An exploratory research into the influence of big data on internationalization decisions

(1)

0

An exploratory research into the influence

of big data on internationalization

decisions

Master Thesis Feiko Schimmel 11083131

MSc Business Administration: International Management University of Amsterdam, Amsterdam Business School Supervisor: Dr. M.P. Paukku

Date: 27-1-2017 Word count: 17.034

(2)

1

Statement of originality

This document is written by Student Feiko Schimmel 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.

(3)

2

Abstract

This research investigates the opportunities of the new phenomenon of big data in international business. The goal of this research is to explore to what extent big data can influence the internationalization strategies of firms. This research is based on the Uppsala model of Johanson and Vahlne (1977) and has added the new phenomenon big data to this model. This exploratory research has used a single embedded case study in which qualitative data is acquired through semi-structured interviews with experts in the fields of market entries. These experts are all consultants working at one of the largest professional service firms in the world. Their overall view on the use of big data in internationalization strategies is asked. The analysis of the interviews has led to two supporting working propositions and one partially supported working proposition; the knowledge derived from big data can influence the entry mode decisions of firms, big data can generate market-specific knowledge and in some cases the use of big data may cause a firm to invest more in a certain country than it would have done without the availability of big data in its decision-making.

(4)

3

Acknowledgements

I would like to thank everybody who has supported me with the development of this thesis:

I would sincerely like to thank my supervisor Dr. M.P. Paukku. His input and feedback were highly valuable, and provided me good insights for constructing this research and writing this thesis. Furthermore, I would like to thank all the respondents that I have interviewed for their time and valuable contributions. Their contributions have been of utmost importance for this research. Also, I would like to thank my parents and friends for their support during my whole study, and especially during the period of conducting research and completing my thesis.

(5)

4

Index of Tables and Figures

Table 1. Nature of 5 V’s of Big Data and examples….…..………..…. 13

Table 2. Links between the interview questions and working propositions …….………..….. 37

Table 3. Characteristics of respondents ……….……….……….. 38

Table 4. Codebook ……….………….………. 42

Table 5. Illustrative quotes big data group 1……… 48

Table 6. Illustrative quotes big data group 2……….……… 52

Table 7. Illustrative quotes internationalization group 1……….……….. 57

Table 8. Illustrative quotes internationalization group 2………...……… 62

Table 9. Illustrative quotes entry mode choice group 1……….………… 67

Table 10. Illustrative quotes entry mode choice group 2………..………. 71

Table 11. Illustrative quotes knowledge group 1………...…75

Table 12. Illustrative quotes knowledge group 2………..…………. 78

Table 13. Results on the working propositions ……….………… 85

Figure 1. A hierarchical model of market entry mode ………...………22

Figure 2. The basic mechanisms of internationalization: state and change aspects………….. 25

Figure 3. Theoretical framework ………..………29

(7)

6

1. Introduction

1.1 Introduction & motivation for research

In 1977, Johanson and Vahlne carried out research on the influence of market knowledge on internationalization decisions for companies. The authors describe market-specific knowledge as knowledge about the characteristics of a particular host country. The most important characteristics are those that are displayed by individual customers. Additional characteristics that are considered are: a country’s business climate, its cultural patterns, and the structure of its market system (Johanson and Vahlne, 1977).

Nowadays, in the era of big data, a great deal of this market-specific knowledge is available to firms (McAfee and Brynjolfsson, 2012). Big data ensures that managers can measure (e.g. customer satisfaction and financial performance) and know more about their businesses and the environment in which it operates (Chen, Chiang, and Storey, 2012). Managers can use the knowledge gathered from this data to improve company decision-making and performance (McAfee and Brynjolfsson, 2012).

The multinational Amazon has already proved this to be the case. Amazon can see what customers buy, the other articles they view during the shopping process, and in what way they are influenced by Amazon’s promotions, reviews, and page layout. It also measures differences in buying habits between target groups. All this knowledge can be used in the decision-making processes of the company (McAfee and Brynjolfsson, 2012). Studies even show that big data might change the business processes involved in decision-making because it improves the visibility of a firm’s operations (Wamba, Akter, Edwards, Chopin, and Gnanzou, 2015).

Moreover, Gobble (2013) states that big data is the next big thing in innovation. Researchers from McKinsey found that “data can create significant value for the world economy, enhancing the productivity and competitiveness of companies and the public sector and creating a substantial economic surplus for consumers” (Manyika et al., 2011, p. 1). EY

(8)

7 Advisory state that firms that use big data from both the inside and outside of the firm are leading change in the world of business (EY, 2014).

Nonetheless, few empirical studies have been conducted to investigate the possibilities of big data. This is partly because big data is a relatively new discipline; the first academic articles to address big data have only been recently published, in 2011 (Wamba et al., 2015). The possibilities of big data can lead to insights for a company’s sustained value delivery, measuring performance, and establishing a competitive advantage (Wamba et al., 2015). One recommendation from Wamba et al. (2015) is to further investigate the influence that big data has on decision-making processes within firms.

Within firms, many strategic decisions have to be made. One decision that firms have to make when they go international concerns entry mode, which is the decision that firms must make when considering how they want to enter a foreign market (Brouthers and Hennart, 2007). Several factors that influence this decision have been investigated. According to Agarwal and Ramaswami (1992), three categories of factors can be distinguished: the ownership advantages of a firm, the location advantages, and the internalization advantages of integrating transactions.

However, current research into entry mode decisions has paid little attention to the impact of knowledge on these decisions. Two researchers who have very clear assumptions in this instance are Johanson and Vahlne (1977), who state that “[k]nowledge is an important obstacle to the development of international operations and that the necessary knowledge can be acquired mainly through operations abroad.” (Johanson and Vahlne, 1977, p. 23). Malhotra (2003) states that research into the influence that knowledge has on entry mode decisions is incomplete.

Based on the above-mentioned points, it can be said that little research has to date been carried out on the impact of big data on management decisions and thus the influence of big data on entry mode decisions. Therefore, a clear gap can be found that is important to fill for

(9)

8 the future development of international management research. It can be assumed that mining big data (extracting knowledge from big data (Provost and Fawcett, 2013)) can provide a significant amount of beneficial knowledge, which according to Johanson and Vahlne (1977) can be acquired by operations abroad. When firms are able to manage this knowledge to make use of it, research needs to be conducted to investigate if this knowledge will influence the internationalization strategy (e.g. entry mode decisions) of firms when expanding their offer abroad.

This leads to the central question of this study:

How does the knowledge management of big data influence the internationalization strategy of firms?

1.2 Goal

The goal of this study is to bridge the gap that has been identified in the previous section. The study will focus on the impact that knowledge from big data has on internationalization strategies and, therefore, on the entry mode decisions of firms. No specific market sector has been researched, as this topic is new in international business literature.

Due to the fact that the study aims to explore the influence that big data can have in the decision-making process when firms want to expand their offer abroad, it is important to review the literature about internationalization, entry mode research, knowledge management and big data. Moreover, semi-structured interviews with experts in advising firms in the process of international expansion will be conducted to gain deeper insight into the current situation and also to obtain the information necessary to answer the research question of this study. As consultants have been chosen from several sectors, this study is not focused on one sector or one market, but focused on their general overview.

(10)

9 1.3 Relevance

This study aims to stress the importance of the new phenomenon of big data and to investigate how this phenomenon influences internationalization strategies for firms. The various opportunities that big data can provide will be discussed in the literature review. The insight that this study provides can help firms better understand the opportunities that big data might offer them, and in what way this relatively new phenomenon can help in the decision-making process when a firm undergoes international expansion. Therefore, this study has practical relevance for managers of firms because it demonstrates a new way of obtaining knowledge for decision-making.

This study solely relies on qualitative data. These data are gathered by semi-structured interviews with internationalization consultants. Through these insights, this study contributes to International Management (IM) literature by exploring the opportunities that big data can provide to firms when they come to decide their internationalization strategies. This research also adds to the multi-cited recommendations proposed by Johanson & Vahlne (1977).

1.4 Outline

This remainder of this paper is structured as follows: The paper will begin with a review of the current literature on big data and the implications it can have for firms in their decision-making on internationalization. Moreover, it will provide an overview of the current literature on knowledge management and entry mode decisions. Thereafter, the working propositions for this research will be given, as derived from the literature review, in order to investigate the opportunities of big data in the internationalization strategy decisions of firms.

The paper will continue by discussing the methodology of the research, utilizing an in-depth case study to collect qualitative data through semi-structured interviews with consultants for market entry strategies. The validity of the working propositions are presented in the results

(11)

10 chapter and are examined in the discussion chapter. This paper will conclude with a comparison of the working propositions and results. The thesis ends with an acknowledgment of the limitations of the study, as well as its managerial implications, scientific relevance and suggestions for future research.

(12)

11

2. Literature review

This literature review will define the main concepts involved in this research. To gain an understanding of this research, it is important to understand the differences between information and knowledge. Therefore, this literature review starts with an explanation of what knowledge and information are. Furthermore, it will discuss the current literature on knowledge management, entry mode decisions and big data. This exploration into the existing literature of these core concepts will help set the stage for this study and to generate the working propositions that are presented in Chapter 3.

2.1 Information versus knowledge

As mentioned before, the difference between information and knowledge is slim. To understand these minor differences, definitions of the two concepts will be given:

In his research R. M. Losee (1997) defines information as: “one or more statements or facts that are received by a human and that have some form of worth to the recipient” (Losee, R.M., 1997, p. 255). The definition of knowledge given in the Oxford English Dictionary is: “Facts, information, and skills acquired through experience or education; the theoretical or practical understanding of a subject”. In other words, information are raw statements that can be received by a human that then become knowledge if they are also interpreted and understood. Thus, based on these two definitions, it can be said that information can create knowledge.

The concept of knowledge is important, as attested by the centuries-long debates about its meaning. Aristotle made a distinction between epistèmè (universal and theoretical knowledge) and technè (instrumental, context specific and practice-related knowledge), which continues to be discussed today (Flyvbjerg, 2004). Recent business literature continues to demarcate between implicit and explicit knowledge. Implicit knowledge of firms consists of informal knowledge that an employee has, while explicit knowledge comprises factual and

(13)

12 declarative knowledge, such as intellectual property rights, contracts and networks (Rasmussen and Nielsen, 2011).

2.2 Definition of big data

To get a better understanding of the phenomenon of big data, it is important to know what it is, how it is generated and what impact it can have on firms.

Big data are generated from an increasing plurality of sources including internet clicks, mobile transactions, user-generated content and social media as well as purposefully generated content through sensor networks or business transactions such as sales queries and purchase transactions. In addition, genomics, healthcare, engineering, operations management, the industrial internet, and finance all add to big data pervasiveness

George, Haas and Pentland, 2014, p. 2.

Several other scholars give similar definitions of big data. Russom (2011) states that big data has three important attributes: volume, velocity and variety. Wamba et al. (2015, p. 235) give a more comprehensive definition and define big data as: “a holistic approach to manage, process and analyse 5 V’s (i.e., volume, variety, velocity, veracity and value) in order to create actionable insights for sustained value delivery, measuring performance and establishing competitive advantages”. To get a good indication of what these 5 V’s exactly mean, and to illustrate it with examples, Wamba et al. (2015, p. 236), have created an overview (see Table 1.). Based on these definitions, it can be said that big data is a digital source of information. Moreover, big data is large and can generate information in real-time. Another aspect of big data is that it ensures we can learn from information: “Big data is about more than just communication: the idea is that we can learn from a large body of information things that we could not comprehend when we used only smaller amounts” (Cukier and Mayer-Schoenberger, 2013, p. 28). According to the definition of knowledge (Section 2.1), information that we have

(14)

13 acquired through learning becomes knowledge. Therefore, it could be said that big data generates knowledge.

Attributes Nature Examples Volume Large volume of data that either

consume huge storage or consist of large number of records

Tesco generates more than 1.5 billion new items of data every month.

Variety Data generated from greater variety of sources and formats, and contain multidimensional data fields

Tata Motors analyzes 4 million text messages every month, spanning everything from product complaints to reminders about service appointments to announcements about new models and also connected with customer satisfaction polling.

Velocity Frequency of data generation and/or frequency of data delivery

Retailers can now track individual customer's data including clickstream data from the Web and can leverage from their behavioral analysis. Moreover, retailers are now capable of updating such increasingly granular data in near real time to track changes in customer behavior.

Veracity Inherent unpredictability of some data requires analysis of big data to gain reliable prediction

eBay Inc. Faced an enormous data replication problem, with between 20 and 50 fold versions of the same data scattered throughout its various datamarts. Later, eBay developed an internal website (datahub) which enables managers to filter data replication.

Value The extent to which big data generates economically worthy insights and or benefits through extraction and transformation.

Match.com reported more than 50% increase in revenue in the last two years' time, with more than 1.8 million paid

subscribers in its core business, most of which driven through data analytics

Table 1: Nature of 5 V’s of Big Data and examples.

Source: Wamba, S.F. , Akter, S., Edwards, A., Chopin, G. and Gnanzou, D. (2015). “How ‘big data’ can make big impact: Findings from a systematic review and a longitudinal case study”. International Journal of Production

Economics, 165, pp.234-246.

2.2.1 Possibilities of Big Data

As stated in the introduction, big data ensures that managers can measure and know more about their businesses and their market environment (Chen, Chiang and Storey, 2012). By using these data, managers can directly translate that information gathered into knowledge that improves decision-making and performance (McAfee and Brynjolfsson, 2012).

Big data differs from traditional analytics in three ways: First, the volume of big data is enormous and can be used in decision-making processes. “Volume” refers to amount of data that is available to firms. For instance, Walmart collects 2.5 petabytes every hour from customer transactions (McAfee and Brynjolfsson, 2012). Second, the velocity of big data is high, providing companies with access to real-time – or nearly real-time – information, which makes a firm more agile than its competitors. Third, the variety of big data is huge. Big data comes in

(15)

14 the form of messages, updates and images posted on social networks, readings from sensors, and GPS signals from cell phones (McAfee and Brynjolfsson, 2012).

The use of big data ensures that new science, discovery and insights can be obtained with highly detailed, contextualized and rich content that has relevance to any firm or industry (Chen et al., 2012). The emergence of a greater amount of data available allows firms to discover the needs of different markets (Doan, Ramakrisnan, and Halevy, 2011).

Every day, 2.5 quintillion bytes of data are generated worldwide. One example to illustrate what this enormous amount of data can do can be found in the 1.8 million photos that are uploaded daily to Flickr (a public picture-sharing site). Since we all know the saying “a picture say more than a thousand words”, the billions of pictures on Flickr are a treasure trove for exploring social events, human society and public affairs (Wu, Zhu, Wu, and Ding, 2014). An example of the use of big data in a business related way can be found in Walmart. During hurricane season in the United States, big data is used to predict local demand for products less requested outside the hurricane season. As Walmart saves all the data gathered from transactions in previous seasons, they can predict what products will increase in sales just before hurricane season. Based on their data-mining (extraction of knowledge from data) Walmart can know in advance that the demand of, say, strawberry pop-tarts will increase sevenfold during hurricane season (Provost and Fawcett, 2013).

2.2.2 Implications of big data

Data driven decision-making means that firm decisions are made based on data analysis instead of manager intuition (Provost and Fawcett, 2013). Data driven decision-making can also be based on big data, as big data is the source of information on which decisions are made. McAfee and Brynjolfsson (2012) found that firms that use data-driven decision-making are 5% more productive and 6% more profitable based on financial and operational measurements. Provost

(16)

15 and Fawcett (2013) state that business performance can substantially improve when a firm adopts data science techniques based on big data. The importance of big data orientations and related managerial and operations issues is an area in which more research is needed (Wamba et al., 2015).

Nowadays, top level firm executives still make most of their decisions based on intuition that has been built over years of experience. McAfee and Brynjolfsson (2012) refer to this form of decision-making as relying on the highest paid person’s opinion (HiPPo). Researchers have found that managers and executives generally rely more on experience and intuition instead of data.

In sum, big data is a great source of information that provides firms with endless opportunities to change their way of doing business (McAfee and Brynjolfsson, 2012). The developments in information gathering towards which big data contributes is enormous. People talk about the world today as being a “digital age” and a “knowledge society”, leading to an understanding of the increasing importance of knowledge for firm development. The handling of knowledge is now part of organizational studies in university education and has given rise to the term “knowledge management” (Lin and Lee, 2006; Lundvall and Nielsen, 2007). The following section details what knowledge management is and what is currently known on this topic.

(17)

16 2.3 Knowledge management

To obtain greater insight into how firms manage knowledge and what is known about this topic, this literature review considers knowledge management in the specific context of business studies.

2.3.1 Definition

Knowledge management is defined as “a range of practices and techniques used by organisations to create, share and exploit knowledge to achieve organisational goals” (Jain and Jeppesen, 2013, p. 348). Another definition given in the literature is “Knowledge Management is the process of creating, sharing, using and managing the knowledge and information of an organization” (Girard and Girard, 2015, p. 14). Based on these two definitions, knowledge management can be said to be about creating, sharing and exploiting knowledge and information. The objective of knowledge management is to improve the performance and sustained viability of a firm (Wiig, 2002).

2.3.2 Origin

The origin of knowledge management lies in asset management, which is concerned with managing the aspects and the outputs of a firm to create a competitive advantage (Greco, Cricelli and Grimaldi, 2013).

Assets can be divided into two categories: tangible and intangible. Both these types of assets can create a competitive advantage for a firm. According to Greco et al. (2013), tangible assets are a firm’s physical and technological assets. An example of a physical asset is the benefit of a firm’s location. A technological asset is an advanced machine that a firm has (Greco et al., 2013).

(18)

17 Various scholars have defined what intangible assets are. Hall (1993, p. 135) states that “intangible resources range from intellectual property rights of patents, trademarks, copyright and registered design; through dependent, or subjective resources of know-how; network; organizational culture, and the reputation of product and company”. In more recent research, other scholars have defined intangible assets as:

[T]he stock of immaterial resources that enter the production process and are necessary for the creation and sale of new or improved products and processes. They include both internally produced assets – e.g., designs, blueprints, brand equity, in-house software and construction projects – and assets acquired externally – e.g., technology licences, patents, and copyrights, and the economic competencies acquired through purchases of management and consulting services Arrighetti, Landini, & Lasagni, 2014, p. 202.

Rasmussen and Nielsen (2011) describe intangible assets as being knowledge assets. For this research the focus is on the knowledge concept and the influence it has on decision-making for firms. Therefore, this literature review continues with the sources of knowledge that exist.

2.3.3 Sources of knowledge

In a study carried out by Castrogiovanni, Ribeiro-Soriano, Mas-Tur and Roig-Tierno (2016) four sources of knowledge are identified. The first is human resources, which are responsible for a large part of a firm’s knowledge. Human resources encapsulate the values, capabilities and experience of a firm’s members (Castrogiovanni, 2016). The second source of knowledge is an organization’s knowledge. This is knowledge that firms build through their own practice (Castrogiovanni et al., 2016). The third source of knowledge is the environment in which a business operates. This is the field of action, the production chain, and the intersection between value systems that can create new or improved products and services (Castrogiovanni et al., 2016). The role of the environment depends on the sector in which a firm is active. Sectors with high flexibility are sectors where firms can acquire knowledge easily and apply it without any

(19)

18 difficulties (e.g. problems to obtain the right knowledge). Another way to obtain knowledge from the business environment is by imitating other firms (Castrogiovanni et al., 2016). The fourth source of knowledge is technology. Jeppesen and Frederiksen (2006) claim that the use of computer software allows for more effective knowledge management. Almost all financial firms use software that connects them to their customers. With this software, firms have access to a vast amount of data about the markets in which they are active or want to expand (Castrogiovanni et al., 2016). Based on the above-mentioned definitions, big data can be seen as falling under this fourth source of knowledge.

2.3.4 Current literature on knowledge management

Knowledge management is an emerging discipline (Darroch, 2005). Knowledge is a strategic resource for a firm that can generate a sustained competitive advantage. Over recent years, a notion of the growing importance of knowledge as a critical resource for firms has stimulated managers to focus on creating a knowledge management strategy (Choi, Poon & Davis, 2006).

Several authors describe the consequences that effective knowledge management can have. First, knowledge management leads to competitive advantage (Connor and Prahalad, 1996; Hall, 1993). Second, it improves financial performance (Teece, 1998; Wiig, 1997). Third, it stimulates innovation (Antonelli, 1999; Carneiro, 2000) and assists with the anticipation of problems (Carneiro, 2000). Moreover, it enhances organizational learning (Buckley and Carter, 2000) and ensures the most effective use of information (Carneiro, 2000).

As mentioned above, effective knowledge management is important for firms. However, in the current literature not a great deal has been written about what effective knowledge management is (Darroch, 2005). Darroch states that this is partly because the current literature has shown that it is difficult to measure and identify knowledge management due to

(20)

19 the intangibility of some knowledge. Another reason given by the author is that knowledge management is a relatively new discipline (Darroch, 2005).

Knowledge may adopt a number of different roles as it is both a tangible and an intangible resource (Hall, 1993). One is the supportive role it takes in decision-making about resources. Furthermore, capability in knowledge management enables those within a firm to leverage the most service from knowledge and other resources (Penrose, 1959). Lastly, effective knowledge management greatly contributes to the innovation and performance of firms (Darroch, 2005).

As outlined before, knowledge can provide information about markets into which firms might want to expand. If a firm wants to expand to a foreign market, it must make several strategic decisions. One of the most important is the entry mode choice (Agarwal and Ramaswami, 1992). As data can give information about such markets, it is also necessary to know what impact data has on the entry mode choices of firms. To give a better insight into entry mode choices and to define current knowledge about the entry mode choice model, the following section discusses this model.

(21)

20 2.4 The entry mode choice model

To get a better understanding of the entry mode choice model, this section focuses on the question: What entry modes exist, and how do firms choose the right entry mode?

2.4.1 Types of entry mode

One of the most important decisions that firms make during their internationalization process is the entry mode in which they want to enter a country (Kogut and Singh, 1988). Therefore, research into entry mode concerns the decisions that firms make in how they choose to enter a foreign market (Brouthers and Hennart, 2007). Sharma and Erramilli describe an entry mode as follows:

[A] structural arrangement that allows a firm to implement its product market strategy in a host country either by carrying out only the marketing operations (i.e., via the export modes), or both production and marketing operations there by itself or in partnership with others (contractual modes, joint ventures, wholly owned operations).

Sharma and Erramilli, 2004, p. 2.

2.4.2 Schools of thought in the literature of entry mode

Several scholars have researched the reasons in which firms have to decide the mode they want to enter a country. This has led to three schools of thought (Pan and Tse, 2000). The first says that starting a business operation in a host country is risky because of political, cultural and market differences. According to this view, the engagement with the host market will gradually increase (Johanson and Vahlne, 1990). The first time a firm enters a host country market, firms tend to prefer the entry mode export, which means that the firm only starts exporting to that country. The reason for this is that this method requires little investment. When a firm obtains more knowledge about the host country’s market, it is willing to take more risks and therefore

(22)

21 invest more resources in that market (Pan and Tse, 2000).

The second school of thought is based on the perspective of transaction costs (Anderson and Gatignon, 1986). This school states that firms will do all activities internally as long as they can do it at a lower cost. Otherwise firms will start to subcontract other firms to perform their activities (Pan and Tse, 2000). According to this school of thought, managers consider all factors of all entry modes as having the same level of relevance in the decision-making process (Kumar and Subramaniam, 1997).

The third school of thought focuses on location-specific factors (Hill, Hwang, and Kim, 1990). Dunning (1988) made an eclectic paradigm to show the advantages and disadvantages of a host country, based on three key elements: ownership-specific factors, location-specific factors and internalization factors. Ownership factors are intangible, and tangible assets. Location factors refer to any (dis)advantages that a potential host country might have, such as natural resources or low wages. Internalization factors concern the choice between the ease of internalization and externalization of activities, largely based on costs (see the transaction cost theory developed by Andersen (1997)). Dunning (1988) emphasizes that location-specific factors are important when deciding on an entry mode.

(23)

22 2.4.3 Hierarchical entry mode model

Pan and Tse (2000) developed a hierarchical entry mode model which is displayed in the figure (Figure 1) below:

Figure 1: A hierarchical model of market entry mode.

Source: Pan and Tse (2000), “The hierarchical model of market entry mode”, Journal of International Business

Studies, vol. 31, no.4, p. 538.

Pan and Tse (2000) argue that managers follow a certain path to decide their mode of market entry strategy. A firm has to make certain choices when entering a foreign market. The first is whether to enter a foreign market in equity or non-equity mode (Pan and Tse, 2000). Once this has been decided, a firm has to make a decision on the second level of the hierarchical entry mode model. Within the equity mode, firms can choose between entering a country and obtaining a wholly owned subsidiary, or to start a joint venture. When a firm decides to enter in non-equity mode, it can choose between making export or contractual agreements (Pan and Tse, 2000). With regards to the level of investment involved, export requires the lowest investment and a wholly owned subsidiary requires the highest investment (Levary and Wan, 1999).

(24)

23 Managers often tackle important decisions by using a hierarchical decision model. They do this because they have limited analytical capacity (Pan and Tse, 2000, p. 538). By using a hierarchical decision model, the process of decision-making becomes a more manageable one. Pan and Tse (2000) tested the hierarchical model of entry modes, examining the influence of cultural distance on entry mode decisions.

2.4.4 Importance of entry mode research

Researchers emphasize the importance of international entry mode research because a firm needs to set correct boundaries when choosing entry modes. These boundaries have significant performance implications (Brouthers, 2002; Brouthers, Brouthers and Werner, 2003). Another reason that stresses the importance of entry mode choice is that once a firm has chosen a certain entry mode, it is difficult to switch to another. This indicates that whatever decision is made will have long-term consequences for the firm (Pedersen, Petersen and Benito, 2002). According to Brouthers and Hennart (2007), a large number of individual mode structures have been examined to show the differences between them. The most frequently investigated choice is the one between wholly owned subsidiaries (WOSs) and joint ventures (JVs).

As mentioned before, several studies show that different types of factors influence the choice of entry mode, such as firm-specific factors (Erramilli and Rao, 1993; Kim, Hwang and Hwang, 1990; Kumar and Subramaniam, 1997; Madhok, 1997), and industry-specific and country-specific factors (Anderson and Gatignon, 1986; Kogut and Singh, 1988; Tse, Pan and Au, 1997).

Brouthers and Hennart (2007) state that there is a lack of consensus about whether the distinctive categories of entry modes, such as contracts, joint ventures and wholly owned subsidiaries can be arranged by the single dimension of increasing control, commitment and risk. The authors propose that more realistic entry mode models need to be developed in order

(25)

24 to help researchers gain a deeper understanding of how firms can grow internationally and still maintain a high level of performance (Brouthers and Hennart, 2007). The internationalization of a firm is a multilevel phenomenon; several internal and external factors influence the final decision. It is important to understand that going international involves a firm from a home country setting up an activity in a host country. The chosen entry mode when entering a foreign market (equity vs. non-equity) depends on the characteristics of the parent firm, the operation, the relationship between the two firms, the situation in the industry entered and the characteristics of the host and home country (Brouthers and Hennart, 2007).

As stated in the introduction, there is not a good understanding of the knowledge construct and its influence on entry mode decisions (Malhotra, 2003). Scott-Kennel and von Batenburg (2012) found that only a few studies examined the influence of knowledge in the internationalization process. The authors concluded that internal knowledge has the highest influence in internal decision-making on a strategy level. Another important factor is that, without effective knowledge, mechanisms firms cannot fully benefit from the advantages of multi-nationality (Mattsson, 2000).

2.5 Internationalization according to Johanson and Vahlne

As mentioned earlier in this literature review, there are three schools of thought on the internationalization decisions of firms. The theory proposed by Johanson and Vahlne (1977) is that knowledge is an important aspect in the internationalization decisions of firms. Their model is one of the most cited in international business, and is known as the Uppsala Model. It is known that big data is a great source of information that can generate knowledge for firms (Section 2.2). Therefore, it is interesting to investigate how this new type of knowledge (big data) can influence the internationalization decisions of firms.

To dig deeper, the literature review will continue with an extended overview of Johanson and Vahlne’s (1977) model. The Uppsala Model is based on several observations;

(26)

25 their studies show that the development of internationalization in operations of firms increases slowly, largely because of the lack of knowledge that firms have of foreign markets. Firms do not begin the process by making a large foreign production investments. First, firms will start to export to a certain country, where a sales subsidiary is then started in that country. Only thereafter firms will establish a production facility in the host country. In addition, firms primarily start in countries that have a small physical distance from them before moving to countries with a successively greater physical distance away from the firm’s headquarters (Johanson and Vahlne, 1977).

One of the reasons firms do this (Johanson and Vahlne, 1977) is that the process of internationalization is an incremental process thanks to the initial lack of market knowledge in the host country and difficulty in obtaining it. Market knowledge is defined as information about markets and operations stored in the mind of people, either in a digital way on computers or in written reports (Johanson and Vahlne, 1977).

Johanson and Vahlne’s (1977) model explains the basic mechanisms in the internationalization process (Figure 2). In this model, there are two state aspects and two change aspects. The state aspects are market knowledge and market commitment; the change aspects are commitment decisions and current activities.

Figure 2: The basic mechanisms of internationalization – state and change aspects

Johanson and Vahlne (1977), “The Internationalization Process of the Firm – A Model of Knowledge Development and Increasing Foreign Market Commitments”, Journal of International Business Studies, vol. 8, no.1, p. 26.

(27)

26 Market commitment consists of two aspects, the amount of resources committed (which involves the size of the firm’s investment in a country) and the degree of commitment (which is the degree to which resources are integrated with other parts of the firm) (Johanson and Vahlne, 1977).

As aforementioned, Johanson and Vahlne (1977) distinguish four types of knowledge in business. First of all, experiential knowledge, which is knowledge that can only be learned through personal experience (Penrose, 1959). Secondly, objective knowledge, this type of knowledge can be taught. Thirdly, general knowledge, this form of knowledge is a common knowledge that is irrespective of the geographical location. Lastly, market-specific knowledge concerns the knowledge about characteristics of a specific national market.

(28)

27

3. Theoretical framework

After outlining the existing literature on the subject, clear links can be made between big data, knowledge management, internationalization and entry mode decisions. As stated before, big data is quite a new phenomenon that can help to improve performance of firms (Provost and Fawcett, 2013). Big data can generate a lot of market-specific information, so it can be seen as a source of knowledge. Knowledge management can improve the financial performance of firms (Darroch, 2005). Jeppesen and Frederiksen (2006) emphasize that the use of digital information (big data) can provide knowledge for firms when they want to expand to markets other than their home market.

As explained in the literature review, information is a source of knowledge (Section 2.1). There are four different types of information, which means that there are four different sources of knowledge. One of these types is digital data, meaning that big data (which is digital) can be considered a source of knowledge (Section 2.3). The huge amount of information that big data can generate for firms needs to be managed effectively so that data-driven decision-making improves performance (Section 2.2).

Knowledge management is about creating, sharing and using knowledge (Section 2.3). In other words, when firms are able to manage the knowledge available to them, they can use it to make strategic decisions that will improve their performance. Moreover, the information big data provides can lead to good decisions. This means that if a firm bases its entry mode decision on big data, the choice of the right entry mode can be made. Therefore the first working proposition is:

WP.1 Big data consists of information; by structuring this information, firms can extract knowledge from this data, which can help with choosing an entry mode.

(29)

28 As outlined in Section 2.5, the internationalization model proposed by Johanson and Vahlne (1977) stresses the importance of knowledge that a firm possesses of a certain country and the increasing commitment towards that country after having obtained that knowledge. While there are different types of knowledge that can be obtained, this research is focused on the impact of big data (digital data knowledge). Obtaining knowledge is often a quite slow process, and therefore the internationalization process of firms is slow (Johanson and Vahlne, 1977). Big data ensures that a large amount of information can be gathered in a short time period (Section 2.2). Considering the basic mechanisms in the Uppsala Model (Section 2.5), big data could speed up the process and generate knowledge that had previously been difficult to obtain. When adding big data to the Uppsala Model, a few interesting propositions for research can be derived.

First, whether or not the knowledge derived from big data can generate market knowledge should be explored. Therefore, the second proposition of this thesis is:

WP. 2 The knowledge derived from big data can generate market-specific knowledge.

Secondly, the theory about big data states that the amount of information that big data can provide is significant (Section 2.2). When the Uppsala Model is taken into account, it could be said that this huge amount of information could create so much knowledge that when a firm wants to expand its offer to a foreign country, the first steps of the learning process can be skipped because the firm has already obtained the knowledge that it would otherwise have to learn incrementally. Therefore, the third proposition of this thesis is:

WP. 3 When a firm uses big data to gain knowledge about a host country, it is willing to commit more resources in that host country than it would have without the use of big data.

(30)

29 Based on these three propositions an interesting model can be drawn as an extension of the Uppsala Model (Figure 3). The new model looks similar, but now big data has been added:

Figure 3: Theoretical framework, source: author.

An important note here is that managers and highly placed executives currently base their decisions on intuition and experience (McAfee and Brynjolfsson, 2012; Blake, 2008). Moreover, previous research does suggest that (senior) managers often use intuition in decision-making (Agor, 1990; Parikh, 1994; Kahtri and Ng, 2000). Intuition means: “being able to bring to bear on situation everything you’ve seen, felt, tasted, and experienced in an industry” (Kahtri, and Ng, 2000, p. 59). As mentioned earlier, this research wants to explore the opportunities of data-driven decision-making (using big data) in internationalization strategies, and how big data can be used to gather knowledge about certain markets. Data-driven decision-making means basing decisions on the analysis of data rather than purely on intuition. (Provost and Fawcett, 2013). This research incorporates the viewpoints of experts in the field who work in various functions. Based on above mentioned points, it could be that people working on a management

(31)

30 level or higher will have a different opinion about the use of big data in decision-making, than people working at a lower level because managers and highly placed executives are used to basing their decisions on intuition and experience.

Therefore, the results of the interviews with managers and higher placed executives (Group 1) will be presented separately from the lower-level respondents (Group 2). However, as this research involves a single-case study and is not focused on the differences in thought between managers and employees working on a lower level, no separate working propositions are presented for the two groups. To get a better indication of the methods used to answer the research question, and to find if the working propositions can or cannot be supported, see Chapter 4.

(32)

31

4. Research design

In this chapter, the research design will be discussed. The research question will be answered by way of a case study. First, this chapter will outline the ontological and epistemological foundation of research design. Second, the case research design will be proposed, especially the selection criteria of the case and the respondents selected for the interviews. Thirdly, the way that the data will be analysed is discussed. Lastly, the quality criteria of the research design is outlined.

4.1 Ontological and epistemological foundations of the research

The research philosophy adopted for this research contains important assumptions about how the world is interpreted by the researcher (Saunders and Lewis, 2012). Ontology deals with the nature of reality and therefore describes the assumptions of what the world is (Fleetwood, 2005). Epistemology means our theory of knowledge, especially how we obtain knowledge (Hirschheim, 1985).

Two aspects of ontology can be indicated: subjectivism and objectivism. The subjectivist view argues “that what is taken as reality is an output of human cognitive process” (Brannick and Coghlan, 2007, p. 62). The objectivist view states that “social and natural reality have an independent existence before human cognition” (Brannick and Coghlan, 2007, p. 62). Knowing this, it can be said that the subjectivist view is relevant when different social actors with potentially diverse perceptions of reality are involved in the research. On the other hand, the objectivist view emphasizes that social objects exist independently of social actors (Saunders and Lewis, 2012). The ontology of this research will be subjectivist in order to construct reality. This decision is based on the definition of the subjectivist view. Moreover, due to the fact that the decision-making during the process of the internationalization of firms is subject to individual actors, the subjective perspective of ontology is most appropriate.

(33)

32 According Saunders, Lewis and Thornhill (2016), there are two major contrasting epistemological assumptions in management and business research: positivism and interpretivism. Positivism is often referred to as the natural science approach for research and requires working with a social reality that is observable to generate law-like generalizations (Saunders et al., 2016). On the other hand, interpretivism has developed as a critique on positivism. This approach focuses on the subjective perspectives of humans on specific situations and experiences (Saunders et al., 2016). Researchers who adopt an interpretative perspective try to create new and richer understandings of organizational realities. For this study an interpretative approach is adopted in order to explore the influence of big data on decision-making during the internationalization process of firms. Questions that address the hows and whys of this process can be better answered using an interpretative approach (Saunders et al., 2016).

4.2 Type of study

The first choice in generating the appropriate research design is the decision between the qualitative and quantitative approach (Saunders et al., 2016). The main difference between the quantitative and the qualitative research approach relates to different ideas about reality. First, it is important to know if the working propositions are measurable and second, whether it is best to use objective or subjective research methods to understand what we know about the topic (Newman and Benz, 1998). When an interpretive research philosophy is adopted, the research design will often have a qualitative nature because the researcher wants to make sense of the subject studied (Saunders et al., 2016). To achieve this goal, it is important to conduct an in-depth study. Therefore, this research will adopt a qualitative research design.

Saunders and Lewis (2012) identify three types of business studies: The first is exploratory. This type of research is conducted if a researcher does not clearly understand a

(34)

33 topic and wants to explore general information about it. The purpose of an exploratory study is to discover new insights, to ask new questions and to evaluate topics in a new light. Second, descriptive studies are done to create an accurate representation of persons, events or situations. Third, explanatory studies focus on a situation or problem to explain the relationship between variables (Saunders and Lewis, 2012).

Knowing that little research has been conducted on how big data influences internationalization decisions, this research can lead to new insights on the use of big data in international business. Based on the above mentioned points, the type of research used in this thesis will be exploratory. New insights can be used to help build theories and find patterns for the use of big data.

This thesis does not, however, adopt a purely inductive approach. A strong theoretical framework has been developed, followed by several propositions (Chapter 3) that are tested during the in-depth interviews.

4.3 Research design

There are several possibilities for obtaining data for qualitative research. Having reviewed the possible options, a single embedded case study was found to be the most suitable option for answering the research question. One reason is that case studies can give researchers a good understanding of why certain decisions are made (Saunders and Lewis, 2012). Another reason for choosing a case study research design is that this method is excellent for conducting exploratory research (Yin, 2014).

When a case study is used, a phenomenon is analysed in its natural setting by conducting multiple methods of data collection (Benbasat Goldstein and Mead, 1987). Yin (2014) states that using a case study give researchers the opportunity to interpret and explain phenomena, which is in line with the interpretative approach of this research. To conduct this type of

(35)

34 research, multiple variables have to be taken into consideration, often leading to a research question that starts with “how” or “why”. These types of research questions stress the importance of contextual aspects, context richness, and a thorough examination of the cases (Yin, 2014). This research aims to understand how a certain phenomenon (big data) influences strategic decisions of firms (internationalization). Using the case study method will give a good understanding of how big data is used these days in the internationalization processes of firms. Therefore, a case study is most appropriate.

4.4 Case criteria & selection

The purpose of this study is to build theory based on a case study. Consequently, for this type of research, theoretical sampling is appropriate (Eisenhardt, 1989). According to Eisenhardt (1989), a random selection of the cases is not necessary and also not preferable. In the literature, the following definition of theoretical sampling is given: “Theoretical sampling means that cases are selected because they are particularly suitable for illuminating and extending relationships and logic among construct” (Eisenhardt and Graebner, 2007, p. 27). The purpose of theoretical sampling is to select cases that develop a theory. One way to conduct an exploratory research is by interviewing experts on a given subject (Saunders et al., 2016). Thus for this research, experts who have specific experience in the internationalization process of firms have been selected. This provides an understanding of the process decisions that were made and the way in which big data influences the internationalization decisions of these firms.

It is possible to find a large number of experts who consult with firms on their market entries within a consultancy firm. In the Netherlands, many of these consultancy firms exist. One of the highest ranked is EY (Formerly Ernst & Young) (Consultancy.nl, 2015). EY is one of the largest professional service firms in the world. After several negotiations, the author was granted access to the firm, creating a unique opportunity to interview the relevant experts. The

(36)

35 reason for choosing this company besides the points mentioned before, was due to the significant experience these consultants have in consulting other firms on an international basis and the huge amount of expertise that is inside this firm. Moreover, they serve a great variety of clients, which ensures a great overview. The consulting section of EY is involved with providing strategic advice: originating and evaluating potential deals and market entries (ey.com, 2016). Firms approach EY to get help with their internationalization strategy. The outcome of the advisory process that consultants of EY will recommend the right market entry strategy. To define this “right” market entry strategy, consultants must carry out research on several factors. First, they must search for a market that can meet the client’s objectives. Second, they must indicate the skills that a firm currently has for operating an international firm and which skills a firm needs to obtain to operate an international firm. Third, consultants must carry out research into the cultural and regulatory differences a firm will face in the target market, and identify how to approach them. Fourth, they must indicate the type of knowledge that the firm needs about the host country. Fifth, they must indicate the best governance structure for operating abroad (ey.com, 2016).

To get the information necessary for answering the research question of this thesis, the units of analyses are (senior) consultants, executive directors and partners of EY that are directly involved in consulting in the internationalization process of firms. Another factor is that they are expected to be the most knowledgeable, as supported by several researchers (Starbuck, 1992; Alvesson, 1992; Alvesson, 1993). As this research takes place in the Netherlands, all firms that the consultants advise are Dutch, with their headquarters in the Netherlands.

(37)

36 4.5 Data collection: Semi-structured interviews

When case studies are used to collect data, scholars often combine several methods. These methods are: archival research, interviews, questionnaires and observations. The data these methods provide can be quantitative or qualitative (Eisenhardt, 1989). For this research, semi-structured interviews are held to ensure that all relevant topics are covered, and that there is space to add questions or further input from interviewees (Saunders and Lewis, 2012).

Eisenhardt and Graebner (2007) argue that interviews are a highly efficient way to gather rich and empirical data, although this approach is often criticized for the existence of possible biases in the data. A key approach to limit this bias is to use experts who have diverse perspectives in a given area (Eisenhardt and Graebner, 2007). To avoid asking biased questions in the interviews, an interview protocol is used that contains the topics in need of discussion (Benbasat et al., 1987). The interview protocol starts with an introduction. After that some general questions are asked. The interview protocol continues with questions about internationalization, thereafter, there are questions on knowledge and big data, and lastly questions are focused on entry mode choices. The interview protocol can be found in Appendix A. Table 2 shows the questions that were asked during the interviews and how the working propositions can be linked to these questions.

(38)

37

Questions WP. 1 WP. 2 WP.3

General Questions

What are the main issues that the EY Advisory is involved with? . . .

What are the key tasks that you perform for clients? . . .

When (for what reasons) customers come to you? . . .

How does the advisory in market entries work? . . .

Internationalization

What reasons do your customers have to expand international or to acquire other

companies or to merge with them on an international basis? X

How are opportunities found in other markets? X

If a customer wants to go to another country, based on what information do you

choose which company is the right candidate or which country is the right country? X How is the country chosen where a customer is going to expand? X

Knowledge / Big data

What kind of information do you provide to your customers during the Advisory

process? X X

What are the sources of knowledge that you have about a certain market? X X

Do you use digital data (data analytics) for giving advice? If so, where do these

data come from? X X

What knowledge is needed to arrive at a decision which market and which

candidate is suitable for the customer? X X X

What knowledge do clients possess in terms of internationalization (e.g. mergers

and acquisitions)? (Is it based on experience, market research etc.) X

What form of knowledge can give more certainty about the right decision? X X

What can big data do for internationalization decisions of firms? X X X

Can big data assist in making an entry mode choice for a firm? If yes, How? If not,

why not? X

Entry mode choice

At which point do firms decide what entry mode choice is best for the firm? X

Who in the company finally decides what is the right choice? X

Are firms that have big data available in decision-making willing to invest more in

a foreign country, than firms that do not? Why? X

Table 2: Links between the interview questions and working propositions, Source: Author

4.6 Respondents

After access to EY was granted, the respondents for conducting the interviews were selected by applying non-probability sampling. Purposive sampling was initially used, meaning the interviewee selection process chose only people within EY who were considered the most suitable for answering the research questions and meeting the objectives of this research (Saunders and Lewis, 2012). The people targeted therefore had direct experience in consulting with other firms during their internationalization process (for a brief description of the process, see Section 4.4). As the EY advisory department is quite diverse, focus was on consultants who are involved in the advisory of market entry strategies. By using an overview of all people working at EY’s Amsterdam office, a selection was drawn based on the aforementioned criteria. These people were contacted by e-mail for interview. Also, the diversity of function level in

(39)

38 the interviewees was taken into consideration. Hence, a variety of partners, executive directors, managers and (senior) consultants were interviewed. Table 3 shows the characteristics of the respondents that were interviewed. Since all employees that met the criteria (12 in total at the Amsterdam office) were asked to participate in the interviews, only eight respondents were available to interview.

Table 3: Characteristics of respondents, Source: Author

To get an overview of how the function levels within the advisory department of EY are built and the role of each function and how competencies are developed, see Figure 4.

Group 1

Respondent Function Department Age Experience in market entry sector Duration (minutes)

Respondent 1 Junior Manager Advisory 33 7 Years 44

Respondent 2 Partner/owner Advisory 53 28 Years 46

Respondent 3 Executive director Advisory 39 12 Years 43

Respondent 4 Senior manager Advisory 52 10 Years 38

Group 2

Respondent 5 Senior Consultant Advisory 31 3 Years 41

(40)

39

Figure 4: Advisory career framework, source: EY.

As this research focuses on the impact that knowledge has on internationalization decisions, viewpoints on the usage of big data of respondents working on a management level or higher could differ from respondents working at junior and senior consultant levels. One of the reasons for making this split is that managers and highly placed executives base their decisions on intuition and experience (McAfee and Brynjolfsson, 2012) (See Chapter 3 for an extensive outline). Another reason for the split is that respondents working on a management level or higher are more involved in leading consulting teams and having contact with the clients. The junior and senior consultants analyse the market, find opportunities, and so on (cf. Function level descriptions, EY, 2016). An additional factor that supports this split is the consultants’ experience. Respondents working on a management level or higher must have at least seven years of experience, while the junior and senior consultants can only have up to three years’ experience in advising firms about market entry strategies. Therefore, as mentioned in the

(41)

40 theoretical framework, the variety of function levels has been taken in consideration in the selection of interviewees.

After the interviews were held, a clear distinction became visible between respondents that work on a junior or senior consultant level and respondents that work on a management level or higher. Consequently, in the results section, the results based on the interviews with the respondents working on a management level or higher (Group 1) are presented separately from the results based on the interviews with the junior and senior consultants (Group 2). This enables results from different viewpoints to be contrasted and compared. All respondents are working at same general department within EY (EY Advisory) and are specialized in market entry strategies of firms. Based on the interviews, it became clear that each employee has a specific focus in the sector of firms they consult (e.g. technology, retail, agriculture and oil- and gas firms). However, the interviews were not focused on the differences between sectors because the respondents serve clients across sectors, and the focus of this research has been on the respondents’ general perspective.

Due to the short time in which this research had to be conducted, it was not possible to gather a large sample of experts. However, those who did contribute to this research have ensured valuable insight into the way Dutch firms act when they consider moving abroad. Moreover, the interviews offer a good understanding of the way that big data influences the internationalization decisions of firms. Interviewing experts in the field of consulting other firms when they want to invest and expand abroad also provides a wide overview of several firms in different sectors.

All interviews were held face-to-face in the office of the interviewee, and all interviews were conducted in Dutch. Before each interview, the interviewee is informed that the interview is anonymous and it will be handled confidentially. This reduces the possibility of participant bias and this increases the reliability of this research (Saunders et al., 2016). Each interview is

(42)

41 fully recorded by a telephone recording device and is afterwards transcribed. This leads to an unbiased and accurate record (Saunders and Lewis, 2012). The transcripts are in the same language in which the interviews were held.

4.7 Data analysis

Data analysis is conducted to gain a deeper understanding of the proposed working propositions by examining the collected data that can lead to answering the research question (Yin, 2014). A generic approach to analysing qualitative data is thematic analysis (Braun and Clarke, 2006). According to Braun and Clarke (2006), thematic analysis allows for a systematic approach to data analysis that is flexible and accessible. The purpose of this method is to search for themes or patterns that occur across a data set (Saunders et al, 2016). Thematic analysis is found the most suitable for exploratory research in identifying patterns from transcripts (Saunders et al, 2016). As this research has an explorative nature, this method is considered the most suitable.

According to Ryan and Bernard (2003), the steps of thematic coding are as follows:

(1) discovering themes and subthemes, (2) winnowing themes to manageable few (i.e. deciding which themes are important in any project), (3) building hierarchies of themes or codebooks, and (4) linking themes into theoretical models.

Ryan and Bernard, 2003, p.85.

This method is used in this thesis.

Fereday and Muir-Cochrane (2006) state that themes can be derived from data, which they consider inductive coding. Moreover, they state that themes can also be derived from the literature, which means that there is an a priori template of codes. This combination is also used for this research because it ensures that all possible explanations of the phenomenon are captured (Fereday and Muir-Cochrane, 2006).

To ensure that this research can be replicated, a codebook (Table 4) is used. This allows others to follow the applied method and validate it. Moreover, it makes the methods that are

An exploratory research into the influence of big data on internationalization decisions