Big Data Analytics adoption and process to achieve competitive advantage: Integrating the Resource- Based View & Institutional Theory

(1)

Big Data Analytics adoption and

process to achieve competitive

advantage: Integrating the

Resource-Based View & Institutional Theory

(A Case Study of the Automotive Environment)

Philipp Kreyenberg

Student ID: NCL: b8018195 / R: S3903443

Double Degree: Advanced International Management

Institutions awarding the degree: University of Newcastle &

University of Groningen

Supervisors:

Dr. Elizabeth Alexander &

Dr.

Rudi de Vries

(2)

Abstract

(3)

Acknowledgment

First of all, I would like to thank all employees at BMW AG who have supported me on my

journey. A very special thanks goes to Tobias Kulzer, my supervisor at BMW, who has been

very supportive. He expressed interest in my work and took the time needed to help me

throughout the process, no matter how busy his schedule was.

A special thanks goes to all my informants who took the time to answer my interview questions.

I am grateful for the openness, commitment and interest that they all expressed towards my

study.

I would also like to thank my supervisors at the University of Newcastle and the University of

Groningen, Dr Elizabeth Alexander and Dr Rudi de Vries. Their advice, questions and

encouragement have always challenged my thought processes and inspired me to push

boundaries.

(4)

List of Figures

FIGURE 1: CONCEPTUAL FRAMEWORK BASED ON THE LITERATURE ... 19 FIGURE 2: CONCEPTUAL FRAMEWORK FOR THE ADAPTION AND PROCESS OF BDA ... 39

List of Tables

TABLE 1:LIST OF KEY INFORMANTS ... ERROR! BOOKMARK NOT DEFINED. TABLE 2: CODE MANAGER SECONDARY DATA ... 34

List of Abbreviations

BDA

Big Data Analytics

CP

Coercive Pressure

GDPR

General Data Protection Regulation

IS

Information Sharing

MP

Mimetic Pressure

NP

Normative Pressure

(7)

Introduction

Big data is seen as an increasingly fast-growing asset which adds potential to an organisation

thus it aroused the interest of all organisations (Amankwah-Amoah,

2016). It is considered an

information asset and has been characterised in three main features, namely the high volume

of data, the velocity in terms of data creation and transmission and the variety of data. These

three features are key in compelling specific technological and analytical methods to transform

the data into measurable value (De Mauro, Greco, and Grimaldi 2016; Salleh & Janczewski,

2016). Further features which can be included are veracity, variability, value and visualisation

(Mikalef et al. 2018). Where big data is seen as an asset, the Resource-Based View (RBV)

provides a framework in which organisations can invest to create value. Nevertheless, due to

the variety of big data, the solutions can be diverse and therefore affect the value in various

ways (Ardito et al., 2018). An example of this diversity becomes clear when analysing the

production domain, which is at the forefront of big data usage (Tan et al., 2017; Zhou et al.,

2019). Manufacturers use a large amount of big data and store more data than any other sector.

The data is used for running performance simulations, discover new patterns, pilot industry 4.0

(Zhou et al. 2019), real-time management of complex systems (van der Spoel, Amrit, and van

Hillegersberg 2017), increasing production yields (Baily and Manyika 2013), and for the

transformation of their supply chains (Baryannis et al. 2019; Hofmann 2017; Pan et al. 2017).

(8)

the format. In other words, it is possible to analyse whether the data sets are structured or

unstructured, such as videos, audio recordings and texts. Furthermore, BDA can be used for

the entire supply chain of the automotive environment, e.g. for product development (Deloitte,

2015).

The importance of data exploration becomes clear when one analyses the worldwide

development of big data and business analysis revenues. The revenue is expected to grow from

$122 billion in 2015 to more than $187 billion in 2019. Over the next five years, revenues are

expected to grow by more than 50% (IDC, 2016). Big data has been identified as the "next big

thing" in innovation (Wamba et al. 2017), attracting large investments and having the potential

to add value in a variety of operations (OECD 2013).

Consequently, a topic on the rise amongst scholars and industry is Big Data Analytics (BDA)

(Wamba et al. 2017). It is also considered to be the most important technological disruption

since the advent of the Internet (Chen, Preston & Swink 2015). The BDA is the organisational

capability (Srinivasan and Swink, 2018) that enables competitive advantage to be gained when

large amounts and variations of data are processed at the speed required for relative insight,

according to information literature (Akter et al., 2016; Gupta and George, 2016; Pauleen and

Wang, 2017).

(9)

concerns, but also by regulatory standards. There are only limited studies which have used a

theory-based approach to explain how BDA can contribute to improve the performance of the

organisation (Chen, Preston and Swink, 2015; Gunasekaran et al., 2016). The organisational

processing theory (Srinivasan and Swink, 2018) or the dynamic capability view (Akter et al.,

2016; Wamba et al., 2017) are the most common theories applied in studies. Nonetheless, a

more holistic view of BDA in relation to issues and the related capabilities of BDA has not

been promoted. In summary, it can be argued that literature provides only limited

understanding of organisational-level of usage of BDA as well as the adaption and process

behind it. To fill this gap, I address the research question: How can big data analytics adaption

and processes be designed to achieve competitive advantage?

To answer the research question, I used an overarching theoretical lens on two major theories:

The institutional theory (DiMaggio and Powell, 1983) and the RBV (Barney, 1991). The

institutional theory breaks up the adaptation of BDA by looking at the interrelation and

coordination between the two major interest groups, which are the stakeholders and the focal

organisation. The RBV stresses the role of internal resources as independent motives for

organisations (Oliver, 1997). In addition, RBV emphasizes the different roles of external

pressure and internal resources and their relationships (Tatoglu, Glaister and Demirbag, 2016;

Zheng et al., 2013). External pressure can influence internal resource development and thus the

adoption and process of BDA to improve operational performance, but this is not well

understood in the context of BDA. Big data-related organisational resources can be exploited,

as Braganza et al. (2017) argue in line with the RBV perspective to gain competitive advantage.

This paper is structured as follows. Firstly, the literature review is presented. Secondly, the

subsequent sections consecutively will describe the methodology followed by the case analysis

and discussion. Finally, my conclusion is drawn including implication for the practice, as well

as limitations and future research.

Literature Review

(10)

and Swink, 2018). Nonetheless, the anticipated performance level of the organisation can only

be achieved if the organisation responds effectively to significant and relevant external

pressures or environmental demands, as Aydiner et al. (2019) noted. Consequently, this study

considers two different perspectives, firstly the RBV (Barney, 1991; Grant, 1991; Wu et al.,

2006) and secondly the institutional theory (Allen, Allen and Lange, 2018; Demirbag, Glaister

and Tatoglu, 2007; Demirbag et al., 2017; DiMaggio and Powell, 1983;), to investigate how

the adaptation and the process of BDA can be designed to achieve a competitive advantage for

the organisation.

Resource-based view

One of the most debated and proven theory in management studies is the resource-based-view

(RBV) (Nason & Wiklund, 2018). The RBV consists of a comprehensive stock of management

literature. The literature deals mainly with the source and nature of an organisation's strategic

resources and capabilities (Priem & Butler, 2001). The RBV argues more precisely that an

organisation must use its resources optimally in order to achieve superior performance through

sustainable competitive advantages. In addition, the RBV applies either the inside-out view or

a firm-specific view when it comes to the question of why an organisation fails or is successful

in the market (Dicksen, 1996). For the organisations resources to be of real benefit, they must

be valuable, rare, inimitable and not substitutable (VRIN), as previous studies have shown

(Wernerfelt, 1984, Barney, 1991). Furthermore, valuable resources “must enable a firm to do

things and behave in ways that lead to high sales, low costs, high margins, or in other ways add

financial value to the firm” (Barney 1986, p.658). Moreover, Barney (1991) highlighted that

“resources are valuable when they enable a firm to conceive of or implement strategies that

improve its efficiency and effectiveness” (p.105).

(11)

productivity of other resources (Makadok, 1999). Thus, capabilities are of high importance for

the organisation (Hitt, Ireland, Sirmon, & Trahms, 2011).

In the perspective of BDA, connectivity (tangible) and information sharing (intangible) are two

key resources to shape BDA capability (Gunasekaran et al., 2016). Furthermore, the BDA

generates new knowledge in connection with large amounts of data. This new knowledge

should be used and exploited together with other important internal and external knowledge

and non-knowledge sources (Ferraris et al. 2018). The study by Dubery et al. (2019) discovered

and empirically proved that human skills and tangible resources of a company have a positive

influence on BDA.

Firm tangible and intangible resources

Resources can also be supplemented by other resources such as financial, technological and

reputational capital (Grant, 1991). Organisational capabilities depend on the bundling of

resources, as the literature of the RBV shows (Wu et al., 2006). In addition, the organisational

ability helps to explain the organisational performance. According to scholars of the BDA

studies, organisational performance and thus competitive advantage can be exploited if

organisations understand how to use the BDA (Gunasekaran et al., 2016; Gupta and George,

2016). The creation of BDA capabilities can be achieved by combining intangible and tangible

strategic resources that can improve the operational performance and thus the competitive

advantage of the organisation (Srinivasan and Swink, 2018). It can be argued, that tangible

resources such as financial (debt, equity) and physical assets (equipment) can be acquired in

the market when following the RBV logic. Tangible Resources can be divided into three

categories when taking the arguments of Gunasekaran et al.’s (2016) and Gupta and George’s

(2016) into consideration:

i.

Connectivity and information sharing (IS) can be viewed as a technological resource

(12)

ii.

Technology is needed to analyse big data. For example, Gupta and George (2016) argue

that in order to extract valuable and authentic information, organisations technologies

must be able to process volume, diversity and velocity from data. A closer look at

relational database management systems (RDBMS) reveals that 80% of stored data is

unstructured. RDBMS thus make it difficult for organisations to analyse the data. This

leads to organisations moving away from RDBMS. As a result, technologies such as

Hadoop and NOQSL have been developed in recent years, in order to facilitate the

storage and parallel processing of unstructured data sets.

iii.

Basic resources are required by the organisation apart from data and technology. Most

organisations need to make significant investments in big data incentives. This means

that standard operating procedures are not yet in place and need to be studied before

implementing. Therefore, investing in big data will not pay off immediately. However,

in order to meet the BDA incentives and desires for success, it is essential that the

organisation remains willing to invest in these projects. Once firms are allowed to make

informed decisions through BDA, the organisation will be able to benefit from the

process. (Gupta and George, 2016). This is a unique resource which is valuable,

inimitable and not substitutable. In the context to Gunasekaran et al. (2016) it can be

argued that big data assimilation is a three-stage process, where time plays a central

role as it is important for the organisation to recognise the potential benefit arising from

big data technology. Following this argumentation, two basic key resources for

organisations are time and investments for a successful BDA integration and process.

New skills and management styles are required to keep up with the growing demand of the data

driven world according to McAfee et al. (2012). In addition, Waller and Fawcett (2013) argue

that all kind of different sets of skills are required for BDA. A subset of these intangible firm

resources creates capabilities within the organisation which are non-transferable and also have

the purpose of increasing the productivity of other resources (Makadok, 1999). Furthermore,

these resources are not substitutable as they are unique for each organisation. Leading us to the

intangible resources which have three important aspects:

I. Human resources can be divided into two groups. The first group focuses on people

(13)

artificial intelligence, statistical analysis, cleaning and extraction of data. In addition,

people in this group have the capability and interest for learning and understanding the

new technological trends. The second group of people are those who have the skills to

manage big data. These people are responsible for planning, implementing and

controlling processes dealing with big data. However, most importantly, they need to

understand how the knowledge gained from big data can be applied to different areas

of the organisation (Wamba et al., 2017; Gupta and George, 2016). Waller and Fawcett

(2013) noted that due to the novelty of big data technology and the capabilities

associated with it, organisations have significant advantages over their competitors

when their employees have big data skills and capabilities. Nevertheless, technical

skills alone cannot provide a competitive advantage in the long term. The reason for

this is that the big data skills may ultimately be spread among people working in the

same or different organisations. As a result, over time this resource becomes common

to all organisations (Nonaka, Toyama and Konno, 2000). The technical skills that an

organisation needs for BDA can be realised by hiring new talents and/or training

existing staff. In addition, management skills are highly organisation-specific and,

according to Gupta and George (2016), are only developed over time by individuals

within the organisation. Management skills can be considered tacit and are therefore

distributed heterogeneously throughout the organisation. In addition, mutual trust has

developed over time and good working relationships have developed between big data

managers and other functional managers. The trust and relationships between these

managers are likely to contribute to superior big data human skills that may be difficult

for other organisations to replicate.

II.

II. Data-driven culture is of great importance as it allows managers to make

(14)

organisation are non-transferable and also have the purpose of increasing the

productivity of other resources (Makadok, 1999).

III. Organisational learning points out that the development of capabilities is necessary to

explore the accumulation, exchange and transformation of knowledge, especially if an

organisation is interested in validating and contextualising the results obtained from big

data. For example, organisational learning helps to adapt informed decision-making

processes within an organisation by enabling the validation of knowledge from large

data sets (Gupta and George, 2016). These capabilities have an impact on the strategic

choice of organisations. Furthermore, the development of capabilities allows the

organisation to add value to their supply chain, develop new products or even expand

in new marketplaces (Madhani, 2010). Organisational learning helps the organisation

to understand the mutual coordination of resources. This is important for two reasons.

Firstly, the heterogeneity of resources and secondly, the differences between diverse

resources, which can vary. This is important in order to compensate deficits and thus

maximise the efficiency of resource usage. In addition, organisational learning helps

management to realise the full potential of scarce resources so that they can focus on

realising these resources while maximising the effectiveness of the scarce resources

(Hung et al. (2011).

It can be concluded that the above listed and explained tangible and intangible resources cannot

be easily transformed or purchased. Therefore, these resources require a longer learning curve

and changes in organisational culture. This makes these resources unique to the organisation

and thus more difficult for the competition to imitate.

Impact of intangible and tangible resources on BDPA acceptance under the mediation

effect of top management commitment

(15)

firm needs top managers, as they are a key factor in building capabilities and orchestrates

resources (Chadwick et al., 2015).

Despite the assimilation of the technology and the importance of top management commitment,

the current literature is underdeveloped when it comes to building BDA acceptance. Looking

more closely at the assimilation process, the first stage is the acceptance of the technology as

advocated by scholars (Davis, 1989), succeeded by routinization and assimilation (Hazen et

al., 2012).

Acceptance, Routinisation and Assimilation of BDA

Routinisation can be defined as the “permanent adjustment of an organisation's governance

system to account for the incorporation of a technology” according to Zmud & Apple (1992,

p.149). Based on Saga and Zmud (1994), Hazen et al (2012) argue that the threefold process

consists of acceptance, routinisation and assimilation. This, post-adaption diffusion of

innovation (Hazen, Overstreet, & Cegielski, 2012) will be used in order to develop and test the

model. This will help to explain the influence of BDA on organisational and cost performance.

Assimilation is the degree to which technology spreads through the organisational processes.

Furthermore, assimilation can be divided into a three-stage post-diffusion process which are,

according to Hazen, Overstreet, & Cegielski (2012):

Acceptance: How BDA is perceived by organisational stakeholders

Routinisation: How well the management systems of an organisation are adapted to the BDA

Assimilation: How well BDA has spread through the organisational process

To reap the benefits of anticipation, organisations must not only accept technology, but

continue to routinise and assimilate it (Hazen et al., 2012). Taking the RBV perspective into

account, the acceptance and assimilation capabilities of an organisation can be developed

through the mediation of a BDA routinisation construct. The routinisation enables acceptance

and assimilation capabilities such as supply chain integration and IT and thus affecting the

performance of the organisation and their competitive advantage.

Institutional Theory

(16)

or acceptable economic behaviour (Oliver, 1997; Peng et al., 2009). The motives of human

behaviour go beyond economic optimisation and extend to social justification as well as social

obligation as assumed by Zukin and DiMaggio (1990). Oliver (1991) argues that since

individuals or organisations are in part prisoners of social conventions, there is a tendency

among the elements of society to make their decisions in accordance with social norms. Thus,

their decisions are often shaped according to social norms so that they are in accordance with

them. In addition, DiMaggio and Powell (1983) and Oliver (1997) argue that the success of

organisations is linked to social norms if they are followed. Drawing on the arguments of

Tatoglu, Glaister and Demirbag (2016) they find when analysing the adoption of corporate

practices at the organisational level, that embedding must be taken into account, even if there

are different nuances in the institutional perspectives.

Nevertheless, the effects of the institutional environment need to be studied more

systematically; regulatory, cognitive and normative institutions need to be taken into account

in relation to practice, creating the capabilities required for adaptation and the processes of

BDA (Kostova & Roth, 2002).

Following this logic, institutional differences influence all aspects of an organisations practice

(Tatoglu, Glaister and Demirbag, 2016). Therefore, by using the RBV and institutional theory,

the adaptation and process of BDA in order to achieve a competitive advantage can be

explained.

The forces from institutions and the resources of the firm

(17)

history and make inappropriate resource decisions; ( 2 ) sunk costs can be cognitive rather than

economic and lead to suboptimal resource choices; ( 3 ) cultural support for resource

investments may be an important determinant of their success; ( 4 ) firms may be unwilling

rather than unable to imitate resources and capabilities, especially when those resources lack

legitimacy or social approval; and ( 5 ) social influences exerted on firms reduce the potential

for firm heterogeneity” (p.700).

Comparing institutional theory with other theories such as transaction cost economics and

resource dependency theory, the difference is that structural and behavioural changes within

an organisation are driven by organisational legitimacy rather than competition and efficiency

under institutional theory, as argued by Liang et al. (2007). DiMaggio and Powell (1983) argue

that this drive for legitimacy causes organisations to institutionalise themselves. As a result,

originally different organisations now develop similarities without necessarily becoming more

efficient. Leading to institutional isomorphism. The institutional isomorphism can be classified

into three types coercive, normative and mimetic according to DiMaggio and Powell (1983).

These three types of institutional isomorphism mechanisms individually also drive regulatory,

cognitive and normative institutions, which according to Scott (2001) belong to the institutional

environment. Thus, institutional isomorphism mechanisms affect the institutional environment.

Coercive pressure (CP) may result from government regulations and guidelines issued by

(18)

access to important scarce resources (Liu et al., 2010). Liu et al. (2010) further argue that when

an organisation chooses to innovate (in this case BDA), it tries to obtain information on

institutional expectations and norms. They use this information to estimate the potential costs

and benefits of introducing BDA. It is important to position oneself accordingly in order to

protect oneself against uncertainties (Choi and Eboch, 1998; Scott, 2008). As a result,

organisations tend to choose safer technologies in order to comply with the social pressure,

regulatory and legal. Jackson and Schuler (1995) argue that in order to respect the legitimacy

of organisations, the recruitment and selection process must be analysed, which can be done

through the theoretical perspective of institutional logic.

Normative pressure (NP) comes from professionalisation and impacts the focal organisation

(Zheng et al., 2013). Through education and professional networks a pool of interchangeable

employees are shaped. This can be found in each industry (DiMaggio and Powell, 1983; Liang

et al., 2007). NP is an incentive for the introduction of inter-organisational information systems

according to Teo, Wei and Benbasat (2003). Furthermore, to prevent being locked out of

cooperative relationships and to ensure access to resources, organisations align with NP. In

addition, organisations join forces with NP to prevent them from being excluded from

cooperative kinship and to ensure access to resources.

Furthermore, if this pressure materializes, organisations are inclined to embrace innovative

technologies as argued by Liu et al. (2010). Hence, organisational behaviour is formed by

individuals who are similar, with regards to orientation and disposition. In the new era of BDA,

managers have technology and management backgrounds inhabiting the upper echelons. It can

be concluded that managers accept and promote data-driven decisions by putting pressure on

the organisation (Akhtar et al., 2018; Gupta and George, 2016).

Mimetic pressures (MP) arises from the tendency of organisations to mimic others. Despite the

(19)

Conceptual Framework Based on the Literature

The importance of a conceptual framework is justified by several researchers who claim that

ideal research should be anchored in theory so that the results can be easily integrated into the

existing knowledge base (Stake 2000, 2005 and Yin 2009, 2012). Thus, this study integrates

two different perspectives: first the RBV (Barney, 1991) and second, the institutional theory

(DiMaggio and Powell, 1983).

This integration helps to explain how the unique capability of BDA within an organisation is

built. This capability is of high importance for the successful adaptation and processes of BDA.

It is of great importance to achieve unique and highly efficient informed decision making

within the organisation, which leads to a competitive advantage. Figure 1 illustrates the

conceptual model of adaption as well as the process of BDA by linking the antecedent factors

(institutional factors and organisational resources), moderating construct (top management

commitment). Consequently, this framework will be used as the guide to the study.

Figure 1: Conceptual Framework Based on the Literature

Methodology

This chapter presents the research method used to answer the research question. Firstly, the

research method is presented. Secondly, the research setting is described as well as how the

data is collected for this research. The last part of this chapter explains the data analysis,

reliability and validity.

Coercive Pressure (CP) Mimetic Pressure (MP) Normative Pressure (NP) Tangible Resources Intangible Resources Top Management Commitment Connectivity (C)

Information Sharing (IS) Technology Basic Resources

(20)

Research Method

With regard to research methods, a decision must be made between qualitative and quantitative

research. This thesis follows the strategy of conducting qualitative research. Quantitative

research emphases on numerical and immutable data, using structured research tools that

capture larger sample sizes that are representative of an entire population (Babbie, 2010). In

contrast, qualitative research aims to understand data in its context, so this method emphasises

words rather than the quantification of data (Bryman, 2012). With the focus on the topic of

BDA and the understanding of its adaptation and processes in the automotive environment, a

qualitative research was considered to gain useful insights. In addition, a qualitative approach

can be useful when researching a topic (in this case BDA) on which only limited information

and studies from previous research are available (Strauss & Corbin, 1990; Hoepfl, 1997). The

choice of a qualitative approach will be further strengthened as certainty and objective truths

about the subject of the adaption and processes of BDA have yet to be developed. In addition

to the distinction between qualitative and quantitative research strategies, the choice of research

design is an important decision as it helps to define a framework for data collection and analysis

(Bryman, 2012). If one follows the description of the five different research designs by Bryman

(2012) and using his terminology, it can be concluded that this master thesis under review

follows a qualitative single case study. This qualitative research method has the emphasis on

the thoughts and views of informants and not on quantifying the collected data (Bryman &

Bell, 2011). This study emphasizes the informants view and thoughts upon the adaptation and

the process of BDA in the automotive environment.

Empirical Context

(21)

to that the analysis of data becomes more and more important in the automotive environment.

Therefore, organisations need to have the ability to analyse data sets which are structured or

unstructured (Deloitte,2015).

As a result of technological changes, the automotive environment is changing and opening up

the entry barrier for new players further, affecting the role of key players and their power in

the automotive environment (Deloitte, 2015). New players in the automotive environment are

software providers. In addition, BDA has great potential in the automotive environment in the

customer behaviour-, marketing mix-, supply chain mix- and predictive quality-analysis

(Deloitte, 2015). This situation makes the automotive environment an interesting environment

to be analysed to answer my research question.

Data Collection

The most regular data source in a qualitative research are face to face, in-depth and detail

interviews according to Creswell (2013) and Yin (2009, 2012). Thus, the primary data are

semi-structured interviews which have been conducted with individual informants. Further, the

informants can be seen as “knowledge agents” because they can explain their thoughts and

knowledge the best (Gioia et al., 2012). It is therefore important to capture how reality is

perceived by the informants (Bryman & Bell, 2011).

(22)

Furthermore, it allowed cross-checking of information provided by the informants (Huber and

Power, 1985). The key informants also provided triangulation to ensure the integrity of the

study and to build rigor, validity, credibility and reliability, as suggested by Creswell (2013)

and Denzin and Lincoln (2011). Table 1 lists the informants who were interviewed for this

study including their organisation, role, job title and how long the informant has been working

for the organisation. I chose five key informants because of the limited time for the research

and the limited interview opportunities due to the corona virus. Furthermore, the number of

five interviews can be justified for two reasons. Firstly, a large number of journals, book

chapters and books propose five to fifty informants and are recommended as sufficient

(Dworkin, 2012). Second, most researchers argue that the concept of saturation is the most

important factor to consider when making sample size decisions in qualitative research (Mason,

2010). A research project can be considered saturated "when gathering fresh data no longer

sparks new theoretical insights, nor reveals new properties of your core theoretical categories’’

(Charmaz, 2006, p. 113). Therefore, the number of five key informants used in this work can

be considered an appropriate sample size.

A first secondary data collection in the form of newspapers and magazines in the field of BDA

was compiled from LexisNexis to prepare the author for the empirical research.

Table 1:List of Key Informants

Interview Organisation Job Title Role within the Organisation

Yearsin the Organisation

INT1 Allianz Head of Accident

Research/Loss Prevention

Manager 20

INT2 BMW Head of the

Transformation Office

Manager 15

INT3 Campana &

Schott

Technology

Consultant for Data Analytics

Executive 3

INT4 BMW Insurance Specialist Senior Executive 7

INT5 Campana &

Schott

(23)

Furthermore, this master thesis adopted the advices of Eisenhardt and Graebner (2007) and Yin

(2009, 2012) in order to ensure a qualitative case study research that is rigorous, highly

comprehensive and has a systematic research methodology. Therefore, the following things

have been done: preparation of data collection, collection of evidence, analysis of evidence and

composition of the qualitative case study report. Furthermore, in this thesis, rigorous data

collection was carried out in carefully interlinked steps, including in-depth interviews to ensure

the use of several sources of information and the creation of a database of case studies.

Thus,

based on the learnings and gaps in the literature, a general interview roadmap was constructed

to ensure all desired topics would be addressed (Appendix 1). Building on this roadmap, a more

detailed semi-structured interview guideline was created to ensure a uniformity of questions

across the interviews (Appendix 2). The guide was tested on its clarity and coherency by a

dummy interview before the in-depth interviews were carried out. The semi-structured

interviews have been used because they gave me the flexibility to analyse the informants for

details and provide the broadest possible scope, while ensuring that we continued to cover the

topics relevant to my research question (Yin, 2003). The interviews were conducted in German

and ranged from 30 to 72 min long.

Secondary data were also collected to obtain an accurate and complete picture (Yin, 200f3).

Secondary data were collected by LexisNexis. Searching for BDA in LexisNexis yielded over

15,000 sources. To narrow down the research, I focused on newspapers and magazines

published between December 2013 and September 2019. To minimize the results found further,

the sources must also contain the words "process", "performance" and "analytics" or " big data"

or "business analytics" or "technology" or "data analysis". In addition, sources also should

focus on "information management and technology" and "computer science and information

technology". The last point to narrow down the research was that the sources should be

available in German or English and should be published in Europe. After removing the

duplicate sources, the final result of my research was 111 newspapers and magazines, which

were analysed to answer my research question.

Data Analysis, reliability and validity

(24)

interest (Gioia et al., 2012). All the conducted interviews have been systematically recorded

and managed as recommended by Yin (2009, 2012). Furthermore, the data which has been

collected during the interviews was transcribed with a true verbatim style and at an appropriate

level of detail. Thereafter, the transcript has been reviewed again.

I have used the tandem reporting, which follows the voice of the informant as well as of the

researcher. The tandem reporting can be used to systematically demonstrate the “1

st

_-order”

analysis (codes) and “2

nd

_{-order” analysis (concepts, themes and dimensions). The coding of}

the primary as well as the secondary data has been done over “ATLAS.ti”. Lastly, the codes

were crosschecked for the first and second draft. During the coding process each data item was

given equal attention. All extracts have been composed per code in an initial coding

framework. The coding framework has been divided into primary and secondary data with the

same codes. This allowed me to demonstrate the links between data but also helped to enable

insights concerning my research question which is a key feature of high-quality qualitative

research (Gioia et al., 2012). The transcripts and the developing analyses and models were

communicated to the informants. However, the informant was not given a veto right over

anything other than the reporting of sensitive data.

The work with the transcript as well as the secondary data collected and analysed formed the

basis for the findings on my research question. In the next section I will describe my findings

from the primary and secondary data.

Findings

(25)

Findings Primary Data

The findings from the interviews can be divided into eleven themes, that will be discussed in

the following.

Influence of coercive pressure on the adaption and process of BDA

This pressure represents a hurdle, especially for globally active organisations. Most countries

have their own specific regulations. Therefore, contracts in different countries have different

regulations (INT2). In the context of the BDA, global organisations must review themselves

with a certain level of expectation, especially with regard to data protection and data ownership.

Organisations must therefore check which guidelines apply in the individual countries (INT4).

As a result, IT security is of great importance for data protection (INT3).

Looking more closely at Germany, regulations such as the General Data Protection Regulation

(GDPR) applies. This regulation stipulates that organisations may only save and collect data if

the organisation has a reason for doing so and clearly defines beforehand what data they need.

This is of course a great challenge for organisations, as it is often not possible to define at the

beginning of the development of a product what data is required (INT4). In addition, the

collected data must be made anonymous (INT 1; INT3; INT4; INT5) and the organisation must

be able to delete certain data. Thus, the user has the possibility to agree to the use of his data

and to revoke it at any time (INT5). Non-compliance with these regulations can have

considerable consequences for the organisation, which was not the case before the introduction

of the GDPR. In the past, the maximum amount an organisation could pay for breaches of the

regulations was 200,000 euros (INT4). Large organisations were therefore not concerned with

breaking the law, as the business that could be created as a result was usually much more

profitable. However, with the introduction of GDPR this has changed. The penalty that can be

imposed can be set at 4% of a company's annual turnover (INT4). Consequently, compliance

with the rules and laws is of great importance for organisations.

(26)

Influence of normative pressure on the adaption and process of BDA

External pressure from other organisations such as Alphabet, which is entering the automotive

industry, is forcing organisations within the automotive industry to adapt new technological

arrangements such as BDA (INT3). Therefore, BDA needs professionalisation, which can only

be driven forward if the organisation ensures that processes and people are taken along (INT1).

This depends on the size of the organisation. The larger the organisation is, the more

complicated it becomes to involve the employees and to implement the process throughout the

organisation. Also, because of the different generations working with an organisation, issues

relating to data are viewed differently (INT2). This makes it more complicated to form a

uniformity of organisational behaviour. Nevertheless, a cultural change can be observed in the

last two years (INT5). In the past, BDA or informed decision-making based on data using

technology was considered a nice gimmick. In the new age of technology, there are various

tools that already show that it makes more sense to work data specific. (INT4).

It can be concluded that the size of the organisation has an influence on the development of

professionalisation. In addition, a cultural change can be observed with regard to BDA

engagement, and it can be observed that managers accept and encourage data-driven decisions.

The NP affects both the resources of the organisation and the management.

Influence of mimetic pressure on the adaption and process of BDA

This pressure can also be found in the automotive environment. For the reasons that vehicles

are becoming more and more technological, the entry barriers of the automotive environment

are becoming weaker. Consequently, new players have the opportunity to enter the market.

Therefore, the tendency shows that it is no longer sufficient to only look at competitors, such

as Tesla which uses the data generated from the vehicle and also uses personal data passing it

to third parties, to have a competitive advantage (INT4). The digital environment is making its

way into the automotive industry. Key players such as Google, Apple and Amazon have created

digital ecosystems that simply have a high level of penetration that plays in different leagues

than car manufacturers do (INT2). “This means that there is perhaps no way around integrating

our car into the Android or Apple world, but it is a great challenge not to lose our own added

value” (16 INT2).

(27)

other industries or from the same industry (INT3). It is important to look at what competitors

are investing in. This means that organisations are looking at their competitors’ actions. If one

organisation in your environment invests 3% of their revenue in BDA, other organisations in

the same environment may not be allowed doing so because the risk of losing an advantage

over others could be highly behind in the near future (INT3).

In conclusion, the MP can be found in the automotive industry. It is important to search not

only in one's own environment, but also in other environments, as different environments

become more and more interrelated. It is a simple principle that once an organisation is

performing well, other organisations will want to add the same value to their organisation and

tend to mimic others.

Required intangible resources for BDA

A closer look at the tangible resources required for BDA, the findings have revealed three key

resources. Firstly, it is important having the required human skills. This means that employees

must have the competence to work with the BDA (INT3). On the one hand, human skills can

be developed through internal or external training (INT1; INT2). Trainings are important to

move employees away from Excel tools to other programs required for BDA (INT2). With

adapting BDA within the organisation, you need certain professionals such as lawyers, data

stewards, analysts, data engineers and someone higher up in the hierarchy to set the central

tone (INT5). If training takes too long and the specific skills are needed immediately, or the

required skills are not available within the organisation, there is also the option of purchasing

the required skills. However, the real challenge is to find the right people with the right spirit.

(28)

Thirdly, organisational learning is crucial for the BDA because of the need to develop skills to

explore, accumulate, transfer and transform knowledge. Organisations have their own internal

innovation departments that can be helpful to better understand and develop big data

approaches, independent analysis or artificial intelligence (INT1). Furthermore, it can be seen

that a department that controls the BDA approach is the one having the most knowledge. This

department knows all processes and helps other departments to build up BDA. The data ideas

are supported specially to roll out data projects faster (INT4).

It can be concluded that the most important intangible resources needed for the adoption and

the process of BDA are human skills, data driven culture and organisational learning. The

employees can be trained internal or external and if required the experience can be bought in.

It is necessary to establish a data-driven culture for the efficient implementation of BDA.

Lastly, organisational learning is important for the development of capabilities, to explore

accumulate, share and transform knowledge that can be established.

Required tangible resources for BDA

The results have identified four tangible resources, which are important for BDA. Firstly,

technology is an important resource for driving BDA forward. It is important that the

organisation has the necessary broadband connection, which means that the organisation has

the ability to transport the information quickly and that the technical requirements are in place.

In addition, technology is needed for the reporting and analysis tools and further technological

instruments such as artificial intelligence and the use of certain hardware to store and process

data, e.g. data lakes are required for BDA (INT2; INT5). Nevertheless, the technology process

is not harmonised in most organisations, which will need to be done in the future to drive the

BDA forward (INT2). Organisations within the automotive industry are in the phase where

individual technologies have been tried and tested for BDA. However, organisations are still

in the phase where they need to find out which technologies are best suited for their

organisation and which they can use (INT3).

(29)

importance to ensure that the organisation links the tools in the best possible way to obtain the

corresponding benefits (INT3). In the organisations of the automotive environment a better IS

takes place, so that a certain degree of connectivity can be affirmed. Connectivity can, for

example, read data from the vehicle and thus analyse the driver's behaviour and help to assess

the risk (INT1). The exchange of knowledge is being increasingly promoted (INT4).

It is of great importance to develop a target architecture so that processes with data have to be

developed (INT2). For example, the organisation must have the data architect who is then

responsible for managing my platform. Then the data engineer, who is ultimately responsible

for development and design. Then there is the Data Scientist, who is responsible for the

"experiments". In addition, there is a product team, a front-end developer and a back-end

developer. Lastly, there are the Business Analysts, who are responsible for the information at

the beginning of the chain (INT3). This means that there are no standard operating procedures

yet and that these need to be investigated for their implementation. In addition, the

implementation of BDA is determined by certain time levels, so that a company does not want

to lag behind, but rather start at a later point in time (INT3). Following this line of reasoning,

the basic key resources for companies are time, investment and the necessary employees for a

successful BDA integration and BDA process (INT3).

Overall it can be concluded that the most important tangible resources needed for the adoption

and the process of BDA are connectivity, IS, technology and basic resources. The technology

is important to develop the right technical architecture with platforms and data lakes for the

BDA. In addition, connectivity and IS are required to achieve sound decision making and

transparency within the organisation. Furthermore, basic resources such as time, investment

and the necessary human resources are important to have the capability to start a project like

BDA.

Top Management Commitment

(30)

value chain, both a top-down and a bottom-up approach are needed. The bottom up approach

is important to show top management the potential BDA has and provide them with ideas

(INT2; INT3). The top down approach is required for the development of the strategy. The

strategy developed by the top management helps to channel resources, to achieve benefits and

steers the entire project (INT3). The strategy is of importance to build a strategic platform

which is central for BDA and helps to identify what the content fields are and what potential

they have for the organisation. In addition, top management needs to move away from the use

case level because a platform needs to be built to use BDA. The platform must be developed

strategically and not around use cases of the organisation. Nowadays, it is also very important

not only to think about driving the feasibility study to the point where it is ready for assembly,

but also to go new ways and try things out (INT1). When it comes to BDA, organisations can

reach four different stages of maturity (INT3):

1. Description

2. Diagnostics

3. Predictive

4. Prescriptive

Therefore, managers from different hierarchical levels must work together to exchange ideas

and create a strategy. Thus, it is important to have a strategy as well as a top-down and

bottom-up approach to successfully implement the BDA processes in the organisation.

BDA Acceptance within the automotive environment

(31)

data and processes are as efficient as possible, still the protection of employees is the main

focus (INT3).

It can be concluded that different interest groups have different perspectives on BDA.

BDA Routinised within the automotive environment

In order to be efficient in BDA, the management systems must adapt to the process. In the

automotive environment, most key players have been around for a long time and their systems

are not built around BDA (INT4). Only in some parts of the value chain are systems adapted

to BDA or can be adapted to BDA without replacing the existing systems. Especially the car

manufacturers have been using data sets for a long time, which are extended and further

developed with regard to the product development of their vehicles. In other parts of the value

chain the adaptation and the process of BDA (INT4) is missing. As a result, an organisation

that has been in existence for a long time prefers to build new management systems rather than

to adapt their existing management system to BDA (INT2). Through technological progress,

BDA becomes a process of continuous improvement (INT1).

However, in many organisations it can be observed that analytical topics are controlled from

the IT system. This can be a problem when an organisation tries to implement BDA, because

it is based on a big data approach. The fact that big data is structured differently means that the

organisation has to handle the process differently (INT3).

It can be concluded that the adaptation or introduction of new management systems to BDA

will take time. In addition, a new architecture for BDA is required. Both the architecture and

the process of BDA must first be understood in order to avoid misunderstandings.

BDA Assimilation within the automotive environment

(32)

helps to identify further potential (INT1). Once the BDA process is implemented in the

organisation, informed decision making can take place. An optimal scenario would make the

decision maker able to open his cockpit and look into a detailed overview to see what the

current situation looks like. In this way, he can immediately see which vehicles are affected

and can therefore take immediate action.

It can be concluded that BDA is only found in some departments along the value chain in the

automotive environment. Nevertheless, it is important that BDA is implemented across the

entire value chain in order to be able to make informed decisions. In addition, once

implemented internally, BDA should be further disseminated to external parties such as

suppliers.

Data Governance

There must be a department, which is the central role that knows exactly who gets what rights

and duties and then manoeuvres them. Everyone within the organisation must then comply

with these standards. It is very important that everything must fit together in this process, both

technically and procedurally (INT 5). There is a need for data governance, which must be set

up to support the positive effects and prevent the negative ones (INT3). Furthermore, due to

the different legal regulations in the various countries, it is important that data governance

within the organisation is rolled out globally (INT3). Unfortunately, this is not the case.

Organisations are only slowly beginning to do so (INT2). Therefore, understanding data

governance has often been a hurdle. The understanding of it has yet to grow.

Summing things up, organisations need to implement a data governance department that has

the overview and control over the BDA adaptation and process within the organisation.

Competitive Advantage

(33)

On the other hand, BDA also offers several advantages for the customer. New BDA processes

will ensure better systems. These are better adapted to the customer and guarantee the

corresponding satisfaction (INT2). Leading to that there is an added value having the product

more adapted to the customers wishes (INT3). In addition, the product has more functions and

can provide these to the customer at a higher speed. A good example in the automotive industry

is a navigation system, that can display the latest traffic jam information. This is a classic BDA

use case in the automotive industry. The customer benefits from this because it is a product

that offers the customer the advantages of avoiding traffic (INT3). Ideally, through the usage

of BDA within the organisation the customer receives a better service. This makes the customer

more satisfied, because the service or product may be cheaper or even more specifically tailored

to the customer’s needs (INT5). From the customer's point of view, however, organisations

must ensure that the service or product that is tailored to the customer by BDA is not too

personalised, as this could lead to problems. The customer can no longer go to his neighbour

and compare his products, because the product for each individual will look completely

different. For example, Facebook uses BDA in the form of a feedback tool, in these scenarios

a customer does not necessarily get added value or too much advertising (INT3).

Thus, the conclusion can be drawn that BDA is a competitive advantage for the organisation.

It helps to develop new product opportunities, which can generate added value to the

organisation. In addition, BDA offers opportunities to realign the business model that can

benefit the customers (better services & products). BDA enables outsourcing, optimises digital

processes, identifies problems earlier, automates them, reduces cost pressure and helps to

transfer human resources competence to better tasks (INT3).

Findings Secondary Data

The analysed newspapers and magazines focused on different industries, which can be found

in table 2. The overview shows how often the themes have appeared in the secondary data and

how often they appeared in the industries, Table 2.

(34)

industries that want to use the BDA in Europe, as the legal framework GDPR was introduced

on May 25, 2018.

Table 2: Code Manager Secondary Data

The GDPR regulates the exposure of consumers to risks, arising from the use of information

about them. It therefore applies to all organisations worldwide that use personal data of EU

residents for collection and processing purposes for their business activities.

NP affects organisations in various sectors not only because of technological progress and the

entry of new players from the IT environment, but also because of their customers.

Customers demand a more responsive and personalised service. They expect seamless channels

and a service that is available in real time. Another aspect is that society is changing due to

technological progress. People are interacting more and more with apps, social media and web

interfaces. As a result, business requirements will no longer be the same as before, as

organisations need to adapt to new methods that affect the culture and technology of the

organisation. It appears that more and more managers understand these changes and try to find

solutions by forming teams, that are adaptable and multi-skilled or by training their employees.

Consequently, it is important to create a consistent organisational behaviour.

The following key resources required by BDA, can be divided into tangible and intangible

resources and can be identified from the secondary data. Tangible resources identified from the

Industry / Code AcceptanceBDA Assimilation BDA RoutniseBDA Coercive Pressure Compeditive Advantage Connectivity Data Driven Culture Human Skills Information Sharing Intangible Resources Mimetric Pressure Normative Pressure Organisational Learning ResourcesTangible Technology Top Management CommitmentGrand Total

Airports 1 1 2

Banking & Financial Services 5 4 17 9 2 7 3 3 4 4 7 3 11 2 81 Construction 1 1 2 Consulting 3 2 1 4 2 1 13 Engery Sector 4 4 3 1 3 1 1 4 1 1 1 24 Engineering 1 1 Farming 1 1 Generall 6 1 3 30 21 13 6 12 16 8 9 9 15 7 14 19 189 Government 1 1 5 1 2 4 3 1 1 19

Healthcare & Pharma 6 2 1 9 4 3 7 3 1 3 1 1 2 4 47

Insurance 2 1 2 3 2 1 1 12

Internet Publishing & Broadcasting & Web search portals

2 4 1 5 3 1 1 3 1 1 7 29

IT & Technology Sector

(35)

secondary resources have been connectivity, IS and technologies. These are the same resources

identified by the primary sources. Technologies such as "Workforce Management Solutions"

can help management to use employee resources more efficiently. In addition, they show that

cloud storage and data lakes can be of great use to back up data needed for BDA. Consequently,

connectivity and IS are important for making informed decisions based on data. These two

resources are high of importance not only because they support strategic decision making,

because they also help developing long-term benefits of the firms in the supply chain. Looking

more closely at intangible resources, the secondary data as well as the primary data showed

that human skills, data-driven culture and organisational learning are important. The

date-driven culture is significant to ensure that the adoption and processes of BDA are accepted by

the employees. This cultural change must be implemented throughout the organisation.

Furthermore, it is important to have the required human skills driving BDA forward. Thus, it

is important that the organisation has the right human skills at the right time. Organisational

learning is essential to understand the need to be agile. Agility is a necessity, if the organisation

wants to keep pace with rapidly changing technologies, competitor initiatives and the needs of

their customers.

Management plays an important role for the adaption and process of BDA, as the primary data

showed already. The decisions made while taking into consideration big data tend to be better

decisions and lead to better performance. Managers should understand this and build

BDA-specific capabilities. The attention of C-level managers shifted towards data due to the big data

hype. There must be a bottom-up and top-down approach between managers at different levels

of the hierarchy in order to develop a strategy for BDA that has also been identified in the

primary sources.

Primary and secondary data show that Management and resources have an influence on BDA

acceptance. BDA is becoming more and more accepted, which can also be seen in major

industries, for example the insurance industry, which is driving BDA to develop their business

further.

(36)

to better monitor the performance of employees, enable managers to better utilize the

workforce and to extend collaborations with partners. BDA will, for example, enable banks to

create faster and more reliable models. This will shorten the development time and reduce the

challenge of limited sample size compared to conventional methods.

The next step would be the BDA assimilation. This is important in order to be able to make

sound decisions along the entire value chain, to achieve a competitive advantage. An excellent

example of BDA assimilation can be found in the banking sector. Compared to other industries

such as retail or telecommunications, banks have a competitive advantage when it comes to

BDA because banks have built up data warehouse and data analysis infrastructures earlier than

other industries. Therefore, their systems have matured over time. Building on this, banks have

the ability to process internal data and integrate, structure and standardise data. Further

examples of BDA assimilations and their success stories can be found in the IT industry. Key

players, such as Amazon and Facebook, perform millions of data-based operations every day

and process millions of pieces of data every hour, all stored in databases for further processing

and analysis.

BDA shows that faster access to data and the analysis of it leads to faster access to insights.

These insights are important to enable a faster response to the market and to identify weakness

within the value chain, which is the key to gain a competitive advantage. This is of great value

to organisations, especially in the retail and manufacturing industries. It can be seen that

different environments are more and more connected by technologies. Consequently, it is

important that technologies and BDA are combined in a way that makes it possible to

efficiently monitor real-time asset performance, utilise historical data to continuously predict

future performance and to identify weaknesses or opportunities in the supply chain, which is

critical to achieve a competitive advantage.

Conclusion of Findings