BACHELOR DEGREE
THESIS WITHIN: Business Administration NUMBER OF CREDITS: 15 ECTS
PROGRAMME OF STUDY: International Management AUTHOR: Simon Cederholm, Anton Medbo, Matilda Varvne TUTOR: Katrine Sonnenschien
JÖNKÖPING May 2020
“What are the Main Challenges for Business-to-Business MNCs
to Implement a Data-Driven Decision-Making Strategy?”
The Major Challenges in DDDM
Implementation: A Single-Case Study
i
Bachelor Thesis in Business Administration
Title: The Major Challenges in DDDM Implementation: A Single-Case Study Authors: Simon Cederholm (970124-7232), Anton Medbo (950823-0951) & Matilda Varvne (950926-3647)
Tutor: Katrine Sonnenschein Date: May 2020
Key terms: DDDM (Data-Driven Decision-Making), DDBM (Data-Driven Business-Model), DDDM challenges, BD (Big Data), BDA (Big Data Analytics), and BI&A (Business
Intelligence and Analysis).
Abstract
Over the past years, the value of data and DDDM have increased significantly as technological advancements have made it possible to store and analyze large amounts of data at a reasonable cost. This has resulted in completely new business models that has disrupt whole industries. DDDM allows businesses to rely their decisions on data, as opposed to on gut feeling. Up until this point, literature is eligible to provide a general view of what are the major challenges corporations encounter when implementing a DDDM strategy. However, as the field is still rather new, the challenges identified are yet very general and many corporations, especially B2B MNCs selling consumer goods, seem to struggle with this implementation. Hence, a single-case study on such a corporation, named Alpha, was carried out with the purpose to explore what are their major challenges in this process. Semi-structured interviews revealed evidence of four major findings, whereas, execution and organizational culture were supported in existing literature, however, two additional findings associated with organizational structure and consumer behavior data were discovered in the case of Alpha. Based on this, the conclusions drawn were that B2B MNCs selling consumer goods encounter the challenges of identifying local markets as frontrunners for strategies such as the one to become more data-driven, as well as the need to find a way to retrieve consumer behavior data. With these two main challenges identified, it can provide a starting point for managers when implementing DDDM strategies in B2B MNCs selling consumer goods in the future.
ii
Acknowledge
This research was done in the spring of 2020 at Jönköping International Business School as the final part of the International Management program. Firstly, we would like to give a thank you to Alpha for sharing their experiences and insights on how the company has been working towards becoming more data-driven. Secondly, we would like to delegate a thank you to the interviewees working at Alpha for sharing their personal views on how this process has been carried out. Furthermore, we would like to thank our tutor, Katrine Sonnenschein, who contributed with knowledge especially in regard to the methodology section, but additionally, she contributed by exchanging ideas with the researchers during the whole process. Lastly, other thesis groups that have participated in seminars have also been a great tool for the development of this research to be carried out. Being authors for this research paper has been challenging and educative, with many new insights and especially those in regard to the field of DDDM. Finally, the authors have developed their abilities in how to conduct research, and hope that the findings established can be useful for similar cases in the future.
iii
Definitions
This section will give the explanations of words and formulations that are connected to the topic and used frequently throughout the paper. If abbreviations are applicable, they are illustrated in this section as well.
Big Data - “is the exponential growth in the volume, variety, and velocity of information, and the development of complex new tools to analyze and create meaning from such data” (Lamb, Hair, & McDaniel, 2018, p.157).
Big Data Analytics (BDA) - “is where advanced analytic techniques are applied on big data sets. Analytics based on large data samples reveals and leverages business change” (Elgendy, & Elragal, 2014, p.215).
Data-Driven Decision-Making (DDDM) - “is the adoption of a real orientation to manage big data throughout the entire decision-making cycle. Data-driven managers should base business decisions on data-analytic thinking in order to use the data collected as a driving force to prescribe actions, predict complexity and “make” the change” (Troisi, Maione, Grimaldi, & Loia, 2019, p.1).
Data-Driven Business Model (DDBM) - “is defined as a business model relying on data as a key resource” (Hartmann, Zaki, Feldmann, & Neely, 2016, p. 1385).
Business Intelligence and Analysis (BI&A) - “is an umbrella term that refers to information systems that transform raw data into meaningful information and help reduce uncertainty in decision-making” (Torres, Sidorova, & Jones, 2018, p.822).
Multinational Corporation (MNC) - “are traditionally thought of as a successful firm that have grown over many years into a large corporation that are international in their operations, vision and strategies” (Aggarwal, Berill, Hutson, & Kearney, 2010, p.557).
(Organizational) Culture - “is defined as a pattern of shared basic assumptions learned by a group as it solved its problems of external adaptation and internal integration, which has worked well enough to be considered valid and, therefore, to be taught to new members as the correct way to perceive, think, and feel in relation to those problems” (Schein, 2010, p.18).
Business-to-Business (B2B) - is when two businesses decide to have a commercial transaction between each other and for example when a company buy products from a producer to be able to create and produce its own products (Ince, D., 2019).
Business-to-Consumer (B2C) - is when a business sells services and/or products to end consumers or audiences (Doyle, 2016).
iv
Table of Contents
1
Introduction ... 1
1.1 Background ... 1 1.2 Problem Discussion ... 3 1.3 Purpose ... 4 1.4 Research Question ... 5 1.5 Perspective ... 5 1.6 Delimitation ... 52
Frame of Reference ... 6
2.1 Frame of Reference Method... 6
2.2 Big Data and Competitive Advantage ... 7
2.3 Big Data Implementation Across Different Types of Corporations ... 10
2.4 Big Data Implementation Challenges ... 12
2.5 Conclusion & Research Gap ... 16
2.6 Theoretical Framework ... 17 2.6.1 Multi-Layered Approach ... 17
3
Methodology ... 20
3.1 The Procedure ... 20 3.2 Methodology... 21 3.2.1 Research Philosophy ... 21 3.2.2 Research Approach ... 22 3.2.3 Research Purpose ... 22 3.3 Research Design ... 23 3.3.1 Research Strategy ... 23 3.3.2 Research Method... 233.4 Data Collection Techniques... 23
3.4.1 Data Collection ... 23
3.4.2 Sampling Process ... 24
3.5 Data Analysis ... 24
3.5.1 Thematic Analysis ... 24
3.5.1.1 Definition & Justification ... 24
3.5.1.2 Familiarizing Yourself With Your Data ... 25
3.5.1.3 Generating Initial Codes ... 25
3.5.1.4 Searching for Themes... 26
3.5.1.5 Reviewing Themes ... 26
3.5.1.6 Defining & Naming Themes ... 27
3.5.1.7 Producing the Report ... 27
3.6 Research Quality ... 27 3.6.1 Reliability ... 27 3.6.2 Validity ... 28 3.6.3 Generalizability ... 28 3.6.4 Ethical Considerations ... 28 3.6.5 Utility ... 29
4
Empirical Findings... 29
4.1 Results ... 30 4.1.1 Execution ... 304.1.1.1 The Decision-Making Process ... 30
v
4.1.1.3 Steps Already Taken Towards DDDM ... 32
4.1.2 Organizational Culture ... 32
4.1.2.1 Culture ... 33
4.1.2.2 Lack of Communication ... 34
4.1.3 Organizational Structure ... 35
4.1.3.1 Nature of MNC ... 35
4.1.3.2 Local Markets Differ ... 37
4.1.3.3 Business Model ... 38
4.1.4 Consumer Behavior Data ... 39
4.1.4.1 Absence of B2C Sales Channel ... 39
4.1.4.2 Lack of Competence ... 40
4.1.4.3 Suggestions ... 40
5
Analysis ... 41
5.1 Summary of Empirical Analysis ... 42
5.1.1 Execution ... 42
5.1.2 Organizational Culture ... 42
5.1.3 Organizational Structure ... 43
5.1.4 Consumer Behavior Data ... 44
5.2 Empirical Findings in the Context of Existing Literature ... 44
5.3 Theoretical Framework ... 45 5.3.1 Multi-Layered Approach ... 45
6
Conclusion ... 47
7
Discussion ... 48
7.1 Future Research ... 49References ... 51
Appendices ... 56
Appendix 1 ... 56 Appendix 2 ... 60 Appendix 3 ... 60 Appendix 4 ... 61 Appendix 5: ... 621
1 Introduction
______________________________________________________________________ This section includes a presentation of the background, problem discussion, purpose, research question, perspective, and delimitation of the research. It educates the reader on the topic, along with an explanation of the research problem and how choices have been made in response to this.
______________________________________________________________________
1.1 Background
Today, there is uncertainty and divided thoughts on how the future of business will look like. According to Wichmann (2018), partly because the technology available for companies has developed rapidly as a result of technological advancements. Even though there are divided thoughts on what are the biggest changes for the future of business, business historians from top universities in the world, including Harvard Business School and Oxford University agree that the change is not happening as fast as people think (Wichmann, 2018). Moreover, the general belief is that the change is exponentially growing, but that is not the case, as the development, in fact, is turning into a plateau phase. Development comes in waves, and during the plateau phase, it is important to implement all new technologies to prepare for the next big wave. The important thing now is to focus on implementing these new technologies in daily operations (Wichmann, 2018). The challenging part lies within that the implementation of new technologies is not only expensive, but also the appropriate competence needed is scarce (Shah, Soriano, & Coutroubis, 2018).
Brynjolfsson & McElheran (2016) explain that there have been major developments when it comes to data-storage and the technologies processing data, thus making it possible for companies to gather and store much more data, and then analyze that data to get more information to use in the decision-making process. They argue that one of the most important outcomes of these new technologies is data-driven decision making, DDDM, which seems to be a slightly misinterpreted phenomenon. A successfully implemented DDDM approach typically results in higher levels of growth, and better data results in the
2
ability to make better decisions (Brynjolfsson & McElheran, 2016). This is possible because it allows decision-makers to collect big data to retrieve information and spot trends in a less time-consuming and expensive way (Shah et al., 2018). This allows data-driven companies to be on the forefront, and the decision-makers are now able to make the decision based on “what they know” instead of “what they think”, which will increase the decision quality (Shamim, Syed, & Khan, 2019). In contrast, for those companies that have not adopted a DDDM approach, spotting trends and retrieving information is so time-consuming so when they are ready to act, the information is outdated (Davenport, Barth, & Bean, 2012).
Brynjolfsson & McElheran (2016) identified that the use of DDDM within manufacturing companies in the US increased significantly between 2005 and 2010. Key advantages in companies that made the transition were size, high level of competence (information technology, workers with the right knowledge, etc.), and awareness (Brynjolfsson & McElheran, 2016). According to Mithas, Lee, Earley, Murugesan, & Djavanshir (2013), today, different industries have changed completely as a result of DDDM, which has provided the opportunity for completely new business models. Netflix is a good example, as they have changed the way ideas for new series and movies are generated. They can be almost certain that the series and movies they release will be successful, based on data from their users, rather than relying on the gut instinct of the directors (Mithas et al., 2013). Key factors differentiating data-driven companies are economies of scale, larger investments in IT and staff with the appropriate competence, and finally, learning modalities within the company (Brynjolfsson & McElheran, 2016). In 2018, it was concluded that 84% of industry-leading firms in the United States used big data analytics, BDA, in their decision-making process to get better accuracy in their decisions (Hallikainen, Savimäki, & Laukkanen, 2019). Perrons & Jensen (2015) argue that there is a big difference in how different industries view data and big data in particular. Also, companies on the forefront view data as a valuable asset in itself, while more traditional industries, such as the oil industry, view data as descriptive information about operations and assets. Companies operating in industries like the oil industry collect a lot of data that is not used in the decision-making process (Perrons & Jensen, 2015).
3
The process of becoming a data-driven company initially includes gathering not only data, but the relevant kind of data. Secondly, the data is to be collected and stored, and finally to be set up into programs for analysis and presentation, which are both manageable and understandable for the business administration part of the company (Brynjolfsson, Hitt, & Kim, 2011). Furthermore, the adaptation of such advanced technologies does not only require capital and the appropriate competence, rather, a managerial challenge emerges in this situation in such way that the responsibility is not only on the IT-department, rather, it is the responsibility of the company as a whole to adopt the new technologies (Brynjolfsson et al., 2011).
For all different types of companies, having a goal to become more data-driven, comes with both challenges and opportunities (Finnegan & Currie, 2010). In this research, one specific corporation currently in the process of becoming more data-driven was studied, referred to as Alpha, due to confidentiality agreement. The confidentiality encompassed the name of the corporation, the names and the positions of the interviewees, as well as the definition of the industry they operate in. These issues are discussed in more detail in the section 3.6.4. Ethical Considerations. Alpha is a multinational corporation, MNC, selling consumer goods through business-to-business, B2B, sales channels and they are one of the strongest players within its field.
1.2 Problem Discussion
Today, managers rely less on gut instinct and more on data when making decisions. This has been made possible because of major developments in data storage possibilities and processing technologies (Brynjolfsson, & McElheran, 2016). It is a general conclusion that most firms know BDA is important, but that they fail to exploit the benefits because they do not incorporate it in the decision-making process (Erevelles, Fukawa, & Swayne, 2016). Furthermore, the use of data-driven decisions can be linked to superior performance amongst public firms and it is concluded that key success factors for the companies that had implemented data-driven decisions are size, awareness, the capability of IT-systems, and education of workers. It follows that the combination of a high number of employees and being part of a multi-unit firm, increased the probabilities of the firm adopting DDDM. It is also agreed that the level of education of the workforce, IT-systems, and learning processes have a significant impact on DDDM (Brynjolfsson &
4
McElheran, 2016). Another established success factor is the one that an analytics culture within the company is vital to gain a competitive advantage through BDA (Hallikainen et al., 2019). Considering the level of education, it is proven to be of importance, but when adding the management capabilities into the equation it offset the level of education to be less important, suggesting that the management capabilities play an important role in the implementation process. Existing research also agrees to the importance of acting on the information retrieved from data, and not only viewing data as a technical asset (Torres et al., 2018). On this note, it was found that more traditional industries, such as the oil industry, is behind other industries in the use of DDDM, in the sense that they ignore available data, and tend to perceive data as descriptive information about other assets. More developed industries instead view data as a valuable asset itself, which further highlights the importance of data scientists, as they operate as the link between data and the decision-makers (Perrons & Jensen, 2015). Lastly, one final conclusion evident in several pieces of research is that culture plays a crucial role in the DDDM implementation process (Lunde, Sjusdal, & Pappas, 2019).Even though there is plenty of research done on general challenges when implementing DDDM, there is less done on what are the main challenges in particular for B2B MNCs selling consumer goods encounter, which make up the major research problem identified in this study.
1.3 Purpose
The authors of this research identified an MNC operating B2B selling consumer goods, that was in the process of becoming more data-driven. It became evident that this process came with a lot of challenges, and that the company had not yet managed to become truly data-driven. When looking at the current literature about DDDM, no research has been conducted to look at the main challenges for B2B MNCs in this process in particular, and how those challenges could be addressed. Hence, the purpose of this research is to contribute to the existing literature by stating the main challenges in becoming more data-driven for B2B MNCs selling consumer goods. Based on this, the following research question was formed to guide the research process.
5
1.4 Research Question
“What are the Main Challenges for Business-to-Business MNCs to
Implement a Data-Driven Decision-Making Strategy?”
1.5 Perspective
This research was conducted by the use of a single-case study with semi-structured interviews, where two subjects currently employed in Alpha were selected as participants. Together, these two employees are responsible for the digital developments in the Nordic countries, which includes working with the transition to becoming more data-driven. It was concluded that these two employees were most relevant to interview about the challenges with implementation of DDDM, after talking to several employees and managers at Alpha. This topic is complex, and most employees and even managers would not be able to identify what are the main challenges, hence the choice of the perspective selected. An alternative approach for this research was to conduct the study from the top management’s point of view, however, the decision was made to pursue with the perspective selected as there is a presumption that the Nordic markets ought to operate as a frontrunner for digitalization, hence the most appropriate decision.
1.6 Delimitation
The purpose of this research was to explore the main challenges when implementing a DDDM strategy and this was done by conducting a single-case study on Alpha. This research was limited to one corporation only, whereas the authors could have taken the approach of using multiple corporations. This limitation was done because it was a struggle to find multiple companies fulfilling all the requirements - that is similar to size, that currently were in the process of becoming more data-driven, and finally have expressed challenges during this process. Moreover, this research was also delimited to one corporation as the aim was not to study a large sample to draw general conclusions. When it comes to the sample selected, another delimitation was made to only proceed with two participants, as these are the ones in charge of the digital developments, and the decision was made that employees further down in the subsidiary would not be able to contribute additionally.
6
2 Frame of Reference
______________________________________________________________________ This part of the research will introduce the current state of the literature within the field of the challenges in DDDM implementation. Initially a method of the frame of reference is provided, followed by the three major sections that make up the literature review - namely, “Big Data and Competitive Advantage”, “Big Data Implementation Across Different Types of Corporations” and “Big Data Implementation Challenges”. Sequent, a section with conclusion and the research gap is presented. Lastly, an appropriate theoretical framework is introduced to enhance the research further.
______________________________________________________________________
2.1 Frame of Reference Method
The authors of this thesis identified some industries to be more developed than others in terms of DDDM. This recognition raised the question if the existing literature is able to explain the phenomenon of why some industries and corporations fall behind. Additionally, it raised the question if there are frameworks and guidelines on how to successfully implement a data-driven approach to decision-making.
The process of creating a frame of reference was initiated by reading peer-reviewed journals about DDDM to initially identify appropriate keywords to use in the search process. The following keywords were identified: DDDM (Data-Driven Decision-Making), DDBM (Data-Driven Business-Model), DDDM challenges, Big Data, BDA (Big Data Analytics), and BI&A (Business Intelligence and Analysis).
Using these keywords, the authors identified relevant journals within this field based on high impact factors. This was followed by finding research that either stated key success factors for data-driven companies, managerial implications, frameworks, and suggestions on how to successfully transition into a DDDM-approach. Before reviewing the literature, a delimitation of the articles was made in terms of only pursuing with articles between the years of 2010-2019, as the study has its focus on DDDM, hence literature from the most recent years was selected as this was considered to be the most relevant. When a sufficient amount of articles fulfilling these criteria were identified, the key findings and
7
contributions of each article were written into a literature review. An illustration of this is provided in the Appendices, with the purpose to provide an organized table, including authors, year, major contributions, and the like (Appendix 1). Furthermore, agreements and contradictions in the existing literature were discussed, and the research gap that there is limited research conducted on what are the main challenges encountered in the process of DDDM in B2B MNCs selling consumer goods were identified. This generated the insight that there is not much literature provided on company-specific cases, and some researchers within the field even suggest that more studies should be done to determine if the general conclusions can be confirmed when conducting qualitative research on specific types of corporations or industries, or if new challenges could be discovered. Based on this, the researchers decided to conduct a qualitative single-case study on a B2B MNC that is a strong player within its field and currently in the process of implementing a DDDM-approach, as this would add to the current literature.
2.2 Big Data and Competitive Advantage
To start out, Davenport et al. (2012) introduced the value of big data, by distinguishing between big data and traditional analytics, and claimed that there are three major ways in which they are different. Companies that capitalize on big data pay attention to data flows as opposed to stock levels, they rely on data scientists and process developers rather than data analysts, and move analytics away from the IT-function and into the core business. In their research, Davenport et al. (2012) highlighted that it is important to switch away from looking at data to assess what occurred in the past, to continuous flows, along with that the whole purpose is for companies to take actions and make decisions in response to data. Moreover, there is a certain need for competence to be able to manage big data - that is, those with a management skillset, including programming, mathematical and statistical skills in combination with the ability to communicate effectively with decision-makers. Big data brings light to the importance for businesses to rethink their relationship between business and IT, as well as their respective roles. The major theoretical contribution by Davenport et al. (2012) is the insight that successful IT organizations will train and recruit people with a new set of skills who can integrate these new analytic capabilities into their production environments (Davenport et al., 2012).
8
Brynjolfsson & McElheran (2016) agreed to Davenport et al. (2012) that the level of education of workers plays a significant role in the use of data-driven decisions. This is also a recurring pattern found in the literature, in terms of the importance to find and maintain the appropriate competence possessing the data skills required in combination with communicative skills to use big data as a source of competitive advantage. Brynjolfsson & McElheran (2016) conducted a systematic empirical study with qualitative interviews on manufacturing companies to determine what factors influenced whether or not data-driven decisions were implemented (Brynjolfsson & McElheran, 2016). It was concluded that manufacturing companies in the US tripled their use of data-driven decisions between 2005-2010, and that plant managers confirmed the importance of using data in operations management. Moreover, better data created a foundation to make better decisions and the amount of data available to managers had increased significantly as a result of new digital technologies. Additionally, that educated workers in combination with capable IT-systems further increased the adaptation of data-driven decisions, economies of scale mattered as the level of adaptation in large companies was more significant compared to small firms. Finally, the number of learning processes within the company played an important role. They did not examine the importance of culture when it comes to the degree of DDDM in this study (Brynjolfsson & McElheran, 2016).
Literature that did consider culture, however, was conducted by Hallikainen et al. (2019), who investigated whether an analytics culture had any effect on how companies improve their long-term profitability by enhancing their management of customer relationships. The study examined what effects the use of BDA has on customer relationship performance and sales growth, and therefore 417 B2B firms were studied in the survey. It was concluded that BDA of customer data enhanced customer relationship performance and that it also affected sales growth. Moreover, an analytics culture played an important role in gaining competitive advantage through customer big data. One limitation with the research was the broad scope with lots of different industries, hence it did not provide evidence of the state in one particular industry. This study contributed to the existing literature by confirming that BDA has a positive effect on the customer relationship, suggesting that the use of BDA contributes to accuracy in decisions, hence an important role in gaining competitive advantage (Hallikainen et al., 2019).
9
Erevelles et al. (2016) agreed to the value of big data, but in addition, they introduced a conceptual framework to illustrate the effect of using big data on various marketing activities. They concluded that big data is considered to be a new form of capital, but even though most firms know it is important, they often fail to exploit the benefits of using big data, because they have not incorporated it in their decision-making process. They argued that an ignorance-based view combined with creative intensity was essential for firms to benefit from big data, as it let firms ask questions that are not necessarily based on the existing data and knowledge, thus it might lead to new insights about consumer preferences. In the conceptual framework, Erevelles et al. (2016), suggest companies who have failed to gain the advantages of big data to consider whether it is the physical, human, or organizational resources that limit the successful implementation of big data in their consumer analytics. Similarly to as highlighted by Hallikainen et al. (2019), they also suggest managers to look at the culture of the company to ensure employees use an ignorance-based view, rather than a knowledge-based view (Erevelles et al., 2016).
Taking the concept of big data a step further, Torres et al. (2018), spotted a gap in the literature in terms of that there is not much insight provided on the most appropriate and fruitful ways of using business intelligence systems for analysis (BI&A). They added to the literature by establishing a theoretical framework for how BI&A systems could improve firms’ performance. Torres et al. (2018) indicated that a sense-seize-transform-view of dynamic capabilities, along with that the importance of change capabilities to gain functional performance of BI&A were important in order to improve firm performance. Furthermore, they agree to both Brynjolfsson & McElheran (2016) and Davenport et al. (2012), that the level of education of the workforce was significant, but contributed to the literature with the finding that when management capabilities were added to the model, it offset the level of education to be less important. Nonetheless, Torres et al. (2018) confirmed the idea touched upon by Davenport et al. (2012), that is, the importance of taking action in response to data. However, this idea is further developed by Torres et al. (2018), who highlighted the significance to act on the information retrieved from BI&A, and that such systems should be valued as a crucial capability for competitive advantage and not just as a technical asset. The latter idea was also confirmed by Erevelles et al. (2016), who emphasized the importance of viewing big
10
data as a new form of capital. The major contribution by Torres et al. (2018), somewhat debates with the findings by Brynjolfsson & McElheran (2016) and Davenport et al. (2012), as their conclusions suggest companies would benefit by investing more in management capabilities, and less in personnel. Finally, they also suggest future research to be done by examining this on one or multiple companies by conducting qualitative case studies (Torres et al., 2018).
2.3 Big Data Implementation Across Different Types of Corporations
Not only is there a general agreement of the value of big data and how this, if implemented successfully, can operate as a source of competitive advantage, this also seems to be true across different types of corporations and industries. Hartmann et al. (2016) spotted a gap in the literature in terms of DDDM implementation and therefore conducted a study to fill this gap by introducing the concept of a data-driven business model, DDBM, in start-up companies. Hartmann et al. (2016), declared that there has been an exponential growth of data due to cloud computing and big data, but that it now is time for this to be commercialized. Companies relying on data is not a new concept, but how companies make use of other data sources and new technologies designed to exploit this data is a contribution to the existing knowledge. New technologies and innovations are often commercialized through start-up companies, hence their choice of sample. They defined a DDBM as “a business model relying on data as a key resource” (Hartmann et al., 2016, p.1385). The major managerial implications of the study are the identification and assessment of potential data sources, key activities, data-related offerings, and revenue models. Also, Hartmann et al. (2016), suggest six types of DDBMs that can be used as inspiration or blueprints for companies who desire to initiate this transition. The research has its focus on DDBM in startups, but the assumption is that the findings also can be applied to established organizations. The major theoretical implication of the research is that it can serve as a reference point for further studies and theory development in the field of DDBM and value creation from big data (Hartmann et al., 2016).
Taking this idea a step further, Shah et al. (2018) conducted a study to examine the challenges associated with big data implementation and therefore used manufacturing small-to-medium enterprises, SMEs, in their sample. They agreed to the existing literature on the importance of big data from a technological perspective but aimed to extend this
11
view by looking at SMEs, which by definition are those companies that are both feasible and flexible in the marketspace (Shah et al., 2018). Moreover, they contribute with insights in which they provided companies with tools to become more data-driven in practice today, but most importantly they also highlight the importance of these tools and methods to be used in a way for companies to continue to change and adapt to market changes in the future. Shah et al. (2018), provided knowledge about the need for SMEs companies to develop a strategy for big data implementation that is both feasible and progressive, but suggest that future research should be done to determine how this plays out in new business settings (Shah et al., 2018).
To complement the contributions of BDA in startups and SMEs elaborated by Hartmann et al. (2016) and Shah et al. (2018) respectively, Cheah & Wang (2017) extended the research by looking at DDBM in established firms. They emphasized that as the volume of big data has increased rapidly, companies have become exposed to new business opportunities displayed in unprecedented ways, but adding to the literature by indicating that this is also true for established corporations. Not only, does big data provide better goods and services, the unexpected source of it also allows for the rise of competitors, no matter the type of corporation. As a result, some corporations have therefore pursued new business models in response to the development of big data. DDBM innovation impacts both established and emerging corporations, and in the study conducted, Chinese established firms were examined. According to Cheah & Wang (2017), established corporations typically demonstrate two characteristics, namely technical stability, and labor intensity, and to minimize risks associated with new technologies, they tend to adopt only proven technologies with a clear return on investment. Based on the findings, there are three key principles that an established corporation would apply to achieve big data innovation, namely: determine market demand based on big data; develop a new business model; and finally, refine its business model. The study by Cheah & Wang (2017) provided the managerial implication insight that DDBM innovation is not limited to emerging high-technology industries or startups (Cheah & Wang, 2017).
Perrons & Jensen (2015) agree to the general idea suggested by Cheah & Wang (2017) that BDA is important also in established industries, and decided to dig deeper into one of them - that is the oil industry. Therefore, they compared the oil industry to other
12
industries, which are more developed when it comes to BDA. They concluded that the industry ignores a lot of the data available and view the data as descriptive information about other assets. In contrast, the leading companies view big data as a valuable asset itself, which was a trend identified by Perrons & Jensen (2015). This finding supports what Erevelles et al. (2016) and Torres et al. (2018) agree upon, that big data ought to be viewed as a crucial capability, but adds to the literature by indicating that because data was not viewed as a crucial capability in the oil corporations, this made up the reason why they were lagging behind in terms of DDDM. Perrons & Jensen (2015) also highlighted the importance of data scientists, as they operate as the link between the data and the decision-makers, making new business models possible as a result of big data and analytics. Lastly, for the purpose of future research they suggested three areas to be examined further: “how oil and gas companies should re-shape their contradicting and collaboration strategies in this new reality”; “how big data change the oil and gas industry models”; and finally, “whether oil and gas companies should cultivate teams of data scientists in order to address the rising importance and multidisciplinary nature of IT and data management” (Perrons & Jensen, 2015, p.120). Looking at the literature conducted on BDA in startups, SMEs, and more established industries, it provides a general confirmation of the value of BDA. Furthermore, Perrons & Jensen (2015) extends this idea further, by looking at the oil industry specifically, which further opens up for the next step to study more case-related studies to confirm or contradict the various patterns found. It is agreed upon that big data is a source of competitive advantage, and even so across different types of corporations, yet many companies still struggle with the implementation of BDA, which therefore opens up to the need to look at the challenges of BDA implementation in more detail.
2.4 Big Data Implementation Challenges
Up to this point, the literature reviewed gives evidence on the value of big data and how this can operate as a source of competitive advantage, and that this also is true across different types of corporations. Yet, companies still seem to struggle, especially the established ones, which reveals that there are challenges that inhibit corporations to become more data-driven. Janssen, Van Der Voort, & Wahyudi (2017) suggested the concept of the big data chain, which begins with the collection of data and end when data-based decisions are taken. This approach contributed to the literature on challenges
13
in big data implementation in which it allows for potential noise along the chain to be identified and removed. According to Janssen et al. (2017), it is often assumed that big data results in better decisions, but it is unclear which factors influence the decision-making quality and how decision-decision-making quality can be improved in organizations, hence the importance of studying the big data chain. For this particular case study, one challenge found in the implementation of BDA was the difficulty to routinize the work and embed the use of BDA and BDA in the operational process. The major challenge found, however, was the lack of appropriate skills and competence needed, resulting in a need for education and training. A conclusion to be drawn based on the research made by Janssen et al. (2017) is that the adoption of BDA is an evolving process. Another major conclusion and suggestion for future research are how to solve the problem of finding competent and skilled people, who also possess appropriate communication skills that effectively can be used across the big data chain (Janssen et al., 2017). The literature, therefore, gives evidence of a general agreement, which also was confirmed by Davenport et al. (2012) - that one challenge in BDA implementation is to find the appropriate competence, possessing both data analytic and communicative skills.
This idea was contradicted by McAfee & Brynjolfsson (2012), who agreed to that there are technical challenges of using big data, but contribute to the literature by emphasizing the idea that managerial challenges are even greater, typically starting with the role of the senior executive team. The five managerial challenges brought to attention in the research are leadership, talent management, technology, decision-making, and company culture. Initially, if a company lacks effective leadership, including a set of shared goals, the implementation becomes a struggle. Talent management is how to find and keep appropriate competence along with the ability to speak the language of business. The technology is a challenge as the tools needed are expensive, and the software is open source. Furthermore, in order to operate effectively, an organization should put information and the relevant decision rights in the same location. Finally, company culture has a huge impact on the success or failure of the implementation of DDDM in companies, as it requires breaking a bad habit of companies to pretend they are more data-driven than they are (McAfee & Brynjolfsson, 2012).
14
Moreover, LaValle, Lesser, Shockley, Hopkins, & Kruschwitz (2011) agrees to the idea by McAfee & Brynjolfsson (2012), in the way that when they considered the adoption barriers in BDA implementation, it was found that these are commonly related to management and culture, rather than data and technology. Based on this, they provided the existing literature with a five-point methodology for managers to apply when encountering these challenges. The methodology suggested, consisted of initially; First,
Think Biggest, which is built on the idea to change people from making decisions based
on personal experiences to base decisions on data, by emphasizing its contribution to a major goal. Secondly, Start in the Middle, indicates that managers should have the employees to initiate the process by defining the insights and questions needed to meet the business objective and then identifying those pieces of data needed for answers, as opposed to the other way around. Thirdly, Make Analytics Come Alive, includes new methods and tools that make insights more understandable and actionable - not only today, but also what will be important in the future. Furthermore, Add, Don’t
Detract, indicates that as executives use analytics more frequently to inform day-to-day
decisions and actions, this increasing demand for insights keeps resources at each level engaged. Lastly, Build the Parts, Plan the Whole, portrays the idea that managers need to understand how each piece of data foundation aligns with an overall information agenda (LaValle et al., 2011).
Shamim et al. (2019) continued down this path, that instead of challenges in BDA implementation being related to the data itself and competence, they argued that big data decision-making capabilities were mainly related to management, leadership, and organizational culture. They based this conclusion by examining these factors using quantitative techniques in Chinese firms. Shamim et al. (2019), found that indeed all factors - leadership, talent management, technology, and organizational culture - were positively related with a company’s big data decision-making capabilities, which supports the wider perspective in the literature in which these seem to be the most recurring challenges companies encounter in this process. These were also the elements considered in the framework suggested in the research. Insightfully, their major contribution is the one that organizational culture had the strongest association with big data decision-making capability. However, the research does face a major limitation in which only Chinese firms were investigated. Nonetheless, Shamim et al. (2019), still argue that the
15
findings can be used as a starting point in other contexts as well, such as business groups and multinational companies, which they also suggested for future research. Another limitation was the quantitative research used, and as this did not allow any in-depth analysis, Shamim et al. (2019), also suggested for more qualitative research on the topic to be made in order to more specifically determine how data management challenges enhance both data decision-making capabilities and quality.
Research that did explore cases with an in-depth approach to challenges encountered in the process of implementing a Customer Relationship Management, CRM, system, and specifically how these ought to be considered as integrated entities, was conducted by Finnegan & Currie (2010). They emphasized the importance of considering challenges not only on cultural, people, processual, and technological layers in isolation, but also how these were integrated, along with the concept of affordance (Appendix 2). A longitudinal multiple-case study was therefore conducted, investigating three different organizations that were in the process of implementing a CRM system. The framework, the Multi-Layered Approach derived from the study was that when considering the different layers, these ought to be taken into account along with the interrelations between affordance, weighting, interoperability, and evolution (Appendix 3). The research by Finnegan & Currie (2010) was limited to SMEs, and implementation of CRM systems, nonetheless, they argue that the findings can be used as a starting point for research investigating in other types of organizations and other kinds of system integrations strategies (Finnegan & Currie, 2010).
The final and, together with to other pieces of literature, most recent research on this topic, was conducted by Lunde et al. (2019), and as brought up in the method section of this frame of reference, BDA implementation challenges is a relatively new topic, hence the importance of recency in this case. In contrast to the previous literature, Lunde et al. (2019) exclusively considered the impact of organizational culture on big data adoption, by conducting a systematic literature review on this relation in different cultural dimensions. They concluded that, even across different cultures, there is evidence that supports the positive impact organizational culture has on big data. More specifically, they concluded a gap between BDA investments and the ability to effectively make decisions derived from data as a way to gain competitive advantage, and that
16
organizational culture played a crucial role in whether or not companies were capable to fill this gap. The research faces some limitations with the systematic literature review used, namely, the risk for bias in the choice of sample, and that it only allowed for rather general conclusions to be made. Nonetheless, Lunde et al. (2019) concluded that this a relatively new topic and because the provided literature up to this point is rather vague for companies to use in their BDA implementation, they, therefore, suggest more research to be made on more exactly how organizational culture impact this process (Lunde et al., 2019).
2.5 Conclusion & Research Gap
To effectively make use of big data, there are a few general conclusions to be drawn. Initially, there is a general agreement evident in the literature by Davenport et al. (2012), Brynjolfsson & McElheran (2016), and Torres et al. (2018) that the appropriate competence, namely a set of skills consisting of data analytical abilities in combination with communicative skills is important, in order for these to be effectively communicated between IT and business departments. Secondly, as according to Perrons & Jensen (2015), Erevelles et al. (2016) and Torres et al. (2018), there is also a need to change the way companies view the data, moving away from consider data as descriptive information, to view data as a valuable asset, or even as a crucial capability. Furthermore, the value of big data and the challenges associated with it are consistent no matter the type of corporation. However, the process of DDDM implementation is an extensive organizational change, suggesting it to be less challenging in startups and emerging industries, thus more of a struggle in more established MNCs. Lastly, it is agreed upon in the literature that most companies encounter technical challenges, but there is a contradiction whether this is the main challenge, or as according to Shamim et al. (2019), McAfee & Brynjolfsson (2012), and LaValle et al. (2011), the major obstacles in practice rather are associated to management and organizational culture. These conclusions are supported in the literature in several ways, initially in the way that both Shamim et al. (2019) and Lunde et al. (2019) found general patterns in their quantitative research, but argue that more organizational-specific research need to be done to confirm their findings. Perrons & Jensen (2015), agree to this and because they researched oil corporations, it gave evidence to that there is a need to particularly look at companies in more established industries. Shamim et al. (2019), also indicated that there is a gap in the research on this
17
topic in MNCs, as the assumption is that these companies struggle even more with BDA challenges. Lastly, there is an agreement that the field is relatively new, further supporting the need for more qualitative research to be conducted. In conclusion, the literature provides insight on what factors corporations that are successfully using DDDM have pursued, and also what factors that less data-driven corporations are lacking. There is limited research done about what are the main challenges encountered in the process of implementing DDDM in more established corporations. The current literature even suggest that more research need to be done on specific cases to gain more detailed knowledge about this topic. Therefore, there is a gap in the literature about what are the main challenges for B2B MNCs selling consumer goods, when implementing a DDDM strategy.
2.6 Theoretical Framework
2.6.1 Multi-Layered Approach
As evident in the existing literature, Finnegan & Currie (2010) suggested the
Multi-Layered Approach which highlights the importance of considering the various layers of
culture, people, process, and technology when implementing a new CRM system. The cultural layer in this framework, concerns where in the organization the new CRM strategy implementation falls, that is, on the IT or the marketing departments. If the cultural layer is not taken into account, there is a risk such strategy implementation will fall between the chairs (Finnegan & Currie, 2010). Secondly, to successfully implement a CRM system, it means involving a wide variety of people, and everyone needs to have a clear definition of the strategy to be implemented. Moreover, the theoretical framework also brings attention to gaining the senior executives support in the beginning of the new initiative, to have everyone on board in this process (Finnegan & Currie, 2010). Thirdly, there is the processual layer which concerns the shift from product-focused to customer-centric, that typically needs the management to redesign core business processes in a way that the process is initiated from the customer perspective and allows for customer feedback. Lastly, the fourth layer is technology-associated, because CRM strategies take full advantage of technological advancements. Moreover, the Multi-Layered
Approach by Finnegan & Currie (2010), argue that it is not rare for companies to tend to
18
process. The theoretical framework also brings attention to the importance of how the various layers ought to be taken into account as integrated variables, as well as along with the concept of affordance, which is illustrated below in Figure 1.
Figure 1: The Concepts of Cultural, People, Processual, and Technological Layers,
along with the Concept of Affordance from Multi-Layered Approach.
Affordance, in this context, brings attention to the adaptive properties of organizations and based on this, challenges could be assessed to further direct companies on how the strategy process ought to be initiated. The theoretical framework further suggest the interrelations between affordance, weighting, interoperability, and evolution, which are illustrated below in Figure 2, and make up an additional perspective of relevance when implementing a CRM strategy.
19
Figure 2: The Concepts of Cultural, Processual, People, Technological Layers along
with the Interrelations of Affordance, Weighting, Interoperability, and Evolution from Multi-Layered Approach.
Weighting, concerns how the layers need to be weighted in relation to their perceived impact and importance in the implementation. By weighting the challenges, it contributes to the knowledge about what direction to take in the CRM implementation process. Additionally, interoperability, could improve the way the senior management identify interpretabilities between and within the layers during different stages of the implementation, in accordance with their affordances. Hence, the Multi-Layered Approach by Finnegan & Currie (2010) makes up a relevant starting point in any case where a CRM strategy ought to be implemented as a tool to identify what are the main challenges in this process. However, the research by Finnegan & Currie (2010) was limited to SMEs, and implementation of CRM systems in particular, nonetheless, they argue that the findings can be used as a starting point for research investigating in other types of organizations and other kinds of system integrations strategies (Finnegan & Currie, 2010). Moreover, because of the ability of the theoretical framework to provide a
20
useful illustration of the integrated entities of culture, people, process, and technology along with the interpretabilities between these stages in the implementation process, the theory serves as a useful framework in addressing challenges for any size of corporation and for any system integrations strategy to be implemented (Finnegan & Currie, 2010).
3 Methodology
______________________________________________________________________ In this section, a story of the procedure, methodology, research design, data collection techniques, data analysis, and research quality of this research are presented. The methodology encompasses the research philosophy, research approach, and research purpose. Furthermore, the research design includes research strategy and research method. In the data collection techniques section, the data collection and sampling process are presented, followed by a data analysis discussion. Finally, the research quality, including reliability, validity, generalizability, ethical considerations and utility are presented.
______________________________________________________________________
3.1 The Procedure
As business students, the researchers identified that despite today’s availability of technology, there are yet many organizations who still rely on gut instinct when making decisions as opposed to rely them on data. Therefore, a search process in the existing literature on DDDM was initiated to dig deeper into this presumed problem, and it was found that indeed there are several challenges that organizations encountered when implementing DDDM strategies, and whereas MNCs in particular seemed to struggle. This gave rise to the idea to study the topic further, and one B2B MNC selling consumer goods in particular was identified who currently was struggling with the process of implementing a DDDM strategy, namely Alpha. Furthermore, Alpha was contacted and based on an initial scanning it was concluded that the Nordic markets, in contrast to other subsidiaries, possessed certain interest in, and knowledge about digitalization, hence DDDM. This initial scanning included that the authors approached multiple employees on different levels in the MNC and asked who were the most appropriate to interview regarding the topic, and all these employees referred to the two people used in the sample. Two interviewees currently employed at Alpha in the Nordic markets with experience of the topic were selected as participants as a purposive sample. As the current literature
21
provided plenty of research of big data and its value, along with common challenges corporations encounter, but less about what are the main challenges for B2B MNCs in particular, an interpretivist study was appropriate, which allowed the researchers to interpret elements emerged in the data collected. In line with this, there was also a need to proceed with semi-structured interviews with the purpose to allow for personal views on how Alpha has carried out the implementation process up until this point. The interviews were transcribed and analyzed into codes, which were studied, organized, and put into themes. Once this was completed, the authors considered the themes, and determined which ones found in Alpha confirmed or contradicted the existing literature. Insightfully, two new aspects were found in the empirical findings, and therefore these contributed to existing literature, and can provide other B2B MNCs a starting point when initiating a similar implementation to become more data-driven.
3.2 Methodology
3.2.1 Research Philosophy
Considering research philosophy, the philosophical assumptions of ontology and epistemology were taken into account, which concerns the nature of reality and what we accept as valid knowledge, respectively (Collis & Hussey, 2014). The ontological assumption concerns the fact that there are multiple realities of what are the challenges in DDDM implementation, meanwhile the epistemological assumption is that knowledge comes from subjective evidence from the participants selected. Hence, this thesis was conducted using an interpretivist paradigm, indicating that the research was an emergent process (Hudson & Ozanne, 1998), in which the researchers interpreted elements of the study, as opposed to test hypotheses. The choice of an interpretivist study allowed the researchers to explore the complexity of a social phenomenon as a basis for further understanding, as opposed to measuring social phenomena in which positivism would have been more appropriate (Collis & Hussey, 2014). The interpretivist paradigm did not indicate that the researchers entered a complete untouched field, as some knowledge and pre understanding was assumed (Hudson & Ozanne, 1988). The motivation behind the interpretivist paradigm was that it allowed the researchers to be open for new information regarding the potential challenges that particularly a B2B MNC selling consumer goods encountered, as this topic within the field of DDDM was discovered to be less touched
22
upon. Therefore, the interpretivist study allowed the researchers to consider multiple realities of the challenges, and to retrieve data from the participants in Alpha that were experienced within the topic studied.
3.2.2 Research Approach
For this research, an inductive reasoning approach was applied, providing the researchers to move from the particular to the general (Mantere & Ketokivi, 2013), which allowed thoughts and ideas in the semi-structured interviews to emerge, and based on that conclusions were drawn. However, the research possessed some deductive elements – that is when a conclusion is drawn about the particular based on the general (Mantere & Ketokivi, 2013), as the theoretical framework Multi-Layered Approach (Finnegan & Currie, 2010) was used as foundation. Yet, the overarching scientific reasoning of what was studied was inductive, as the aim was not to test predetermined factors that would prevent a successful adoption of DDDM. Rather, as existing literature suggested knowledge on the challenges B2B MNC selling consumer goods was limited, the major purpose was to develop an understanding of what these challenges were. One example of how the researchers operated inductively was in the data analysis procedure whereas the data collected in the semi-structured interviews was interpreted, codes were generated, and these were further organized into themes, and finally conclusions were drawn. This allowed the researcher to confirm or contradict existing literature, but most importantly, it was because they operated inductively as they were able to detect two completely new findings of value in Alpha. As these findings contributed to the existing knowledge, a deductive approach would not have been able provide such insights because such reasoning is bound to only one solution (Saunders, Lewis, & Thornhill, 2019).
3.2.3 Research Purpose
This research had an exploratory purpose, which was useful when the aim was “to find new insights or to assess phenomena in a new light” (Saunders et al., 2019, p.139). Considering the existing literature, up until this point, the purpose has been to generate general conclusions about the challenges in the implementation of DDDM. However, the purpose of this thesis was exploratory as the aim was to consider these challenges in a new light – that is, what are the challenges particularly for a B2B MNC selling consumer goods, along with the fact that the researchers were open to new challenges to be explored.
23
3.3 Research Design
3.3.1 Research Strategy
Along with this being an interpretivist study came also to the appropriateness to use qualitative research, and such data may be defined as “normally transient, understood only within a context and are associated with an interpretivist methodology that usually results in findings with a high degree of validity” (Collis & Hussey, 2014, p.130). The major justification for this choice was that the research gap suggested that more qualitative case studies needed to be conducted to be able to determine whether the findings discovered in Alpha confirmed, contradicted, or provided new insights that contributed to the existing literature.
3.3.2 Research Method
In the process of exploring what are the main challenges that a B2B MNC selling consumer goods encounter when implementing a DDDM strategy, a single-case study was conducted. The research examined only one unit of analysis (Saunders et al., 2019), as the data collected from Alpha solely focused on the DDDM operations that have been carried out so far, and what the major challenges have been. The researchers’ aim was to select a B2B MNC that currently are in the process of becoming more data-driven, and have expressed challenges during this process, which together make up a rather unique state. Sampling multiple cases in such a unique state, that additionally were all the same type of corporations, which would have been required for comparison, was a struggle. Therefore, a single-case study was the most appropriate choice as it allowed detailed data of high validity, case-specific for Alpha to emerge.
3.4 Data Collection Techniques
3.4.1 Data Collection
The aim of the research was to look at challenges Alpha encounter in their implementation of DDDM, and the purpose was to explore what these challenges might be, whether they confirm, contradict existing literature, or whether new challenges emerged. Therefore, the most appropriate method for data collection was semi-structured interviews, as this allowed the researchers to have a list of themes and questions to be examined, but with a
24
less structured interview technique (Saunders et al., 2019). How the interviews were made is evident in the interview guide, where the open-ended questions used are illustrated (Appendix 4). This was a beneficial data collection method, as opposed to structured interviews which would discourage the interviewees to speak freely and sharing their opinions. However, this also indicated that the thematic analysis was more time-consuming, in which structured interviews are more straight-forward in the analysis procedure (Saunders et al., 2019). In this research, the interviews were made with two employees responsible for the digital developments in the Nordic countries and they lasted for 41 minutes and 1h 32 minutes respectively. Moreover, due to the outbreak of the COVID-19 virus, the interviews were conducted on Google Hangout and these were recorded. Finally, the recorded interviews were then transcribed and analyzed using a thematic analysis (Braun & Clarke, 2006).
3.4.2 Sampling Process
The sampling process of the research was purposive sampling, which is when the researchers select participants based on their experience of the phenomenon being studied (Collis & Hussey, 2014). In practice, this indicated that the authors selected those employees in Alpha with the most appropriate understanding and knowledge of the current IT-systems and state of DDDM in the MNC, with the assumption that these employees were the most likely to contribute with data of value for the research to be carried out effectively. The major advantage of proceeding with purposive sampling was that the process was convenient, not very time-consuming and a relatively small sample was used. As the assumption was that the field of DDDM in B2B MNCs is rather untouched and complex, it justified the purposive sample to be used in order for valuable data to emerge (Collis & Hussey, 2014).
3.5 Data Analysis
3.5.1 Thematic Analysis
3.5.1.1 Definition & Justification
To effectively answer the research question and to ensure high validity of this research, a thematic analysis was used to analyze the data, which may be referred to as a “method for identifying, analysing and reporting patterns (themes) within data” (Braun & Clarke,
25
2006, p.79). Based on existing literature in the frame of reference, various themes and debates were evident, followed by a major research gap of that more qualitative case studies of B2B MNCs selling consumer goods needed to be conducted in order to determine what challenges they encountered in the process of implementing DDDM. As an attempt to fill this research gap, the data collected from Alpha was analyzed using a thematic analysis to allow the researchers to determine whether themes found in Alpha confirmed, contradicted or developed new themes. Therefore, it was of importance to go through all stages necessary to avoid jumping into conclusions of the presented qualitative data, hence, this thesis followed the phases of thematic analysis suggested by Braun & Clarke (2006), including familiarizing yourself with your data, generating initial codes, searching for themes, reviewing themes, defining and naming themes, and finally producing the report. The use of a thematic analysis by Braun & Clarke (2006) allowed the researchers to inductively analyze the data, whereas by transcribing the data, codes that represented the interpretations by the researchers were assigned, and further these were put into themes. However, slight deductive elements were evident in the thematic analysis, as themes already evident in the current literature were used as a foundation, but moving from the particular to the general was the major approach to the data analysis, which is inductive (Mantere & Ketokivi, 2013).
3.5.1.2 Familiarizing Yourself With Your Data
For this thesis, the data was collected through semi-structured interviews, that were recorded and transcribed by the researchers (Braun & Clarke, 2006). The transcription of the data allowed the researchers to ensure all data of value was captured, especially because the semi-structured interviews generated plenty of data whereas the interviewees were encouraged to speak freely about their challenges related to DDDM. Moreover, according to Braun & Clarke (2006), the transcription also allowed the researchers to become familiar with the current processes and views on how DDDM had previously been managed in Alpha.
3.5.1.3 Generating Initial Codes
The second phase encompasses generating initial codes of data extracts from the data collected (Braun & Clarke, 2006). Based on the literature review, existing knowledge gave evidence about certain challenges of importance, but as the assumption was that these findings would be extended in the case of Alpha, the interview questions were
