manage IT resources in accordance with business needs and priorities”.

Akter et al. (2016) assembles literature from previous research and concludes four core themes that emerged:

Planning (i) identifies big-data models and business opportunities that align to improve firm

performance.

Coordination (ii) refers to a cross-functional synchronizing and implementation of BDA across the

different functions of the firm.

Investment (iii) decisions are a critical part of BDAC-management as they entail reflection of

cost-benefit analyses, providing firms with strengthened ability to meet business objectives.

Control (iv) ensures sufficient utilization and commitment of resources which for example could in

Ferraris et al. (2019) conducted further research based upon these managerial capabilities and the correlation between these factors and knowledge management to determine their effect on firm performance. The experiment concluded that there are significant relationships between these factors. Another study concluded that substantial Big Data investments does generate excess returns in the form of competitive advantage for companies, putting the ones without Big Data investment at risk (Ramaswamy, 2013). Wamba et al. (2017) identifies the most crucial management capabilities as; planning, coordination, control and decision-making which shows resemblance to Akter et al. (2016) identified subcategories with decision-making as a difference.

The current literature within management capabilities offers a vast range of different resources and capabilities that are considered important. Other researchers like Kiron et al. (2014) highlight the importance of organizational culture to leverage the use of management capabilities. This is further supported by McAfee and Brynjolfsson (2012) which mentions company culture as a crucial capability due to the shift in decision-making that can be made based upon data rather than data and the enabling and acceptance of this fact. Thus, decision-making and data driven decision is also highlighted as an important capability to manage. Davenport & Patil (2012) rather focuses on structural aspects of management capabilities like moving analytics of IT and Big Data into the core operational functions of the business and the optimal structure related to this.

2.4.2 Talent Capabilities

Talent capabilities refers to the organizational ability a firm receives from having skilled employees (e.g., analytics skills or knowledge) adequately equipped to act upon and perform tasks within a Big Data environment. This knowledge is referred to as talent capabilities and can help to create a sustainable competitive advantage (Akter et al., 2016).

Akter et al. (2016) identified four emerging themes within the subcategory of talent capabilities:

Technical Knowledge (i) refers to the knowledge of the technical aspects of Big Data. These can for

example include; operational systems, programming languages and database management systems.

Business knowledge (ii) refers to the acknowledgement and understanding of different business

environments and functions related to Big Data.

Technology management knowledge (iii) refers to the need of efficient management knowledge to

support business alignment and goals.

Relation Knowledge (iv) refers to the ability of employees to be able to work and cooperate within

Mikalef et al. (2018) identified similar topics when reviewing existing literature within the field of talent capabilities. The authors argue, similarly to Akter et al. (2016), that technical knowledge, business knowledge, relational knowledge and business analytics knowledge were the most critical capabilities, and business analytics knowledge is the only difference from Akter et al. (2016) categorization.

Other researchers like Davenport et al. (2012) argue that to gain a sustainable advantage through the use of talent, one must leverage discovery and agility. By training and recruiting employees with new sets of skills and distributing these analytic capabilities into new production environments to continuously screen for patterns, events and opportunities is vital to achieve the sustainable advantage. Mikalef et al. (2018, p. 567) further reinforces this statement by arguing that “Due to the fusion of business and IT departments in BDA firms, the importance of a liaison person has emerged: that is, a person capable of bridging the siloed departments and making them work collaboratively’’. Mikalef et al. (2020) also denotes a critical aspect to be an overall competence within BDA. Not only within the Data Scientist and IT department, but particularly in managerial positions in order to achieve a data-driven culture.

2.5 Ambidexterity

Ambidexterity is defined as a firm's ability to explore new opportunities while simultaneously exploiting existing capabilities within the firm (O’Reilly & Tushman 1996). In order to reach operational success, O’Reilly & Tushman (2013) and Teece, Peteraf, & Leih, (2016) argue that companies must increase their fit and alignment continually since evolutionary change is a prominent characteristic of today's world. Additionally, they highlight the importance of organizational change in order to adapt to be better suited for the new wave of technology and competition.

Despite the many opinions of how to achieve ambidexterity, scholars have united in the conclusion of innovation as a fundamental cornerstone and objective of the concept (March 1991; Heckmann & Maedche, 2018). March (1991) early on categorized innovation into the two subfields of exploration and exploitation. O’Reilly and Tushman (2013) and Luger et al. (2018) further elaborates on the two divisions within ambidexterity. Exploration is a firm's ability to explore and capture new resources and knowledge, and then capitalize on that information. Exploitation is the ability to better utilize already existing resources in order to create operational efficiency. The problem with exploration and exploitation is the fact that they are contradictory forces which means that organizations often focus on one of them and the key to successful utilization of an ambidextrous organisation is to be able to efficiently manage exploitation and exploration simultaneously (Raisch & Birkinshaw, 2008). Benitez, Castillo, Llorens & Braojos, (2018) extended research of the subfields of ambidexterity in relation to technology and innovation management, indicating that exploitation refers to innovation of an

incremental nature of small continuous improvements whereas exploration of new trends and opportunities refers to radical innovation of larger scale. Moreover, Benitez et al. (2018) theorizes the division in relation to knowledge management, where exploitation of knowledge refers to monitoring and improving existential knowledge within a firm and exploration entails the investigation and acquirement of nuanced and novel knowledge.

In the exploration of external insights, Kauppila, (2010) highlights partnerships as a tool to acquire new and improved knowledge. An organization alone cannot have the capabilities to solely rely on the internal knowledge and should therefore use partnerships as an element to enhance explorative innovation and knowledge (Kauppila, 2010). Mikalef et al. (2019) agrees with Kauppila’s (2010) conclusion of the importance of leveraging knowledge from outside the firm such as partnerships, but also to engage in shared problem solving to enhance innovations.

Scholars have identified that different setups of organizations are appropriate for realizing different types of innovative development Arora, Belenzon & Rios (2014). A centralized structure is said to better foster novel and radical innovations where extensive and specialized knowledge and experience are shared between groups to better acquire tacit and complex information (Arora et al., 2014). On the other hand, a decentralized structure is argued to be more successfully applied to nourish innovations of an incremental nature.

Ambidexterity has furthermore been frequently linked to enabling firm performance and agility and is therefore seen as a fundamental capability to reach long term success (Lee, Sambamurthy, Lim & Wei, 2015). Sprung from this research, ambidexterity and its relation to leveraging a firm’s capabilities, has been linked to BDAC and the effect it has on performance (Lee, et al., 2015; Wamba et al., 2017). Further, Rialti, Zollo, Ferraris & Alon., (2019) tested if ambidexterity and agility could mediate the relationship between BDAC and firm performance and the results showed a positive relation but did not investigate how specific traits of these mediators and their interplay impacted BDAC and performance. Also, previous research has predominantly emphasized the technical parts of BDA, partly excluding the managerial and talent perspective of the phenomena, which further calls for an investigation of how ambidexterity enhances capabilities of these sectors (Mikalef et al., 2020). Consequently, researchers within this topic can unite in the fact that ambidexterity is important and strongly connected to agility. Researchers, therefore, argue that ambidexterity is a core reason in organizations strive to achieve agility (O’Reilly & Tushman, 2013; Lee, et al., 2015).

2.6 Agility

Agility refers to a firm's ability to cope with uncertain and rapid change and thrive in an environment of unpredictable and changing opportunities (Lu & Ramamurthy, 2011). Blank (2013) further elaborated upon this arguing that corporations within industries of rapid transformation ought to be prepared to change and constantly update their business models to cope with uncertain internal and external conditions.

Grewal & Tansuhaj (2001) extended the research of agility, emphasizing that it should not be confused with or mistaken to be the same as flexibility in a business context. Flexibility is defined as the predesigned set of organizational resources that enables it to efficiently react and respond to changes. Hence, flexibility is a component of agility, aiding an organization’s capabilities when responding to changes with predictable features (Van Oosterhout, Waarts & Van Hillegersberg, 2006). Organizations will find themselves in situations involving unpredictable elements, and conditions of rapid unpredictable change are of constant occurrence in not only IT (Van Oosterhout et al., 2006), but particularly BDA-environments (Davenport, 2014; Mikalef, 2020). During circumstances with these characteristics, an agile response is in its place. Agility then refers to “The ability to rapidly implement an effective response to unforeseen opportunities and threats” (Li, et al., 2019, p.2) and achieving this ability in industries characterized by a high degree of uncertainty, is argued a foundation to a sustainable competitive advantage (Pavlou & El Sawy, 2006)

Research emphasizes information sharing and communication as important compounds of agility (Harraf, Wanasika, Tate & Talbott, 2015). Internal and open communication is important to circulate information throughout the organization. Three channels are considered important for the information to flow in: top-down, horizontal and bottom-up. Agility is further indicated to improve critical business functions through coordination and collaboration between internal processes, individuals and teams (Felipe, Roldán & Leal-Rodriguez, 2016). As to research regarding the organizational setup in relation to agility, many scholars accentuate a non-hierarchical structure, which enhances innovation and facilitates the organization to better tackle uncertainty (Davenport, 2014; Rialti, Marzi, Silic & Ciappei, 2018).

Regarding operationalizing agility in a business, there are two major methods of agile practices, Kanban and Scrum (Middleton & Joyce, 2012). These agile practices were first developed to suit software and IT operations but are now used among many different industries. Scrum is derived from the Kanban method, both entailing working with small units, visual boards of operative status, daily meetings and updates to allow for control, orientation, receptiveness and alteration in an operation (Middleton & Joyce, 2012; Pries-Heje & Baskerville, 2017).

Agile performance is not stated to be a universal practice towards success but can have harmful consequences to organizations operating in reasonably stable environments (Nadkarni and Narayanan, 2007). However, in rapidly changing markets, a positive correlation between agility and firm performance is maintained to exist (Tallon & Pinsonneault, 2011). There are even arguments stating that” agile is too slow” as sublimation towards the rapid development and increased uncertainty within the field of Big Data (Cukier & Mayer-Schönberger, 2013). However, in general, studies have shown that high agility is interlinked to companies becoming more likely to embrace and adapt to managerial processes of a dynamic nature as well as to integrate more innovative IT-systems within their organization (Wei, Yi, & Guo, 2014). Cegarra-Navarro, Soto-Acosta & Wensley (2016) even states that agility itself can be argued a corporate capability facilitating integration, structure and application of knowledge and assets within an organization.

Rialti et al. (2019) further posit two connections between BDAC and agility, the first being that agility and BDAC is positively interrelated and the secondly that the combination can lead to increased firm performance. In other words, agility works as an intermediary between capabilities within BDA and a corporation’s operational performance. This is supported by Mikalef et al. (2020) view that firm’s dynamic capacity positively mediates between BDAC and their performance. The correlation is further strengthened by Dubey, Gunasekaran, Childe, Roubaud, Wamba, Giannakis & Foropon (2019), accentuating that deep understanding of agility is a necessity in order to achieve and develop beneficial BDAC. However, investigations of how specific traits of agility affects components of BDAC in a BDA-environment, remains scant.

As the review of the existing literature states, there is some evidence of proven interconnections between agility and BDAC. However, the research is not very extensive hence, it could be argued that current research within the field is not sufficient enough to provide reliable conclusions in the everchanging and global business environment of BDA. Furthermore, Rialti et al. (2019) and Mikalef et al. (2020) claims that there is a need for qualitative and empirical research of how these concepts are connected and dependent on each other to provide further insight of how they impact the utilization of BDAC.

2.7 Conceptual Framework

As the topic of this study regards dynamic mediators in business and its connection to and interplay within BDA, a conceptual framework has been developed building upon the literature reviewed that aims to provide guidance to the following sections. The framework is derived from Akter et al. (2016) outline of categories and subcategories of talent- and management BDAC. Further, the addition of theory regarding dynamic mediators of ambidexterity and agility and the impact they might have on

BDAC was developed from the literature as they frequently were mentioned in relation to each other and BDAC (Wamba et al., 2017; Rialti et al., 2019). The setting of this framework has additionally emerged from a combination of the findings obtained in the theoretical framework with the empirical insights gained from the pilot interviews.

In detail, the purpose of the framework is to aid the analysis of empirical findings by enabling provenience in answering the established research questions, entailing what specific traits that emerged from the dynamic mediators, how they are interconnected and what impact they have on a firm’s managerial and talent BDAC. The extracted findings from the framework and its analysis will be presented in section 4 and 5 respectively.

3 Methodology and Method

_________________________________________________________________________________

This chapter provides motivation for the chosen methodology and outlines how the research was conducted. The first section highlights how the research has been conducted based on the topics of research philosophy, approach, strategy and case study design. The second section elaborates on the collected data, sampling process, and ethical considerations of the findings.

__________________________________________________________________________________

3.1 Methodology

3.1.1 Research Philosophy

A relativistic interpretivist philosophy was adopted in this study. The research paradigm has been drawn from the two main philosophical compounds of conducting research, ontology and epistemology (Easterby-Smith et al., 2013). Ontology refers to the philosophical stance towards the existence of objects in the world and whether they are subject to human dependence and interpretation, or not. The relativistic viewpoint of ontology is argued to suit this study, as the philosophy claims that multiple truths exist, and facts are relevant depending on the viewer's perceptions and perspectives. To contextualize, BDAC’s may differ from industries and companies or even teams, and dynamic mediators may be perceived and acted upon differently by different managers or teams.

Epistemology entails the acquirement of knowledge and learning about the social world (Easterby-Smith et al. 2013). The epistemologist philosophy of interpretivism argues to see reality as highly subjective and puts emphasis on analyzing the complexity of a social phenomenon and through this gain an interpretive understanding (Krauss, 2005; Weber, 2004). This philosophy complies both with the choice of a qualitative study and an exploratory purpose. The choice of exploratory research is supported by scant research regarding which concepts to utilize in regard to the research question. Furthermore, it enables the study to interpret and create meaningful understanding from the qualitative semi-structured interview answers to provide conclusions for the research question. The aim of the research is to gain deep-knowledge and strive to understand the emergent traits of agility and ambidexterity and how the combination of them impacts a firm's BDAC. Saunders et al. (2016) argues that observations that are heterogeneous, complex and subjective complies with the interpretivism philosophy which is in line with this study.

3.1.2 Research Approach

To employ an interpretivist philosophy in the study to gain comprehensive understanding of BDAC and affecting factors, the authors have adopted a qualitative research approach (Saunders et al., 2016).

Induction and deduction are the two most discussed research approaches where inductive approach refers to the exploration of phenomena through observations and deduction aims to validate assumptions drawn from theory (Collis & Hussey, 2014). The study's major approach is deductive but has traits of inductive elements. A conceptual framework has deductively been developed from previous literature. This framework has been utilized to design an interview guide with pre-set concepts derived from previous literature and acts as support throughout the empirical findings and analysis which complies with a deductive research approach (Saunders et al., 2016). However, inductive elements were used in the data analysis of ambidextrous and agile traits, as these emergent themes were unbiasedly established without adherence to previous literature. This complies with an inductive approach, investigating a relatively new and unknown phenomena where there is a lack of theory developed within the field (Andersen, 2012), as in the case of the lack of empirical research regarding traits of dynamic mediators. Emphasizing the combination between a deductive research approach within a qualitative study has been heavily discussed, since a combination of deductive research approach with quantitative studies are the most common (Saunders et al., 2016). However, it is argued that a deductive research will help link a qualitative study to existing research.

According to Alvesson and Skoldberg (2009), a purely deductive or inductive methodology has been used less frequently in modern studies since solely applying one approach has been considered to produce biased and less credible findings. Instead, a combinational approach does not exclude deductive theoretical influences or inductive new insights when collecting empirical data. Hence, the arrangement of this dual methodological approach extracts beneficial features of both approaches and is therefore argued to be considered an excellent choice in this study’s process of discovering patterns and explaining phenomena that are partly underdeveloped.

3.1.3 Research Strategy

From the objective of an explorative research, the authors constructed a plan to receive arguments and perspectives from the participants in order to provide an answer to the research questions and realizing the study’s purpose. This aligns with a qualitative approach, as the authors aim to extract an underlying implicit connotation of the findings rather than analyzing numerical discoveries of an explicit quantitative character (Krauss, 2005). A case study has been deemed most appropriate to accomplish this in an efficient manner. The authors conducted the interviews to attain an understanding of common

behaviors between a small sample of participants rather than less qualitative data from a larger sample. This was conducted through comprehensive interviews where the answers were interpreted and analyzed. A case study could successfully aid in this process by facilitating comprehensive data gathering of the processes and environmental factors of the context studied (Eisenhardt & Graebner, 2007). Moreover, such strategy is argued to be effective when studying a phenomenon empirically in its natural context (Robson, 2002), as in the case of the application of BDA in the retail sector. A case study can also provide extensive in-depth knowledge of a wide array of environmental variables and effectively answers “how” or “what” questions better than other strategies such as for example experiments or surveys (Yin, 2014; Saunders et al., 2016). Hence, a case study strategy is reasoned most appropriate to provide understanding of the investigated phenomena.

According to Yin, (2014) two different case studies are available, single and multiple case study. In this study, a single case study was appointed. Saunders et al. (2016) argues that a single case study is suitable when a single example provides an opportunity of analyzing certain aspects that previously has not been studied extensively (Saunders et al., 2016). Due to the scant research investigating the triangular interconnection between ambidexterity, agility and BDAC, the choice of a single case study is further emphasized. In situations where a certain case or particular representatives in the circumstance may provide extensive knowledge regarding insights related to a common field or phenomena, a single case study is deemed appropriate (Yin, 2014). Hence, Åhlen’s has been selected since they are seen as frontrunner within BDA in the Swedish retail sector and are therefore able to provide in-depth information. An additional reflection that strengthened this choice of research method was the required resource of time needed for investigations involving case studies.

Given the scope of this thesis and the circumstances concerning the emergent COVID-19 virus, the authors have deemed that conducting multiple case studies with the same extent would consume too much resources which could harm the overall quality of this study. Single case study is also argued to produce extra and more valuable theory since the observation time is larger at the specific case compared to a multiple case study (Dyer, Wilkin and Eisenhardt 1991).

3.1.4 Case Company - Åhlens

Åhlens was established in 1899 and is today one of the biggest retailers in Sweden with 60 stores throughout Sweden (Ahlens, 2020). Last year, Åhlens sold for more than 4.8 billion SEK to over 80 million customers. Main products are fashion, beauty and interior design which are exactly the same products that they started to sell in 1899. In 1996 a customer club was created which today has more than 2 million members, currently being the largest customer loyalty program in Sweden, containing abundant of consumer data. They are also part of the Axel Johnson Group which owns several

organizations in Sweden, such as Kicks, Axfood and Dustin to name a few. The case study was conducted on the two departments ‘’Customer insight and CRM’’ and ‘’E-commerce’’. Both these departments are heavily involved with BDA-operations and were therefore deemed suitable to investigate for the purpose of this study.

3.2 Method

3.2.1 Primary Data

Due to the nascent emergence of the studied phenomena, two pilot interviews were carried out to attain practical knowledge to validate the study’s topic, research questions and interview questions. The first pilot was conducted with the Chief of the Customer Insights-department at Åhlens, which enabled the authors to receive a perception of whether a single case study and the conceptual framework could be successfully applied in relation to the researched area. Furthermore, the authors interviewed a Business Intelligence consultant in the second pilot. The interviewee possessed objective yet extensive knowledge of the field, which complemented the preparation of data collection. More specifically, it widened the authors’ perception of the area through providing a somewhat contrasting perspective of the different phenomena by not being directly or solely linked to the specific retail company nor its industry. This process was essential in developing an in-depth and concrete understanding of the investigated area as well as to assess the viability of not only the study but its designed methodology (Van Teijlingen & Hundley, 2002).

The collection of primary data represents empirical findings of the research that aims to produce insightful and expressive answers to the constructed research questions. By conducting three semi-structured interviews and three follow-up interviews (see table 2), this research attempts to provide a comprehensive yet in-depth and heterogenous knowledge of the subjects discussed (Longhurst, 2003). A semi-structured questionnaire enables the authors to be adaptive in the interview yet adhering to the fundamental topics and the research questions in the interview. Moreover, it also enabled the interviewees to fully express their thoughts and arguments which enabled collection of significant data (Longhurst, 2003). The final design and questions of the interview guide emerged from insights obtained through the pilot interviews and mainly the secondary data collection. Follow-up interviews were conducted to allow for complementary elaboration on questions and insights from the previous interviews.

3.2.3 Sampling Process

Gathered data was sampled using a non-probability sampling design, since the authors selected interview participants grounded in their expertise rather than through a random selection. As the authors had a clear picture of the purpose regarding how and why to sample the units in a certain way, purposive sampling was primarily conducted, where interviewees were identified and selected based on their understanding and experience within the area (Creswell & Plano Clark, 2007). Three criteria had to be fulfilled for a candidate to be selected. Firstly, they had to be directly involved in the BDA operation. Secondly, they had to attain knowledge related to the talent and management aspect of BDA. Thirdly, a minimum of 2 years’ work experience within the field of BDA was required. The empirical research of this paper also followed a snowball sampling, as the first interviewee suggested and arranged for contact with several other suitable candidates within the corporation. This technique of chain-sampling could conceivably generate a somewhat biased setting and inaccurate results, since the sampling is heavily dependent on the initial participant, however, it is a favorable way of receiving access to sealed contexts of interview candidates (Halvorsen, 1992). Hence, the positive outcomes arguably outweigh the negatives of snowball sampling in this regard, especially during the current COVID-19 situation.

Participant Role Date of interview Duration Interview-Type Pilot Interview

1 Chief of CRM and Customer Insight at Åhlens

18/3-2020 45 min Microsoft Teams _{Video Conference}

2 IT Consultant 24/3-2020 37 min Zoom

Interview

1

Chief of CRM and Customer insgiht and Åhlens

1/4-2020 80 min Microsoft Teams _{Video Conference}

2

Team Leader for Data Scientists and Analytics Translators

7/4-2020 66 min Microsoft Teams _{Video Conference}

3 Platform owner for E-_commerce 8/4-2020 67 min Microsoft Teams _{Video Conference}

Follow up Interview

1 Chief of CRM and Customer Insight and Åhlens

21/4-2020 35 min Microsoft Teams_{Video Conference}

2

Team Leader for Data Scientists and Analytics Translators

21/4-2020 29 min Microsoft Teams_{Video Conference}

3 Platform owner for E-_commerce 22/4-2020 33 min Microsoft Teams_{Video Conference} Table 2. Interview Overview

3.2.4 Semi-Structured Interviews

Three main interviews with additional follow-up interviews were conducted (see table 2). Each main interview lasted for approximately one hour, and the follow-ups for around 30 minutes. All interviews were held in Swedish to avoid any language barriers. The interviews were informal and semi-structurally organized. This allowed for the exploratory purpose and gained flexibility and room to maneuver the discussions to attain the interviewees’ full perception of the given question (Saunders et al., 2016). The form of semi-structured interviews was also favorable to the research since open-ended and follow-up questions allowed the authors to expand, explore and develop a participant’s answer (Bryman & Bell, 2016). As the case study included respondents from different sections and positions of the company, it was pivotal to enable follow-up on given answers to reveal for instance personal arguments for similarities or differences within the firm. The semi-structured approach has been supported by several scholars (Saunders et al., 2016; Rennstam & Wästerfors, 2015) to develop an understanding and collect data of social complexities and interactions. Although, it is vital to keep in mind that a critical stance to an interview-based approach has to be kept by the authors as an interviewee can be influenced by many factors which causes their answers to be distorted and less truthful (Silverman, 2005).

Due to the COVID-19 virus, all interviews had to be redesigned from what initially was planned as face-to-face interviews, to instead be done digitally through Zoom and Microsoft Teams. The reason for conducting video calls instead of phone calls is since it allows for notification and consideration of signals, body language and non-verbal prompts, which strengthens the quality of the data collected (Saunders et al., 2016). The use of internet-based interview tools is also considered viable as a methodological approach within the social sciences (Evans et. al, 2008; Flick, 2009).

3.2.5 Interview Guide; Operationalized Template

To enable a link between the study’s literature review, identified concepts and empirical findings, an interview model was created which can be seen in Table 3. The model is derived from the conceptual framework and aims to answer the research questions in this sense. The template provided the authors with a device that allowed them to derive and ingrain theoretical concepts and viewpoints into interview questions, allowing for rich, consistent and reliable answers (Saunders et al., 2016).

Firstly, a general, open-ended question was asked in each section to gather and explore information (Collis & Hussey, 2014). Secondly, Situational Follow-up Questions were prepared and asked if more detailed answers were needed or if unclarity arose. These questions aimed to answer the Operationalized Question which states for what purpose the questions were asked in each section. The interview model

contains four Categorizations (General BDAC, management capability, talent capability and Impact on BDAC). Each of these categories were appointed subcategorization which was referred to as Categorization Proxy to deductively separate which concept or topic the questions aimed to investigate in relation to the Categorization.

Prior to the interviews, the authors sent out a brief summary including a background and the purpose of the study, definitions of theoretical terms and lastly the interview model including upcoming questions. At the start of every interview, the team briefly went through this document with the respondent to assist with any questions or unclear sections of the interview.

Table 3. Interview Guide

CATEGORIZATIONCATEGORIZATION _PROXY OPERATIONALIZED _QUESTION SITUATIONAL FOLLOW UP QUESTIONS GENERAL QUESTION

General BDAC

Roll and position

Identification of the participants position, roll and the respnosbility related to the BDA

operation

(2) Which tasks and responsbilities does you roll entail? (3) For how long have you worked at Åhlens?

(1) What is your current roll at Åhelns and how does it relate to the BDA-operation?

Use of BDA

How does the company use BDA and what are the end goals with their BDA operation?

(5) Describe the organizational structure (centralized/decetralized) in realtion to the BDA-operation

(6) Since the implemntation of BDA, how has your business procedures changed?

(4) For what reasons does Åhlens use BDA and how si the organization structured around it?

Management Capabilities

Agility How does Agility affect Åhlens Managment Capabilties?

(8) What are the effects of Agility at Åhlens? (9) Do you work with cross-functional teams? If yes, what effects do you see with this work proceedure in terms of BDA?

(10) How does Agility effect availbity of crossfuntional resources?

(11) How would you describe Åhlens decision making process? Are decision taken based on intuion or data?

(7) Explain how coping with change and responsiveness in your organization has assisted in your utilization of Managerial capabilities?

Ambidexterity How does Ambidexterity affect Åhlens Managment Capabilties?

(12) How do you in an effective way use existing resources and capabilties in strategies, planning and work processes?

(13) How would you describe the ability to identify and cope with external changees? How do you deal with this?

(14) How do you work with the combination between exploration and explotation?

(15) How would you describe your organizational culture? Does it enhance or hinder your BDA operation? How does it affect your innovation?

Talent Capabilities

Agility How does Agility affect Åhlens Talent Capabilties?

(17) Does Agility have an impact of the talent you recruit?

(18) What do you do to coordinate your techical and business knowledge within the company? How would you say that agility effects this process?

(19) How do you work with innovation and new opportunities and how do you motivate your talent to explore for new innovations within BDA? (20) How does Agility effect cooperation and utilization of the talent knowledge?

(16)Explain how coping with change and responsiveness in your organization has assisted in your utilization Talent Capabilties?

Ambidexterity How does Ambidexterity affect Åhlens Talent Capabilties?

(21) How do you continusly screen and update your BDA knowledge? How does this knowledge effect your ability to adapt to change and innovate within BDA? (22) How do you educate your talent within BDA and do you ellaborate on using existing resources and work procedures?

(23) Is there ever resistence to changes? If yes, how do you mitiage the resistance?

Impact on BDAC

Agility and Ambidexterity

What impact does Ambidexterity and Agility have

on Åhlens BDAC?

(25) How important do you consider Agility and Ambidexterity for your BDA-operation?

(26) Related to Agility and Ambidexterity, how would you argue that Åhlens could use their existing BDAC to achieve improved BDA performance?

(27) How do you measure your BDA operation?

(24) How does the concepts of Ambidexterity and Agility impact BDAC?

3.2.6 Data Analysis

The authors manually transcribed the audio recordings from the main and follow-up interviews to allow for a structured and upright data analysis. All authors individually and collectively proof-read the separate transcriptions to avoid missing important data hence attaining a credible ground for analysis (Saunders et al., 2016).

A thematic analysis to structure the analytical process was employed (Saunders et al., 2016). This method was deemed appropriate to firstly deductively outline the established concepts accordingly from the conceptual framework. Additionally, enabling a thorough inductive analysis of emerging aspects from the collected data not directly bound to literature. In detail, this process followed four main phases. Firstly, the authors noted their received perceptions from the transcribed interviews. Secondly, these notions were labeled into codes. Next, a categorization of the codes into wider themes were conducted. Finally, emergent patterns and implications of the themes were summarized and elaborated upon.

In practice, the transcripts were firstly deductively analyzed according to Akter et al. (2016) framework of BDAC categorization. The empirical findings were scrutinized according to talent and managerial categorizations and their subdivisions of Akter et al. (2016) theory. Furthermore, the researchers analyzed the emergence of these main and subdivisions from the transcripts and applicable findings from the data was allocated to the related divisions.

Next, identification of data connected to the concepts of ambidexterity and agility in the organization were deductively analyzed through the operationalization of the interview guide. More specifically, an analysis of the emergence of these concepts in Åhlen’s BDA-operations were conducted. Further, an inductive analysis of the emergence of specific traits of these dynamic mediators were conducted. From this, five themes of ambidexterity and four related to agility emerged. These were ultimately examined and discussed by the research team, elaborating on the interconnection of these dynamic concepts as well as their impact on BDAC. A thematic analysis allowed for rich identification, analysis and structure of patterns, providing a detailed explanation and elaboration of the gathered data (Braun & Clarke, 2006).

3.2.7 Ethical Considerations and Data Quality

Anonymity and Confidentiality

As sensitive data has been gathered and handled through our qualitative data collection, the Swedish Research Councils ethical research principles were followed to ensure anonymity and confidentiality. These are: 1) Distribution of Information, 2) Informed consent 3) Protection of confidentiality, and 4)