• No results found

Big Data : the engine of personalization? : the case of the music streaming industry

N/A
N/A
Protected

Academic year: 2021

Share "Big Data : the engine of personalization? : the case of the music streaming industry"

Copied!
88
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Big Data: The Engine of Personalization?

The Case of the Music Streaming Industry

Master Thesis

MSc. Business Administration International Management Track

Maximilian Bronowizki 11974842

Amsterdam Business SchoolUniversity of Amsterdam Dr. M.P. Paukku

(2)

Table of Contents

1. Introduction……… 6

2. Literature review ………... 8

2.1 How was it done before digitalisation?...11

2.2 Big Data………..16

2.2.1 The Five V’s………...17

2.2.2 Controversy around big data………..18

2.3 The role of technology……….………..20

2.4 Personalization……….………..23

2.5 The bigger picture……….……….26

2.6 The music streaming industry………...………30

2.6.1 Deezer………...………34

2.6.2 The flow tab………..35

2.7 Conceptual model………...…………..35 3. Methodology………..…………....37 3.1 Research strategy………..…………...…37 3.2 Research design………..……….37 3.3 Data collection………..………...…39 3.4 Data analysis………...……….41 3.5 Coding………..41 3.6 Coding analysis………...………….……43 3.7 Transferability………..………45 4. Results………...……….45

(3)

4.2 User behaviour………49

4.3 Exploitation of insights………...50

4.4 Personalization………....53

5. Discussion……….56

5.1 Implications………56

5.1.1 Implications for the significance of big data………..58

5.1.2 Implications for user behaviour……….59

5.1.3 Implications for the exploitation of insights………..60

5.1.4 Implications for personalization………....62

5.2 Comparison models………...64 5.3 Contribution………..65 5.4 Limitations………67 5.5 Future research……….67 6. Conclusion………..……...68 7. References………...………...71 8. Appendix………..…..80 8.1 Appendix 1………...…....80 8.2 Appendix 2………...80 8.3 Appendix 3………...81 8.4 Appendix 4………...81 8.5 Appendix 5………..84 8.6 Appendix 6………..86

(4)

Abstract

Market orientation and in particular market research which constitutes a major part of that field went through tremendous changes in the last years. Progress is made in terms of insights that are gathered by companies nowadays and notably, technology seems to be a factor that has influenced this factor extensively. This study aims to research on how a specific technology, namely big data analytics, is affecting market orientation or more specifically the exploitation of identified latent customer needs and how this process contributes to the phenomenon of personalization. A qualitative approach was chosen to investigate this matter. Therefore, we conducted structured interviews and identified big data analytics, user behaviour, exploitation of insights and personalization as the most central aspects of this research. Moreover, this research identifies an advanced form of personalisation namely, the discovery process in content streaming based industries. The study discusses the role of each aspect of the whole phenomenon of personalization and offers new perspectives on the role of each part and their interplay. This paper provides a foundation for further research and academic as well as managerial implications that can be utilised by managers to design a successful personalization process.

Keywords: Market orientation, market research, big data analytics, exploitation of identified latent customer needs, user behaviour, personalisation, discovery process

(5)

Acknowledgement

I would like to articulate my profound gratitude to Markus. Your guidance, supervision and feedback were very valuable to me in the writing process. Thank you very much for your support. Furthermore, I would like to thank all the Deezer employees, who volunteered to participate in the interviews and thereby enabled me to conduct this research.

Statement of originality

This document is written by Student Maximilian Bronowizki (11974842) who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(6)

Introduction

How do companies know what I would like to buy next? The internet gets more and more personalized, so that different people have individual consumption experiences. “Big data opens the windows to give us an in-depth view of the customer’s mind and where trends may bend toward in the future.” (Newman, 2015). The products consumers search for and purchase are juxtaposed to the consuming habits of millions of other customers. Companies such as Amazon build profiles out of our consumption habits and match us with products that we are likely to purchase (Marr, 2017). In today’s world, companies can easily get access to big data, since in many industries it is a by-product that the customers provide by making use of a product or service. The corporations use that data to learn more about our consumption behaviour. Big data is gathered, analysed and interpreted by businesses, so that they can utilize the interpretations derived from the analysed information to make strategic decisions or gain insides for the market research (McAfee & Brynjolfsson, 2012). This paper investigates whether and if yes how this technology changed the market research strategies up to this point. Firstly, this study looks at the technology in isolation and the most relevant characteristics of this technology (Russom, 2011). Afterwards, this technology is connected to the concept of market orientation. The market-oriented approach is a long-term oriented concept, and is focusing not solely on the expressed but also the latent customer demands (Slater & Narver 1998). This study focuses on a very specific and yet very important direction in the field of market orientation. More specifically, this research digs deeper into the companies’ efforts to identify latent customer demands and how that technology is involved in that process. Latent customer needs can be reformulated into demand that the customers themselves are not aware of. Consequently, identifying latent customer needs means knowing what your customers want before they do. But why would this be desirable for companies? What kind of value can they

(7)

create from this? One of the main answers to these questions is personalization. Humans like the feeling of “this product/ this service was made specifically for me”.

This is particularly true for the case of the music streaming industry. The case of the music streaming industry is helpful in order to analyse the phenomenon of interest at the firm-level and provide some context of an industry as a proxy for digital born businesses and compare that to best practices that were common before the digitalization of the market research departments. The music streaming industry is an industry that is highly affected by our main independent variable, namely technology or in this specific case big data. Hereby, we would like to detect specific developments in the field of market research and look at the specific mechanisms that dominate this field nowadays. The market research in this industry is one of the major sources of value creation, which makes absolutely sense because if this function is working well within a company and the service or product matches exactly the need of the consumer, that service or product also creates more value for that consumer. Particularly, the business of providing music content to a big number of users, which differ in their preferences and individual tastes, allows for massive value creation extracted from personalized offers. Consequently, we discuss the phenomenon of personalization as an outcome of the identification process of latent customer needs. After discussing the phenomenon of personalization in general, we will pay attention to a specific type of personalization, namely the discovery process. Discovery is establishing itself as a substantial part of personalized content providers’ services. However, it is a feature which is very recent and has not got a lot of academic attention in terms of research. Consequently, this investigation is located in the consumer research field and provides a better understanding of customer satisfaction and how it is achieved nowadays. Therefore, this study also touches upon the bigger context of the previously described concepts, so that we try to not only research the connection of big data analytics to the exploitation of identified customer needs and

(8)

different parts of the puzzle work together and if there are some parts that play a surprisingly big or small role in the researched phenomenon. It would also be desirable to identify new directions in personalization in order to provide completely new implications for scholars and managers.

Firstly, this paper commences by reviewing the existing literature containing all relevant concepts, which will lead to the research question of that study. Secondly, the theoretical framework and the research design is illustrated. Thirdly, the results will be demonstrated and afterwards discussed. Fourthly, the insights gained in the discussion part, including implications, further research suggestions and limitations of this paper will be emphasized in the conclusion.

Literature Review

The following section has the purpose of introducing the reader to the field of market-orientation and specify the research subject of this paper. We do this by analysing and evaluating the existing literature on this topic.

The marketing perspective regards the detection of consumer wants and needs as the company's purpose (Slater & Narver, 1998). The company, which is able to do that more effectively and efficiently than its competitors will acquire a competitive advantage. Scholars distinguish between two separate views here. On the one hand, there is the customer-led view, which represents a short-term approach of responding to expressed customer needs. On the other hand, there is the market-oriented approach that is more long-term oriented, and that is focusing not only on the expressed but also the latent customer demands (Slater & Narver 1998). In this research, we will focus more on the market-oriented approach and the identification of latent customer needs. The Oxford dictionary defines latent as "existing but not yet developed or

(9)

manifest", which consequently means that latent customer needs exist, but the customers themselves are not aware of them as they are concealed to some extent. Nevertheless, it is very significant to determine the latent needs as well. Market-oriented businesses acquire and interpret market information or data, not only to detect the previously mentioned customer needs but also to keep track of the competition's strategic moves. The purpose of this activity is to provide superior customer value and stay ahead of the competition (Slater & Narver, 1998). Consequently, market-orientation is a concept that provides value and the ability to stay competitive for companies in many ways. “Almost every company competes to some degree on the basis of continual innovation.” (Leonard & Rayport, P:103, 1997). That is the reason why managers want their companies to stay in close contact with their customers, so that the company can “listen” to what the customer wants next, if there is a change in the demand structure or not and many other insights that might be derived (Leonard & Rayport, 1997). Market-oriented companies “are better able to understand customer needs (and) they more readily adapt offering to meet changing preferences” (Hills & Sarin, P:14, 2003). Another scholar stresses that only the company which shapes markets, consumption behaviours and customer relationships proactively, is able to make optimal use of its own resources (Meyer, 2017). This insight underscores the significance of a market-oriented approach. Nonetheless, one has also to highlight that even though more and more companies become market-oriented, it is nowadays a cost of doing business rather than a source of competitive advantage (Kumar, Jones, Venkatesan & Leone, 2011). This is due to the fact that market orientation is a source of competitive advantage for early adopters and that the benefit decreases for the following vast majority (Kumar et al, 2011). The fact that it is a cost of doing business underscores that it is not a choice anymore between market-oriented and customer-led view but market orientation is mandatory (Kumar et al, 2011). Nonetheless, some scholars have a slightly different opinion

(10)

growth and low-competition environment (Slater & Narver, 1994). However, they also state that in the long-term businesses will feel the need to be market-driven and it is advisable to invest in a developing market-orientation (Slater & Narver, 1994).

The question remains how to identify the needs that the customers themselves are not aware of? In order to specify that more, we provide the first model of this paper. Hereby we make the assumption that there is a spectrum from unknown and unmet demand to known and met demand. This assumption goes hand in hand with an understanding that latent customer needs can be identified and thus the emphasis is on the tools and practices to reach that objective. As pointed out before this research investigates in the field of consumer research in order to get a better understanding of customer satisfaction. Therefore, we assume that companies make efforts to move in the right direction on that spectrum and that companies that are able to position themselves more to the right have a higher probability of achieving customer satisfaction, so that the companies' motivation is a prerequisite for the following model.

(11)

How was it done before digitalisation?

In the following section, we look at the past path that firms followed to determine latent customers’ needs in order to illustrate the developments in this area before we dig deeper into the practices that are common nowadays. Hereby, we discuss some popular practices that were established as best practices at certain points of time before digitalisation.

Notwithstanding, before further analysing the existing literature about market orientation we take a look at how consumer research was done before that and with which practices it was associated. In 1971, Bettman introduced a paradigm that aims to explain the structure of consumer choice processes. This paradigm illustrates how individuals make a first decision (which is not further explained in the model) and that environmental factors influence a possible perception of incongruity of expectations and delivered value, which leads to an active search or other forms of problem solving (Bettman, 1971). However, the author emphasizes that the attempt to put such a complicated process in a simplified model is incomplete and flawed at best, but it also shows first interest in the consumer research area (Bettman, 1971). This interest kept increasing, so that researchers commenced to ask multiple questions about consumer behaviour. For instance, another scholar investigated the buyers subjective price perceptions and questioned general basic economic models such as the inverse price-demand relationship (Monroe, 1973). This additionally illustrates the researchers’ interest in more specific insights about consumers, their perceptions and needs. Consumer research is interested in the purchasing decisions of individuals and in the past this research was a difficult and painful process, which was primarily conducted in the form of continuous surveys (West, 1974). These are surveys, which gathered information in regular time intervals from a stable but also quite limited sample. There were several objectives that the market researchers intended to achieve

(12)

The insights that researchers were focused on at that time are clearly more general insights than what market research is trying to accomplish nowadays. What is striking here is that amongst those aims and many others there was also the aim to forecast trends or seek new products. Although we can ascertain that market researchers in the seventies must have already thought of similar constructs such as latent consumer needs, there was clearly a lack of attention paid to this concept (West, 1974). In the discussion part of his paper, West (1974) acknowledges that market research faces several issues in some areas and that one of the most problematic areas is the new product research, which is deeply connected to a central phenomenon of this paper: the latent consumer needs.

The following paragraphs introduce two central literature pieces that represent the state of the art within each time frame. The aim is to assess how the literature and best practices have evolved in that time. Both articles are well-cited, published in renowned magazines and used by later scholars as bases for their research.

Consequently, we move on to the first piece, which is an early paper that cogitated firstly about market orientation, market intelligence and latent consumer needs in detail. We also have to state that this field of research did not get a lot of attention in the recent years as the model proposed by the authors from 1990 is still one of the most significant models in this field. The first key article for this paper’s research identifies three most significant pillars of market orientation, namely, customer focus, coordinated marketing and profitability. Those pillars represent general concepts that have to be integrated profoundly into the company’s culture and the way it operates (Kohli & Jaworski, 1990). Customer focus involves decision making on the basis of market intelligence. Whereas market intelligence can be defined as taking market factors into consideration, which affect customers’ preferences and needs, not only current but also future. This highlights the fact that back then researchers thought in terms of whole markets in the market-oriented view so that none or only minimal market segmentation took place. Coordinated marketing is a concept that emphasizes “that a market orientation is not solely the

(13)

responsibility of the marketing department.” (Kohli & Jaworski, p: 3, 1990). Finally, profitability is identified by the researchers as the consequence of a market orientation (Kohli & Jaworski, 1990).

Furthermore, the authors identify three tasks that have to be fulfilled by a company that is engaging in a market-oriented approach. Firstly, there is intelligence generation, which is the process of gathering relevant information and is the most important part for this paper. There are multiple mechanisms through which the desired information can be obtained. "The mechanisms include meetings and discussions with customers and trade partners (…), analysis of sales reports, analysis of worldwide customer databases (…), sales response in test markets and so on." (Kohli & Jaworski, p: 4, 1990). After the intelligence generation, the next step is the intelligence dissemination. This concept deals with the necessity that multiple if not all departments of a company are involved and informed so that the firm can effectively respond to the market. The last element is the responsiveness to the insights or the intelligence that was first generated and then disseminated, consequently, the action taking (Kohli & Jaworski, 1990). In addition to that, the authors proposed a conceptual framework, which formed the foundation for many market orientation studies afterwards and can be seen in the following figure.

(14)

This framework illustrates three types of antecedents to a market orientation (Kohli & Jaworski, 1990). There are the senior management factors, which represent the substantial impact that senior managers have on the organization and its market orientation. Moreover, there are the interdepartmental dynamics, which epitomize the formal as well as informal relationships and interdependencies between departments, so that it is closely related to the previously explained concept of intelligence dissemination. Other scholars that labelled that concept as cross-functional integration confirm the meaningfulness of interdepartmental dynamics and its effect on a company's market orientation (Im & Workman, 2004). The last category of antecedents are the organizational systems, which stand for the organizational structure. For instance, whether a firm is centralized in its decision making or the degree of formalization etcetera (Kohli & Jaworski, 1990). These three types of antecedents lead to a market-orientation according to the authors. Otherwise they also identify supply-side and demand-side moderators, which influence the relationship of market-orientation to general identified consequences such as customer responses, business performance and employee responses (Kohli & Jaworski, 1990). Supply-side moderators refer to the nature of competition among suppliers while

(15)

demand-side factors are defined by the structure of demand in an industry through for instance consumer preferences (Kohli & Jaworski, 1990).

This piece of literature provides an insightful view on the methods of identification of latent customer needs and also on the factors that the scholars researched as significant for a market orientation. However, the two most relevant insight are that firstly in 1990 technology was not regarded as a relevant variable for a market orientation and thus was not included in the conceptual framework at all and secondly, that such concepts as consumer preferences were identified as demand-side moderators and were left in a black-box without further explanation rather than putting the identification of the demand-side moderators in the centre of the model.

In 1997 the answer was to make use of empathic design according to a well cited article from the Harvard Business Review. Although, this method should not be seen as the only solution at that time, it is a good representative of a development in this field. The methods introduced previously lacked an integrative approach, whereas empathic design is described as a technique that integrates the process of gathering, analysing and applying the information in the observed target market (Leonard & Rayport, 1997). This approach is different from the old-fashioned continuous surveys that were explained previously because it emphasizes an integrated view of such activities as information acquiring or interpreting and does not look at for instance information gathering in isolation (Leonard & Rayport, 1997). The first step of this process is to make observations to catch the unfiltered reaction of the customers; hereby the researchers would pay attention to details such as consumer's hesitation to purchase some item in a particular situation. The second step is that of capturing data with the help of photographs or videos in order to analyse details that might go unnoticed if not recorded in some way. The third step is the reflecting and analysing, at this point the researchers try to determine all

(16)

develop prototypes of solutions and test those prototypes with some kind of sample (Leonard & Rayport, 1997).

After analysing two literature pieces from 1990 and 1997, we can extrapolate some developments in our research area. It can be stated that there has been a development towards more reliance on technology and more specifically data as firstly the customer database was the only channel to extract information from and in contrast to that multiple databases play a central role in the empathic design method. Moreover, we can observe a development towards observing rather than asking or interviewing, so that the customers or other stakeholder groups cannot filter easily which information to share with the researchers. Another development that we can detect are the steps towards an integrative approach. Already in 1990, the importance of intelligence dissemination was highlighted, however, the empathic design approach puts the integration aspect at its core by coalescing distinct tasks into one method. By doing that this approach creates overlapping responsibilities for different departments of the firm and "forces" them to collaborate more intensely. Notwithstanding, it can be punctuated that although there is a development towards more reliance on data, the analysed literature presents a very limited diversity of data sources. However, the most significant development seems to be the role that technology plays in the market research process so that the next section will focus on the technology of interest of this paper.

Big data

In the following section, we introduce the concept of big data as it is assumed to be a central variable in the progress that was made in the field of market research and we need to understand its nature and its potential.

(17)

Big data is a concept that does not have a single, precise definition. In this research, we define big data as significant amounts of digital data that managers can measure and thus make use of when making a strategic decision. Big data is characterized by the 5 Vs (See Figure 1), which are explained in the following paragraph.

The Five Vs Volume

The first is the volume, which is the size of the data. Already in 2012 2.5 Exabyte of data were created each day (McAfee & Brynjolfsson, 2012). It is obviously the primary characteristic of big data (Russom, 2011). Furthermore, the volume of big data is continually increasing for more than 2 decades now (McAfee & Brynjolfsson, 2012).

Velocity

The second one is the velocity, which is the speed at which that data is generated. Another way of thinking of it is as “the frequency of data generation and/or frequency of data delivery" (Wamba, Akter, Edwards, Chopin & Gnanzou, p: 236, 2015). Due to the availability of high volumes of data, the need for companies emerged to react quickly to the insights generated from big data analytics. In other words, with data coming at you permanently in real-time,

(18)

The third V is the variety, and it is connected to the different types of data. As mentioned before there is no universal definition for big data, so that big data includes different kinds amongst others traditional data as well (Lee, 2017). Nowadays, data can be retrieved from multiple different sources such as web sources or social media (Russom, 2011). Clearly, this heterogeneity adds to the complexity of this technology. So that, a natural tension can be acknowledged between the velocity and the variety attributes of big data.

Veracity

The fourth V represents the veracity or in other words how accurate and reliable the data is. This attribute deals with the uncertainty or ambiguity of big data. When data is acquired, it is in an unstructured state, which in turn highlights the significance of the analytical work performed by the company (Wamba et al, 2015). Therefore, it is substantial that the structuring and the analytical work, in general, are able to mitigate the ambiguity and uncertainty of big data in its initial state.

Value

Finally, the last V is the value of the data, and it was introduced by Oracle, and it means that firms have to understand how to extract value out of big data (Lee, 2017). Basically, this is the extent to which insights that are generated through big data analytics will be exploited and transformed in order to create value for the firm (Wamba et al, 2015). This is also the attribute, which is profoundly connected to the phenomenon researched in this paper.

Controversy around big data

The concept of big data involves much ambiguity as it does not have a general commonly accepted definition and there are elements in these concepts that make people uncertain whether to consider this technology a thread or an opportunity. A concept that is associated with such

(19)

an ambiguity usually is also connected to many controversies. In the previous paragraphs, we have introduced multiple characteristics that define big data, now we induct another one that is more specific to the phenomenon researched in this paper. “Big Data is commonly defined as the process whereby computers sift through enormous quantities of data to identify patterns that can predict individual’s future behavior.” (Porat & Strahilevitz, p: 1434, 2013). As the spending on IT and specifically big data is constantly increasing, so that companies spent $28 billion in 2012 on big data (Porat & Strahilevitz, 2013), many scholars and practitioners wondered whether this technology itself could be the source of competitive advantage. “Big data isn’t new, but the effective analytical leveraging of big data is.” (Russom, p:7, 2011). Consequently, we assume that not big data itself is a valuable resource but the exploitation process of the insights generated through big data. Furthermore, some others doubt the usefulness of information technology (IT) in general, which also includes the big data technology, as IT gains more characteristics of a commodity with constantly decreasing prices and increasing availability. Consequently, no firm can differentiate itself from the competition through the usage of IT or big data (Carr, 2003). Nevertheless, we also have to emphasize that this critical perspective is rather unconventional and that most of the scholars agree on the usefulness of IT and big data.

However, the relevant question here is how is that technology connected to the phenomenon that we are interested in? Big data led to the emergence of personalized search engines and targeted online marketing. Even very personal information such as the sexual orientation can be derived from the internet (Kosinski, Stillwell & Graepel, 2013). Big data did not only change the way consumer information is obtained but also added a variety of potential information sources. For instance, the social media platforms such as Facebook are considered to be great

(20)

sources nowadays. These amounts of information make personalized customer experiences possible and can be seen as a tool to improved market segmentation (Pridmor & Hämäläinen, 2017). Moreover, it is important to mention here that the information the companies gather is not filtered by the consumers, but the information is given "involuntarily", which makes it less biased. To sum it up, big data gives companies the opportunity to collect vast amounts of unbiased information about our preferences and consumption patterns. We assume that this is a very valuable asset that assists the firms in determining latent customer needs and thus achieve a higher customer satisfaction indirectly. However, we also have to acknowledge the fact that some scholars doubt the role of big data analytics as a source of competitive advantage and that some even doubt the usefulness of the technology in general.

The Role of Technology

The next section deals with how technology or more specifically big data analytics fits in the framework of market orientation and the identification of latent customer needs.

Previously, we made the assumption that latent customer needs are on one scale and can be pushed from unknown and unmet demand to known and met demand. Moreover, we underscored that we expect companies that are able to position themselves on the right side of the spectrum to perform better in terms of customer satisfaction. Now we propose the next model of this paper, which illustrates that we expect technology to have a positive effect on this process, hereby we introduce a variable, namely technology, that was utterly absent in the model of Kohli and Jaworski (Appendix 1). This positive effect is expected to be manifested in the form of quality and velocity. Meaning that companies are able to get valuable insights with the assistance of this technology that would not be discovered without it and also that big data is expected to shorten the period from identifying latent customer needs to exploiting that knowledge. In other words, we identified a major variable that was not acknowledged in the

(21)

existing literature on this topic, and which is expected to influence the process of interest profoundly.

Nonetheless, we also have to mention the natural tension between those two variables. Processes that are said to have a high quality, whereas high quality is defined as adhering to an excellent standard of measure, often sacrifice velocity for it and vice versa. Although this paper will not focus on this tension, we would like to acknowledge it. Moreover, this research is focused on the quality part or more specifically on better knowledge of existing customers, so that velocity takes only a minor role here. This means that our expectation is that technology enhances particularly the knowledge that companies have of their existing customers. However, we will dig deeper into that phenomenon when we propose the conceptual framework of this paper.

Model 2

In order to specify our area of interest more, we will focus on the exploitation process of the identified latent customer demand. In other words, we will look at how technology enhances the exploitation process once the latent consumer needs are identified and consequently on the

(22)

assimilate, transform and exploit knowledge (Zahra & George, 2002). Nonetheless, we will only pay attention to the fourth role of this concept, namely the exploitation of knowledge as the other three dimensions are more connected to exploration of knowledge than exploitation (Van den Bosch, Van Wijk & Volberda, 2003). The ability to exploit knowledge that has been gathered is a crucial element of innovative capabilities (Cohen & Levinthal, 2000). The authors also stress that in order to be able to exploit knowledge the company has to invest internally in that activity as it will not just come as a by-product of something else (Cohen & Levinthal, 2000). This indicates that the exploitation of gathered knowledge is a challenging task. Some authors specify that exploiting knowledge is “based on the routines that allow firms to refine, extend, and leverage existing competencies or to create new ones by incorporating acquired and transformed knowledge into its operations.” (Zahra & George, p: 190, 2002). Consequently, these scholars put the emphasis on routines and their effect of providing system and structure. This leads to this paper's next working proposition. The usage of big data as a systematic routine enhances the exploitation of the identified latent consumer needs, which is a valuable insight as the discovery process of latent customer needs is expected to be a non-continuous change process, and big data is theorized to bring structure and a systematic approach to this process.

However, we are not only interested in whether big data improves the exploitation process of the identified latent customer needs but also in which way it does. Thus, we are also interested in the consequences of the usage of that technology. Yli-Renko, Autio and Sapienza conducted a research which thematises the knowledge acquisition and knowledge exploitation processes of young technology-based firms. The empirical findings of these scholars underscore that "young technology-based firms that acquired greater market and technological knowledge through their key customer relationships (…) developed greater technological distinctiveness" (Yli-Renko et al, p:608, 2001). This insight is applicable to the music streaming industry, which will be introduced at a later point in this paper, in a sense that the companies in this industry

(23)

make use of big data analytics in order to acquire knowledge about their users and exploit it afterwards to achieve greater distinctiveness. That distinctiveness is manifested for instance in the form of a personalized service. Although the previously mentioned distinctiveness can take multiple forms, this paper focuses on only one major form, personalization. Therefore, we analyze some relevant literature pieces about the personalization process.

Personalization

Personalization seems to be a concept that gained much attention in the recent decade. On the one hand, the number of scientific papers that focus on personalization or specific aspects of personalization has increased dramatically in the recent years (Tseng & Piller, 2011). On the other hand, one can observe the significance of this concept in many companies and their product or service offerings. For instance, companies such as sportswear giant Nike offers their customers the possibility to customize their shoes (“Nike Store“, (n.d.)).

However, "personalization is only possible if reliable, and projectable customer data are available" (Arora, Dreze, Ghose, Hess, Ivengar, Jing, Sajeesh, 2008, p: 318). Connecting this to the insight we previously gained from the literature piece about absorptive capacity, we can underline that in order to establish a routine the company needs reliable and stable inputs which is in this particular case customer data. An essential prerequisite for many forms of personalization is the availability of user profiles that contain personal preferences and interests of the user (Ferman, Errico, Beek & Sezan, 2002). Ferman et al introduced a model that illustrates user information is used for the purpose of personalization.

(24)

You can see user data appearing in multiple points in the framework in form of user preferences, usage history or usage history, so that, user data is a necessary ingredient when engaging in the process of personalization, and we will come back to this aspect when evaluating the results of our research. But for now, we can state that content in the form of data about user behavior and technology are the two necessary ingredients for personalization attempts (Kaptein & Parvinen, 2015).

Personalization is the most ultimate form of segmentation and therefore very difficult to achieve since the company is working on a one-to-one basis (Arora et al, 2008). That process enables the firm to decide, with the support of previously collected customer data, how the specific service should look for each customer (Arora et al, 2008). Researchers noticed that "personalization proved to add value and was therefore worth pursuing" (Van Velsen, Van der Geest & Steehouder, p:183, 2010), which also makes sense that a one-to-one relationship to customers is valued by the customer and therefore crucial in the value-creation process. There are many studies conducted that confirm that. For instance, one scholar tested how web personalization affected the user assessment of the website value and came to the conclusion that personalization influenced the outcome significantly (Benlian, 2015). The value of personalization is also affirmed by some other studies that identify a clear connection between personalization and customer loyalty (Thirumalai & Sinha, 2013). Nonetheless, the authors also

(25)

emphasize that the value added through personalization is context dependent so that a personalized service or product might add more value in one industry than in another (Thirumalai & Sinha, 2013). It seems that this process gained more importance and attention from scholars. This development is expected to be facilitated by technology (Thirumalai & Sinha, 2013).

The next paragraph aims to provide a short general overview in the field of personalization. Some scholars present three identified types of personalization. Firstly, there is personalization in the form of recommendations. This type has been the most widely used form of personalization, and it is also the most relevant type for this research since the proxy that we will use is a personalized recommendation list of songs, namely the "Flow Tab". Secondly, there is personalization in the form of guidance and orientation, which involves providing a personal path through the system. Notwithstanding, this kind of personalization has not been used extensively by the market yet (Van Velsen et al, 2010). Thirdly, there is the last type of personalization, namely the personal views and spaces. This type can be illustrated by personalized home pages or personally relevant content. A straightforward example of this type of personalization is a mobile application that shows you your personal mobile data usage for the past period (Van Velsen et al, 2010). However, all of these three types can be categorized as adaptive personalization types, because the company has to adapt the service based on data that has been observed about the previous customer behavior (Chung, Wedel & Rust, 2016).

The next paragraph has the objective to shortly explain the difference between personalization and customization as these two concepts are used often as synonyms even though they are not interchangeable. For this purpose, we would like to present a model that was introduced by Arora et al in 2008.

(26)

This figure emphasizes that customization is a process that is customer initiated, so that for instance customers have the ability to design their Dell computers according to their needs and wants (Arora et al, 2008). Hereby, the company just has to provide the necessary means to the customers to do it themselves. In contrast to that, personalization is a process that is initiated by the company (Arora et al, 2008). Consequently, personalization requires way more work in terms of market intelligence as more information has to be gathered and analyzed. This is the reason why personalization is more important for this research, because the identification of latent customer needs is way more significant in the context of personalization than for customization.

The Bigger Picture

In order to emphasize the significance of these concepts as elements that are part of a bigger chain that eventually leads to a better overall performance (in form of higher profits) and provide more context to the reader, the next section will thematise the contextual factors, in which the phenomena of interest are embedded. The concepts that are introduced in this section are very wide so that we cannot perform an in-depth analysis of all of them. The objective of this section is instead to underline the connections between the different elements and therefore

(27)

addresses the concepts simplistically.

There are multiple reasons why companies invest much time, money and efforts in order to establish a one-to-one relationship with their customers. One of the major reasons is that personalization enhances companies’ ability to meet the consumer expectations and needs, which results in increased customer satisfaction and ultimately customer loyalty (Riemer & Totz, 2003; Kompella, 2017). As this research has the objective to investigate further in the field of consumer research, this is a very significant insight to which we will come back in a later stage of this paper. Moreover, personalization does not only potentially increase customer satisfaction but also creates so-called lock-in effects, because "individualization of products and services decreases product comparability” (Riemer & Totz, p: 1, 2003). Personalized products and services are more difficult to compare to similar services; consequently, personalization increases the switching costs for the customers (Riemer & Totz, 2003). Switching costs are costs that the customer associates with changing their current service provider or current supplier. These costs can be economical, however also psychological, since the customer has to put time and mental effort into finding a new service or product and afterwards getting used to the new service or product. Furthermore, these costs are maximized when the company is able to meet the customer's needs and expectations in a personal manner (Riemer & Totz, 2003). To sum it up, it can be emphasized that personalization is expected to increase customer satisfaction on the one hand and increase switching costs on the other hand

(Kompella, 2017).

In order to get a better understanding of the dynamics between the concepts that were touched upon, we take a deeper look at switching costs. There are three types of switching costs: transaction costs, learning costs and contract costs (Klemperer, 1987). Transaction costs refer

(28)

another. Learning costs represent the time and the effort of learning about a new product or service. Lastly, contract costs are associated with for instance penalties for quitting a contract when it's still premature (Klemperer, 1987). We identified learning costs like the ones that are most applicable to the music streaming industry.

Given that higher switching costs and increased customer satisfaction are the outcomes of personalization, the next step would be to investigate how these outcomes influence customer loyalty. On the one hand, customer satisfaction is expected to lead to customer loyalty, because the higher the perceived value of the service or product provided by the company, the unlikelier it is for the customer to get a better similar service or product by another provider (Yang & Peterson, 2004). On the other hand, switching costs have only a moderating effect under the condition that the customer satisfaction is above average (Yang & Peterson, 2004). This can be analyzed from the cost-benefit perspective. If the customer satisfaction is above average, then the switching costs have to be subtracted from the benefit of switching to another provider and thus decreasing the net utility from the switching action (Yang & Peterson, 2004). If the customer satisfaction is not above average, then the customers might perceive that their opportunity costs are high and therefore higher switching costs do not lead to higher customer loyalty. However, applying these insights to this particular study leads us to the expectation, that since personalization increases customer satisfaction, both identified consequences of personalization, customer satisfaction and higher switching prices will influence customer loyalty (Kompella, 2017). Customer loyalty is directly connected to higher profits because it is very costly for companies to gain new customers so that keeping your current customers becomes a major condition for performing well. "The companies with the highest retention rates (…) also earn the best profits." (Reichheld, Teal & Smith, p:135, 1996).

Higher profits result in more resources and more resources lead to higher budgets for different functions such as marketing or IT spending. As IT is regarded by the companies as a very

(29)

critical resource, this perception is also reflected in their spending habits on IT (Carr, 2003). Consequently, higher profits result indirectly in more IT spending or more specifically in bigger investments in big data, which gives the whole phenomenon a circular structure. Apparently, this is a simplified chain of connections as the budget for a specific function depends on many things such as the performance of the specific unit (Otley, 1978), also higher IT spending can occur without higher profits as a prerequisite, so that the following model only depicts one of many possible structures and solely is here for illustration purposes.

The grey elements in the model represent the concepts of interest, whereas the black units are utilized in order to shed light on the bigger picture, in which big data analytics, the exploitation of identified latent customer needs and personalization are embedded.

(30)

The Music Streaming Industry

In the succeeding paragraph, we interpose the music streaming industry, because we will use it as our industry of interest so that future theoretical insights can be connected to that industry immediately and ideally also to some similar industries.

Digitalization caused major changes for the music industry, which led to entirely new music distribution channels and thus also wholly new business models (Wlömert & Papies, 2016). Companies such as Spotify which was formed in October 2006 and Deezer which was founded in August 2007, provide access to a comprehensive library for customers who pay a subscription fee on a monthly basis (“Deezer”, (n.d.)). “Music-streaming services encompass aggregative features that invite participation and enable listeners to perform as content curators of their music consumption.” (Hagen, 2015). This is the main reason for choosing this industry as the context for this research, because one distinctive feature of this industry is the active participation of users. This means that apart from being a technology-driven industry, it is also an industry, in which personalization is expected to be a very relevant concept.

The previously mentioned service providers are so-called paid streaming services (PSS). There are also free streaming services (FSS), which do not demand a monthly subscription fee but derive revenues from advertisement. Moreover, there are also hybrid models as for instance Deezer, which have multiple revenue streams (Wlömert & Papies, 2016). There have been many discussions about whether streaming services add value to the music industry or rather diminish value through channel cannibalization (Wlömert & Papies, 2016). One research accentuates that there is an overall "positive net effect for paid streaming services on revenue" (Wlömert & Papies, p: 315, 2016). On the one hand, the research clearly identifies that PSS substitute different channels. On the other hand, the authors point out that the revenue generation and the steadily increasing market for PSS result in a positive net effect for the music industry (Wlömert & Papies, 2016). Furthermore, the music industry experienced in the late

(31)

1990s and 2000s a significant declination in revenues due to digital piracy (Thomes, 2013). The music streaming services can be considered as a legal alternative that mitigated digital piracy and led to an increase in revenues for the entire music industry (Dow & Kuiper, 2013).

The music streaming industry has experienced tremendous developments in the recent decade. The big players in this industry such as for instance Deezer went from providing limited content and constrained availability to almost indefinite and diversified content as well as unconstrained availability (“Deezer plans to launch”, 2012). Moreover, these service providers did not only develop in this field but also numerous others, which bestow them platform character. In other words, nowadays many of the music streaming service providers also offer other services on their apps such as access to exclusive content like interviews with your favourite artists or tailored information about their recent concert tour dates.

Spotify, which was the first PSS, experiences the recent years exponential growth. The company announced in January 2018 that it has reached 70 million subscribers (See Figure 3) and was valued in 2017 as worth 19 billion $ (Reuters, 2018). On the 4th of April Spotify went public and commenced with a stock price of 165.90$. Since then it peaked at 170$ and is currently (07.05.2018) at approximately 155$ per share, which is associated with a market capitalization of approximately 30 billion $ (“Spotify Technology”,2018). This kind of growth is responsible for the positive net effect. However other distribution channels such as retailers of physical music units have to expect a "declining relevance and less revenue" (Wlömert & Papies, p: 325, 2016). Also, Deezer experienced exponential growth on a smaller scale in the recent years (See Figure 2). The figures in this section show that development. It can also be seen from these figures that Spotify operates on a bigger scale than Deezer. This can be partially connected to the concept of first mover advantage, since Spotify was the first music streaming

(32)

(Midia, 2017)

As the first music streaming provider, Spotify separated itself from all the other available options by providing an extensive music library of more than 20 million tracks (Dow & Kuiper, 2013). A prevalent feature of Spotify are the personalized playlists and recommendation engines such as the “Discover Weekly”. This is a playlist, which is updated every Monday and contains thirty songs that the specific user did not hear before but will like with a high probability (Heath, 2015). In other words, Spotify creates every week personalized playlists for millions of users and the majority of the users are impressed by the choices that Spotify makes (Heath, 2015). Although, Spotify was the first music streaming service provider to develop such discover playlists, other players such as Deezer have invested much time and financial resources to develop discover playlists features and enhance this element continuously. There will be more information on that in the next section.

In this paper, we will look further into these features of the music streaming applications and we will treat them as a proxy of the company’s efforts to identify and satisfy the customers’ latent needs and preferences.

(33)

Figure 2 (Deezer subscriber numbers):

(https://www.statista.com/statistics/321559/deezer-paying-subscribers/)

(34)

Deezer

From the few players that are in this industry, we chose to target Deezer as our primary company of interest. There are multiple reasons for that. Firstly, Deezer aims to be the most personalized music streaming service in the market (“Soundplate”, 2018) .Deezer introduced recently new features to the application that stress that however this will be discussed in a second (Deahl, 2018) .As the founder of Deezer, Daniel Marhely, emphasizes "To launch a product, the most important thing is that it answers a personal need." (“Deezer”, (n.d.)). Secondly, the availability of these companies and their employees is very limited or almost absent. Thirdly, I am a year-long Deezer user myself, so that I am very familiar with the product, which makes investigating the specific phenomenon that this research is interested in, easier and more profound. Deezer is a French company which was founded in August 2007 and has grown to have 14 million active users in April 2018 (“Deezer”, (n.d.)). The streaming service is in 182 countries available and contains over 53 million different titles and over 100 million playlists that have been created by the users or Deezer editors (“Deezer”, (n.d.)). As mentioned before Deezer aims to be the most personalized music streaming service provider, this is underscored by the investments in features that enhance the personalized user experience but also by the way of thinking or the corporate culture at Deezer. For instance, Deezer published recently an article, in which it is explained how Deezer “Visualize(s) your music DNA with Data” (Thary, 2018). This article emphasizes how Deezer makes use of the user's listening history and categorizes the music preferences of the user systematically by genre, daytime or other aspects that help to recognize a pattern in the user's listening habits. This is just a minor example that illustrates Deezer's objective of providing a personalized service and other bigger features and adjustments have also been launched recently.

(35)

The Flow Tab

Coincidentally, Deezer just launched at the end of April 2018 a new feature called Flow tab. There was already a feature called Flow, which was responsible for creating an endless stream of user's favourite songs and additionally new tracks that are picked by Deezer. On the company's website, it is described as a flow that "is a curated endless stream of your favourite songs plus new recommendations picked just for you" (“Deezer”, (n.d.)). The Flow tab, however, gives the user the opportunity to listen to personalized track lists which are based on the individual music preferences. This means that users are now able to listen to several tracklists that mark the artist name that they are inspired by (Deahl, 2018). By providing multiple discover playlists instead of only one, Deezer ensures to offer more specific playlists, so that users do not only get a recommendation list per se, but they can categorize it further down to recommendations for a specific genre they would like to discover at the moment (“Deezer’s new Flow tab”, 2018). We can consider this as a further investment in the enhancement of personalized content. Consequently, we will use this new feature as a proxy for the phenomenon that this research is focused on.

Conceptual Model

After highlighting the context and the relevance of our phenomena of interest, we would like to introduce the conceptual framework of this paper and narrow the focus down.

After we have made the assumption that unknown latent customer needs are on a spectrum with known latent customer needs and therefore can be pushed towards identification, we made a second assumption that technology or big data enhances this process primary in terms of quality

(36)

big data analytics, which was represented in the second model as the moderator, as our routine and the exploitation, which happened at the right end of the spectrum in the second model, as the process of our interest. Lastly, we underscored that in order to better analyze the role of big data and the exploitation process, we also need to identify the main consequence of this routine and the following process. Thus, we introduced the concept of personalization as our major consequence, although this phenomenon is expected to have multiple consequences, and discussed the contextual factors of these concepts briefly. All these insights lead to the third model of this paper.

Conceptual framework

Therefore, we formulate the central research question of this paper.

How does big data analytics contribute to the exploitation of identified latent customer needs and in which way does that enhance the company’s ability to provide a personalized product

(37)

Methodology

Research Strategy

In order to ensure a certain quality standard for this research, a research strategy is designed. This research is characterized by the qualitative research approach in order to gather an in-depth understanding of the variables presented in the conceptual framework of this paper and their relationship to each other. The main objective is to understand how big data influences the exploitation of identified latent customer needs and how that affects the personalization process. In order to understand the highly complicated contextual factors that are associated with big data and personalization and answer the “why” and “how” questions, qualitative research is the best option to fulfil that task and exhibit the complexity of the phenomenon of interest (Miles, Huberman & Saldana, 2003).

Research Design

This paper contains a qualitative research approach. The preferred method of qualitative research are multiple interviews with employees in the music streaming industry; consequently, this paper will make use of a single case study, therefore we make use of a holistic approach as we have only one unit of analysis (Yin, 2003). Case studies can be explanatory, exploratory or descriptive. Explanatory case studies are utilized in order to explain causal relationships. In contrast to that, exploratory studies fit researches that contain a not clearly defined problem. Descriptive studies are simply describing problems or processes that are taking place. Consequently, this study is exploratory since we mentioned previously that there has been paid only a limited amount of attention to the phenomenon that is researched in this paper. The case study approach makes sense if certain conditions are given that enable the researcher to examine

(38)

cannot be manipulated (Baxter & Jack, 2008). Third, the contextual conditions are considered as relevant and will be covered in the paper. Finally, the demarcations are not evident between context and phenomenon (Baxter & Jack, 2008).

The objective was to realise interviews with one of the big players in the music streaming industry. However, it is very important here to stress that the music streaming industry turned out to be a rather closed industry, which made this research a lot more complicated. Nonetheless, we were able to build contact to one company that is considered to be one of the major players in this industry, namely, Deezer. Although, employees of only one company were interviewed, we can regard the insights as industry-wide given the limited number of players in this industry and the geographical dispersion of the participants. The interviews were conducted with employees from different management levels. We developed three different versions of interviews. In order to ensure that this research is providing a multidimensional view on the topic at hand, some participants contributed to this research from the business/ management side, some from the purely technical side and finally some that emphasize the consumer perspective and therefore the value-adding aspects of the researched phenomenon. One version aims to research the senior management’s view on the matter at hand, the second version’s objective is to investigate the data analysts’ point of view and the third one is intended to look into the perspective of marketing/ product management department. Moreover, two interviews are executed with senior management representatives in order to hear the overall perspective to the researched phenomenon and therefore analyse the relationships that were introduced in the big picture part. Four interviews were conducted with software engineers, data scientists or data analysts in order to gain insights from the people who are connected the most to this technology and a specific part of the researched phenomenon. Nonetheless, these participants were asked to formulate their answers in a way that enables readers with limited technological expertise to understand the line of argumentation. Moreover, eight interviews were executed with product, project, marketing and content managers. This represents the majority of the interviews as this

(39)

group was identified to be the most qualified to provide insights from the business perspective on the researched phenomenon. The minimum number of interviews was planned to be not less than ten in order to guarantee that the research allows for triangulation and indeed at the end there were fourteen interviews conducted in total. The interviews were conducted with Google form as a tool due to the fact that the music industry is a rather closed one. This allowed us to execute multiple structured interviews with the option to ask follow up questions later on via email, so that we can consider the whole interview process as a semi-structured one (Saunders & Lewis, 2014). These interviews were performed with several employees of Deezer, who I approached through various channels whereas LinkedIn is my primary communication channel to the respondents. The interviewees were selected based on the position that they own in the company. These participants were selected based on their extensive expertise and experience with big data analytics, the transformation of insights coming from big data analytics into strategic decisions and personalization as a whole. The respondents were required have at least have two or more years of working experience at Deezer.

Data Collection

Fourteen interviews were conducted, and all of them are made use of for this research. Since Deezer is not physically present in the Netherlands and the limited availability of the participants, as mentioned previously the interviews were conducted mainly via Google form and partly via email. The interviews are characterized by open questions in order to ensure that participants can express their perspectives and provide in-depth insights about the phenomenon of interest. This is expected to lead to high-quality answers and therefore deeper insights (Horton, Macve & Struyven, 2004). All of the interviews were conducted in English, whereas

(40)

United Kingdom. The following table provides a comprehensive overview of the performed interviews.

Description of Participants

Interview title Job title Years of experience (In the company) Data Analyst Data Scientist 2

Data Analyst Data & Research Officer

4 Marketing/ Product

Management

Product Manager 2 Data Analyst Software Engineer 3 Marketing/ Product

Management

Content Manager 5 Senior Management Senior Customer

Relationship Manager 8 Marketing/ Product Management Project Manager 3 Marketing/ Product Management

Marketing & Sales Manager

4 Marketing/ Product

Management

Product Manager 5 Data Analyst Software Engineer 5

(41)

Senior Management Senior Product Manager 7 Marketing/ Product Management Marketing Manager 4 Marketing/ Product Management Product Manager 2 Marketing/ Product Management Project Manager 5 Data Analysis

After finalizing the data collection process, the next step is to analyse the gathered data in order to show the resulting findings. Due to the fact that the interviews were conducted via Google form and via email, the transcribing process is less complicated than with audio-recorded interviews. After the data is processed into the suitable format, the further step is to commence with the coding. Coding is the process in which certain data fractions are matched to efficient data-labels. This process is done in order to make the structured identification of factors that are relevant for answering the research question of this paper possible.

Coding

This process requires us to utilize deductive and inductive research methods. In order to do that we make use of Nvivo 12, which is an analysis program that supports us in identifying patterns in the gathered data (Bazeley & Jackson, 2013). For this purpose, we associate statements about specific topics with codes. A code is a ‘'qualitative inquiry is most often a word or short phrase that symbolically assigns a summative, salient, essence-capturing, and/or evocative attribute for a portion of language-based or visual data'' (Saldana, 2012). In the beginning, the codes are very close to the actual statements of the interview participants, which implies an inductive approach (Saunders & Lewis, 2014). Once relevant statements are structured under the

(42)

data

bi

g

pe

rs

on

al

iz

at

io

n

de

ez

er

analytics

industry

us

er

experience

process yes business company in si gh ts cust om er department users research ne w w it hi n m us ic market playlists te ch no lo gy va lu e co m pe ti ti ve ad va nt ag e cost discover content example ge t one decisions to ol us e ac ti vi ti es able focus in fo rm at io n no w he lp in st an ce make much al so pr oj ec ts st ra te gi c analysis better fo rm long product service behavior co m pe ti ti on personalized re ga rd su pe ri or usage ye ar s ad di ng challenges create developed important kp is talk us ed coming comparison customers fe ed ba ck flow model analysed collected m an y tab tr an sf or m in g va lu ab le worked according app co nn ec ti on co re en ha nc e ex pl oi ta ti on m an ag em en t needs st re am in g w or k co lle ag ue s' ha pp en te ch no lo gi es w ay add bda dr iv er s in sp ir ed main

under the categories presented in the conceptual framework of this paper and look for causal relationships (Miles & Huberman, 1994). Consequently, this represents a deductive approach and the combination of both approaches enables us to seek patterns in the gathered data while revisiting predefined general categories (Saunders & Lewis 2014). However, we have to state here that due to the relatively small number of participants making extensive use of Nvivo 12 did not make much sense, so that most of the analysis and the grouping of different codes were done by me personally rather than with the support of this software.

The first step in working with Nvivo 12 was to create a word cloud that is based on the most frequently used words in the interviews and stresses the significance of the data content by a set of representative words (Cui et al, 2010). The following figure illustrates the previously described process.

(43)

Coding Analysis

Here is a short overview of the chosen codes and the number of the references made can be found in the appendix (Can also be found in appendix 3).

This code table emphasizes that all of the significant variables of the primary research question and the theoretical framework of this research are represented through codes. So that, the technology or the routine in the framework is represented by the code Big Data (The Technology), the process of the model is represented by the code Exploitation of Insights, and finally, the consequence in the model is equal to the code Personalization. After establishing these codes, it was possible to focus on a specific relationship that we have assumed in the literature review and look at all the answers by all the respondents to this specific relationship. Naturally, many codes have also been dropped due to the fact that parts of the participants' answers were not relevant for the specific phenomenon that is researched here in this study. For instance, the codes General Information or Ad-hoc Analysis are codes that have been used and

Referenties

GERELATEERDE DOCUMENTEN

Electronic Data Interchange, procure-to-pay process, Robotic Process Automation, purchasing and supply management, facilitators and inhibitors, intention, implementation... Table

After establishing the current situation, a structured literature review on material planning and change management were conducted to design the ideal situation..

Our main argument for involving patients in translational research was that their input may help to make an innovation more relevant and useable for patients, and that it may

Agility [1] is achieved both through dynamic business processes and strategy (also think of alignment between different organizational levels), optimization [2] is primarily

Drawing on the RBV of IT is important to our understanding as it explains how BDA allows firms to systematically prioritize, categorize and manage data that provide firms with

 Toepassing Social Media Data-Analytics voor het ministerie van Veiligheid en Justitie, toelichting, beschrijving en aanbevelingen (Coosto m.m.v. WODC), inclusief het gebruik

Er is geen plaats voor het voorschrijven van combinatiepreparaten met cyproteron (merkloos, Diane-35®), omdat deze niet effectiever zijn dan andere combinatiepreparaten, terwijl ze

A study on the professional development of teachers who participated in such a typical context-based education professional development programme reported that teachers who gained