Effects of Data Science, Predictive Analytics & Big Data (DPB) on the Supply Chain

Table of contents

Declarations

Specimen Abstract

Abstract

Acknowledgements

Table of contents

Glossary

List of figures

List of tables

1 Introduction
1.1 Research Topic, Relevance and Focus
1.2 Research Objectives
1.3 Structure

2 Literature Review
2.1 Overview
2.2 The Terminology
2.3 Big Data
2.4 Data Science
2.5 Supply Chain and Supply Chain Management
2.6 Applied SCM Data Science
2.7 Implementation of SCM Data Science
2.8 Predictive Analytics
2.9 Financial Sector
2.10 Services Supply Chain
2.11 Summary

3 Methodology
3.1 Introduction
3.2 Approach
3.3 Design
3.4 Strategy & Role
3.5 Methods
3.6 Credibility & Limitations
3.7 Ethics
3.8 Data Collection
3.9 Interviews

4 Findings, Application and Discussion
4.1 Introduction
4.2 Services Supply Chain
4.3 Big Data
4.4 Decision Making
4.5 Summary

5 Conclusion
5.1 Results and Answers
5.2 Reflection and Outlook

Reference List

6 Appendices
6.1 Definition: Bandwagon Effect
6.2 Applications of Analytics
6.3 Effectiveness vs. Domain Knowledge
6.4 Definition: Sexy
6.5 Implementation of Big Data
6.6 Maturity Map
6.7 Methodological Pyramid
6.8 Research Onion
6.9 Individual Consent
6.10 Organisational Consent
6.11 Interview Guide
6.12 Extract of transcription and translation
6.13 Extracts from the interview analysis process
6.14 Interview Participant Coding
6.15 Customer Journey
6.16 Definition: Buzzword
6.17 Definition: Opportunity Costs
6.18 Explanation: SWOT
6.19 Definition: Micro-Management
6.20 Definition: Telematics
6.21 Definition: Generation Y
6.22 Reflective Statement

Last Page

Declarations

I declare the following:-

(1) that the material contained in this dissertation is the end result of my own work and that due acknowledgement has been given in the bibliography and references to ALL sources be they printed, electronic or personal.
(2) the Word Count of this Dissertation is +/- 10.933 due to anonymization.
(3) that unless this dissertation has been confirmed as confidential, I agree to an entire electronic copy or sections of the dissertation to being placed on Blackboard, if deemed appropriate, to allow future students the opportunity to see examples of past dissertations. I understand that if displayed on Blackboard it would be made available for no longer than five years and that students would be able to print off copies or download. The authorship would remain anonymous.
(4) I agree to my dissertation being submitted to a plagiarism detection service, where it will be stored in a database and compared against work submitted from this or any other School or from other institutions using the service. In the event of the service detecting a high degree of similarity between content within the service this will be reported back to my supervisor and second marker, who may decide to undertake further investigation which may ultimately lead to disciplinary actions, should instances of plagiarism be detected.
(5) I have read the University Policy Statement on Ethics in Research and Consultancy and the Policy for Informed Consent in Research and Consultancy and I declare that ethical issues have been considered and taken into account in this research.
(6) I have read the University Policy Statement on Data Protection in Research and Consultancy and I declare that the data collected for use in this dissertation has been properly safeguarded and will be destroyed once the dissertation or subsequent research activity has been concluded. I acknowledge that it is my responsibility to destroy the information with due regard to confidentiality.

SIGNED: Lukas Ebert…...

DATE: March 29th 2017

Specimen Abstract

STUDENT NAME Lukas Ebert

DEGREE BA (Hons) International Business Administration (Completion)

DISSERTATION SUPERVISOR n/a

DISSERTATION TITLE Effects of DPB on the Supply Chain

DATE March 2017

KEYWORDS Supply-Chain-Management Business-Intelligence Data-Analytics

Preamble In order to ensure the protection of data and privacy, certain parts of the paper had to be anonymized and censored (“n/a”, “OI”, etc.)

Abstract

Working for a financial service provider raises the awareness for certain issues within this sector and how those could be solved by means of new trends such as big data.

This undergraduate thesis presents an exploration of the effects of big data on the supply chain and on decision making. A topic of significant relevance, especially since the literature review discovered that there had been insufficient research conducted on the supply chains of service providers. Furthermore, this investigation of existing literature created a frame for the content of the dissertation by displaying the roots and development of big data and its far-reaching impact in and beyond the business context.

The research method that was used for this dissertation consists of qualitative data collection by interviewing employees of a key player in the German leasing and asset finance market, providing insight to the business from the industry leaders viewpoint.

After conducting and analysing the input, the key findings on one hand corresponded with parts of the reviewed literature in terms of the application of big data and on the other hand filled the identified gap regarding services supply chains. More specifically, as a tangible outcome an exemplary supply chain framework was created, based on the identified opportunities and threats regarding an implementation of big data as well as their critical evaluation.

The results ranged from chances for increased efficiency of background processes and improved effectivity of sales processes, leading to a greater profitability on one side, to being confronted with issues of privacy, economic viability and the adaptions caused by the better decision making stemming from enhanced insights through data science, predictive analytics and big data, on the other side.

In order to fully exploit these identified opportunities, further research within this field is recommenced, especially in regard to the increasing relevance of the sector and the mega-trend itself as well as the actual technical implementation.

Acknowledgements

First of all, I would like to thank my supervisor from the Newcastle Business School at Northumbria University who supported me vigorously and advised me constantly through the progress of this dissertation.

Furthermore, I would like to show my gratitude to the examined organization and its employees for their willingness to participate in the interviews, their time commitment and help for this research.

Finally, I would like to appreciate my family’s and friend’s understanding and emotional support.

Glossary

B2B Business to Business

B2C Business to Customer

Dr Doctor

DPB Data Science, Predictive Analytics and Big Data

GO Governmental Organization

IO Investigated Organization (Focus of this case study)

IT Information Technology

KPI Key Performance Indicator

NGO Non-Governmental Organization

S&P 500 Standard & Poor's 500; an American stock market index

SC Supply Chain

SCM Supply Chain Management

USP Unique selling proposition

List of figures

Figure 1 Conversion of data to information and to knowledge

Figure 2 Sources of Big Data

Figure 3 The house of research methodology

Figure 4 Process of Deduction and Induction

Figure 5 Result from the analysis regarding the product identification

Figure 6 Average Services Supply Chain

Figure 7 Word cloud of opportunities

Figure 8 Life cylce as sales platform

Figure 9 Word cloud of threats

Figure 10 Applications of Analytics across the Supply Chain

Figure 11 Dependency of effectiveness on domain knowledge and analytical skills

Figure 12 Framework for implementing big data on the supply chain

Figure 13 Maturity Map for implementing big data on the supply chain

Figure 14 Methodological Pyramid

Figure 15 Research Onion

Figure 16 Informed Consent Form for research participants (1/2)

Figure 17 Informed Consent Form for research participants (2/2)

Figure 18 Research Organization Informed Consent Form (2/2)

Figure 19 Research Organization Informed Consent Form (2/2)

Figure 20 Supply Chain Analysis

Figure 21 Big Data Analysis

Figure 22 SWOT Analysis

Figure 23 Investigated Organization Customer Journey

List of tables

Table 1 Issue regarding credibility & limitation

Table 2 Potential risks and preventive measures

Table 3 SWOT

1 Introduction

1.1 Research Topic, Relevance and Focus

“There is nothing permanent except change.”

With this statement, Heraclitus of Ephesus (n.d.) implied that we need to be able to adapt. Applying his credo to a business context of general and personal relevance, one can observe that the financial service industry has been impacted negatively by the financial crisis and more recently by the low interest environment. The competition through banks, strengthened by their access to cheap refinancing, and existing peers, with leaner processes, increases(Gelfarth, 2015). Affected companies need to adapt to this change in the market. Therefore, the overall aim must be to increase profitability through the sub-targets of decreasing costs while increasing the revenue, in order to defend one’s market position and grow sustainably.

A possible way to reaching the goal could arise from a current trend regarding technological advancement. In fact, the life of each and every one of us has been shaped by industrial revolutions. The first one enabled us to use water and steam to power mechanical production facilities, the second one gave us electricity which made mass production possible and the third one supplied us with IT, computers and the automatization of processes (Marr, 2016). Now we are facing the fourth industrial revolution labelled as Industry 4.0 by the German federal government(BMBF, n.d.). It consists of the internet of things and goes as far as machines communicating with each other using cyber-physical systems and cloud computing. However, all of those advancements are rooted in the new ways of collecting, storing, transferring and especially analysing very large amounts of data, commonly referred to as Big Data (Lee, Bagheri, & Kao, 2014).

Prof. Klaus Schwab, founder and executive chairman of the World Economic Forum, argues that this fourth revolution will be and already is fundamentally different from the previous three. According to him, we do not only face advances mainly in technology, but for the first time in a combination merging physical, digital and biological areas and effecting, once again, the life of each and every one of us in a way we have not experienced it before(World Economic Forum, n.d.).

It revolutionized our private lives and how we live, from owning self-replenishing refrigerators (wiseGEEK, n.d.; Bennett, 2009), over smartphone payment and virtual reality to monitoring the own body and self-driving cars(Lohr, 2012). However, big data also impacts us against our will and with us being unaware of it. The mega trend was for instance used to influence a democratic political process as seen on the election of President Trump with big data driven digital marketing (Chester, 2017) and once again was Sir Francis Bacon proven right, who said in 1597 that “knowledge is power”(BrainyQuote.com, n.d.).

Those examples illustrate not only how big data connects various subject areas, but also that the possible scope of impact is too large to be completely covered in this dissertation. Therefore, the author wants to focus on the economic perspective and analyse what effects big data and associated components have on a particular segment of business, in this case the struggling financial services sector(O’Dowd, 2017).

1.2 Research Objectives

Based on the aforementioned explanation the following research question was formulated: What are the effects of DPB (data science, predictive analytics and Big Data) on the supply chain and the decision making of companies in the financial service sector in Germany, elaborated on the example of a key player of this very market?

An answer shall be found through the structured and holistic approach of setting and achieving the following goals, arisen from the deconstruction of the research question:

1. Addressing the issue of the big data’s bandwagon effect (see Appendix 6.1) by developing and communicating a correct and common understanding of the term and the context(Shankar, 2015).
2. Displaying the gap in literature in terms of services supply chains and filling it by creating and providing a generalizable example based on the examined organisation.
3. Presenting evaluated findings of the analysis on the effects big data has on this exemplary services supply chain and the associated decision making (process).

1.3 Structure

The structure of the dissertation orientates itself on the four frameworks approach, a concept specially developed for a purpose such as this dissertation with a business topic and characterized by simplicity and versatile possibilities of application(Quinlan, 2011). Applying the scheme onto the outline of the chapters, one observes that the conceptual framework is contained in the first chapter by introducing the reader to the topic, justifying its selection and developing a precisely formulated research question, followed by the theoretical framework being contained in the literature review and concerned with underpinning the research with appropriate theory. The methodological framework occupies the third chapter by giving guidelines regarding the content and the technique to gather the research. The final parts include the analytical framework containing the data analysis and findings as well as the conclusion and therefore the answer to the research question.

2 Literature Review

2.1 Overview

This chapter aims to provide the reader with an introduction into the terminology of DPB as well as its development and importance. Furthermore, the topic of SC, its management and the influence of trends will be examined. Based on this knowledge the author aims to ascertain and display the gap in theory through the review of up to date academic literature, which is soon to be filled by means of this undergraduate dissertation.

2.2 The Terminology

Almost everything we do nowadays is either trackable or already being tracked. This development combined with the fact that this data became much cheaper to access and store (Bradbury, n.d.; Sanders, 2016) leads to the fact that we generated almost as much data in the last two years as in the entire history of mankind(IBM, n.d.). This phenomenon as well as the attempts to generate value from it can be summarized under the abbreviation DPB, representing the three main parts data science, predictive analytics and big data. One could argue that the nomenclature is debatable since predictive analytics stem from data science which then again rely on big data(Maurer, 2015). To be able to comprehend this statement one needs to be aware of the exact and underlying definitions of the terms data and information, since those are crucial for understanding the processes and the value of big data. Whitney (2007) provides a comprehensible definition of the process from data to knowledge and how those terms can be defined. As one can observe from the illustration below, data is the input but knowledge is the output one actually and ultimately aims for when applying big data.

Abbildung in dieser Leseprobe nicht enthalten

Figure1Conversion of data to information and to knowledge Illustration adapted from Whitney (2007)

2.3 Big Data

Bachmann, Kemper, & Gerzer describe big data as the phenomenon of very large, exponentially growing amounts of data high in variety and context (2014). However, the absence of a uniform definition perfectly illustrates the reason why society is unable to estimate the effects this megatrend has. What academia agrees upon though, are the sources of the high volume data, which can be clustered in five main categories and are displayed in the illustration below.

Abbildung in dieser Leseprobe nicht enthalten

Figure 2 Sources of Big Data Illustration developed by the author based on George, Haas, & Pentland (2014).

Furthermore, big data brings along issues regarding ethics and security. Data sharing is a very important part of big data infrastructure, hence George, Haas, & Pentland (2014) call for more protection of privacy and more control when handling sensitive data, one of the reasons why all categories from the figure above except for public data are relevant for this dissertation. Very often the data goes back to individuals which need to be protected from unwanted parties intruding their privacy and constricting their freedom illegally, especially since big data grants deeper insights than ever before (Warren & Brandeis, 1890).

After understanding where the data stems from it is necessary to understand what makes big data unique. One of the many, but a leading definitions describes it as the following: “Big data is high-volume, high-velocity and high-variety information assets that demand cost-effective, innovative forms of information processing that enable enhanced insight, decision making, and process automation”(Gartner, Inc., n.d., p. 1). Deconstructing this description, we find the three Vs of big data, which are volume, velocity and variety(Waller & Fawcett, 2013).

Volume describes the amount of data at rest, which is increasing faster than ever before at a high and continuously, exponentially increasing rate so that by the year 2020 (Marr, 2015), there will be about 1.7 MB of data be created - every second - for every single one of the forecasted almost eight billion human beings on the planet(Worldometers, 2017).

Secondly, velocity describes the data that is in motion. This consists of two aspects, on one hand the velocity regarding how much data is being processed over how much time. The systems responsible for those actions have been modified to such a level of efficiency that data can be processed in real time(Stedman, 2016). On the other hand, velocity concerns the speed of change in meaning and relevance. The older the data is, the more it loses value e.g. GPS tracking.

Thirdly, variety delineates the many forms of data. It can occur as structured, unstructured, text, numbers or multimedia. Bachmann, Kemper, & Gerzer argue that those three Vs are an insufficient description of the uniqueness of big data (2014). According to them veracity or the truthfulness and correctness of the data that is at doubt is an important addition, since there is uncertainty due to data inconsistency or incomplete model approximations. IBM agrees up to this point by displaying similar findings in their visual analysis (IBM, n.d.).

However, the preceding authors add a final V, partially anticipating the goals of exploiting big data, which is value. All the data that is being collected, stored and analysed must provide added value in order to make the whole project profitable. Revising the beforementioned definition, this shall be achieved by making more intelligent decisions based on a better information base with enhanced insights as well as a reduction of cost and failure in addition to increased efficiency through process automation.

To possibly answer the question of how this goal shall be reached, Gartner Inc.’s (n.d.) description is reviewed one final time resulting in the hypothesis that innovative forms of information processing are required which then again would fall into the second part of DPB, which is data science.

2.4 Data Science

In recent years increasing attention came to this field of innovative forms of processing data, more specifically the scientific analysis of data(Google Trends, 2013). While analytics are in this context defined as machine learning techniques(Lehmann, 2017), data science is introduced as “the application of quantitative and qualitative methods to solve relevant problems and predict outcomes”(Waller & Fawcett, 2013, p. 78). In other words, it is the process that converts big data into a far more valuable source – smart data or knowledge.(George, Haas, & Pentland, 2014). Self-evidently, this is nothing new considering intelligence in business has been used since the 1990s. However, its subpart business analytics and later big data analytics evolved through the need for a unique data handling, since the data became too much and too complex to be handled by conventional analytics(Chen, Chiang, & Storey, 2012).

Big data therefore changed the game, as Fawcett & Waller (2014) put it, of how to gain a competitive advantage based on one’s supply chain. In contrast to conventional analysis where one asks a question and looks for an answer, in big data analytics one does not have an exact question but looks for patterns and relationships instead, for a reason one might uncover subsequently(Bachmann, Kemper, & Gerzer, 2014). Sanders (2016) agrees in his seminal work on the point made and labels it a change in the nature of inquiries.

However, she recognizes two more important changes. First, the change in the opportunity for inquiry, stemming from the velocity of data capturing in almost every field has enabled real time data analysis. Secondly, the change in the nature of experimentation. This goes back to increased accessibility especially through the internet(International Telecommunications Union, 2007), enabling large scale experiments analysing many social and economic phenomena.

2.5 Supply Chain and Supply Chain Management

By reason of the supply chain, from now on referred to as SC, having a strategic role that cannot be overemphasized(Coyle, Langley, Gibson, Novack, & Bardi, 2009)the key to reaching the set objective is rooted in its optimization. It is described as the system or process, which relocates the product or service from the supplier to the consumer or customer, which includes the allocation and transformation of various resources into a final product(Melzer-Ridinger, 2012). It is important to be aware of the distinction between supply and value chain and that SCs in fact link multiple value chains(Nagurney, 2006). Oehlrich (2012) defines the latter as a set of essential activities, which an organization performs in order to deliver product or service. The value chain represents the total value of the company and is composed of the value activities, in this context the physical and technological activities from which a company creates a valuable product, and the profit margin.

The management of this very supply chain, in this dissertation from now on abbreviated as SCM, is defined as “the systemic, strategic coordination of the traditional business functions within a particular company and across businesses within the SC, for the purpose of improving the long-term performance of the individual companies and the SC as a whole”(Mentzer, Myers, & Stank, 2006, p. 5). An excellent design of the organizations SC as well as appropriate strategies and operations within the SCM lay the groundwork for success and advantages over competitors, as illustrated on the example of Walmart(Kozlenkova, Hult, Lund, Mena, & Kekec, 2015).

2.6 Applied SCM Data Science

The synthesis of the aforementioned topics results in SCM data science, an application of various quantitative and qualitative analytics on frameworks of SCM to solve relevant problems and predict certain outcomes on the basis of big data.(Waller & Fawcett, 2013). The latter topic of predictive analytics will be referred to later in the literature review. Sanders (2016) goes into more detail, dividing the SC into parts to which such analytics could be applied to and gives examples of what such could focus on (See Appendix 6.2). She states that the largest growth has been experienced on the marketing side, continuously gaining better knowledge of the market and its participants. Its applications are on the sell side of the SC that revolve around the customer(Lohr, 2012)and may result in better adjustment of strategy and operations leading to higher margins(Sanders, 2016).

Nonetheless, an important issue raised by Waller & Fawcett (2013) is the fact that effectiveness in terms of data science can only be achieved when the human resource, known as data scientist, can combine domain knowledge of the specialized field of application with his/her breadth of analytical skill set (See Appendix 6.3). This results in a high demand for the sexiest job of the 21st century (See Appendix 6.4) but insufficient supply since there are no degrees in data science being offered by universities(Davenport & Patil, 2012). As a consequence, those specialists are high in difficulty and cost to hire and retain(Press, 2015).

Another issue is the not slowing pace at which big data advances. If companies do not keep up with the trend due to a lack of talent, they will lose market share to their competitors because of them gaining almost unassailable advantages(Davenport & Patil, 2012).

2.7 Implementation of SCM Data Science

From many studies conducted on companies that successfully implemented those practices one can extract certain guidelines. Sanders (2016) structured those approaches in a framework for implementation (See Appendix 6.5). The gist of it however goes back to three key lessons. First, it is inevitable to coordinate and link the applications of big data across the entire SC. It is not conducive to analyse a single function or solve an isolated problem. Second, their course of action is not random and driven by the search for coincidences, but fully in line with their strategy developed in beforehand, resulting in specific tactical actions.

Third, performance needs to be measured constantly using appropriate metrics and KPIs to efficiently engage in taking measurements as part of continuous improvement process, inspired by Kaizen, the Japanese management philosophy(Prošić, 2011). Fourth, experience shows that the implementation needs to mature in certain steps. This evolution is comprehensibly illustrated in a maturity map (See Appendix 6.6). It structures the process by not starting disorientedly with an aimless analysis, but to prepare the data and the digital infrastructure bevor initiating the project. A successful implementation may culminate in predictive analytics or automated algorithms, which directly increase the SC’s efficiency or discover new forms of value generation(Ericsson, 2014).

2.8 Predictive Analytics

Predictive analytics which have been mentioned many times up until this point are a subset of data science. They are defined as “the use of statistical or machine learning methods to make predictions about future or unknown outcomes”(Brown, Abbasi, & Lau, 2015, p. 6). Those conclusions are being drawn relatively quickly and inexpensively through the approximation of relationships between variables in combination with mathematical methods. If those are applied as part of the maturity map, SCM predictive analytics are created, which are defined as the synthesis of the use of “both quantitative and qualitative methods to improve supply chain design and competitiveness by estimating past and future levels of integration of business processes among functions or companies, as well as the associated costs and service levels”(Waller & Fawcett, 2013, p. 80).

That predictive analytics are a driver of change shall be outlined using two examples(Fawcett & Waller, 2014). On one hand, one has to get accustomed to the fact that decision will be driven by correlation and not causality. This goes back to predictive analytics providing great insight, but changing a general decision making paradigm to knowing what, but not why, since the algorithms are far too complex and automated for decision makers or effected parties to understand. On the other hand, and arguably far more important is the development that tracking data enables profiling, which then again enables prediction(Lohr, 2012). Through the collection of data and its analysis organizations are able to create customer profiles and identify certain routines in behaviour which allow them to predict future moves with a very high probability(Fawcett & Waller, 2014). This goes as far as e.g. Amazon introducing anticipatory shipping. Their patent claims to enable them to ship products to customers, who are expected to order this item soon, before they have even placed an order, based on his/her shopping pattern(Welch, 2014).

2.9 Financial Sector

A sector that falls in the category of allowing more research to be conducted on is the financial services sector. The importance of the financial markets stems from the fact that they perform an essential economic task by transmitting financial funds from parties where the ROI is rather low to such which can achieve a higher return. Thereby, they contribute to higher production and a higher efficiency leading to more welfare for in the overall economy and society(Crockett, Harris, Mishkin, & White, 2003). The finance sector’s market weight increased by 123% in the last five years making up for almost 15% of the S&P 500 Index, compared to the real estate sectors market weight of not even 3%(Fidelity, 2017).

Focussing on the sub-category leasing, as this is the core competence of the examined organization, there are a few features one needs to be aware of when analysing the supply chain, which is the USP of leasing that makes it such a promising industry. Firstly, it is characterized by the pay-as-you earn model, which provides one with financial flexibility and instead of the need to commit a significant amount of liquid funds to an investment. Furthermore, the lease instalments are tax deductible and will improve the balance sheet as well as the capital structure represented by numerous key figures such as the equity ratio of the affected company when compared to traditional buying. Lastly, the short leasing cycle of an average of 48 month give the customers the chance to stay up to date with the newest technology and equipment (Confidential information of the investigated organisation, 2015).

2.10 Services Supply Chain

Even though there is a sufficient number of articles examining DPB, contrastingly there is an insufficient amount of literature explaining its effects on the SC. It goes as far as having journal authors invoking their readers to conduct more research on the topic and even providing them with research questions. This has been the case in Waller & Fawcett’s (2013) journal article particularly focussing on this matter and contributing signifacntly to the research aim. More specifically and in view of the targeted (financial) services sector, various literature agrees that there has not been engaged enough in research on this special form of the SC, known as the services supply chain(Ellram, Tate, & Billington, 2004). Breidbach, Reefke, & Wood (2015, p. 2) agree on this statement, adding that it is also necessary to do so, since “service supply chains are traditionally perceived as distinct from goods-centric or ‘product’ supply chains (…). Thus, attempts to transfer insights gained from ‘product’ to service supply chains have not been particularly successful“.

In order to examine the effects DPB may or may not have on the SC of a financial service provider, one needs to analyse its SC in beforehand (Dietl, n.d.). Without this step data scientists will be unable to generate value through the application of advanced analytics. However, an issue that one needs to bear in mind is the complicatedness. It is fair to assume that services SCs are higher in complexity than manufacturing ones, at least when it comes to transferring findings across industries(Digabit, 2015). Metaphorically speaking, it may appear simpler to compare a company that manufactures wooden chairs with one that manufactures automobile chassis in terms of their basic SC elements than it would be for service providers from the financial and the health care sector(Athens University of Economics and Business, n.d.). This may restrain the degree of generalization for possible findings of this dissertation across various services sectors but not within the financial one.

2.11 Summary

Summarizing, the trend of big data is of high and continuously increasing importance. Based on organizations need to generate value from this trend, various new ways of analysing have been developed. Data science carried out by highly demanded and highly skilled data scientist is aiming at improving SC processes and its sub-category predictive analytics has the objective of identifying and exploiting new market opportunities. However, in terms of the (financial) services sector there has been insufficient research conducted on the SC and consequently on the effects of DPB on the SC of such organizations. Bearing the issues of comparability as well as ethics and security in mind, the gap identified by reviewing the literature is aiming to be filled by the means of this dissertation driven not only by curiosity and the need for a complete coverage through literature but also by the goal of exploiting advantages and possibilities of generating more value.

3 Methodology

3.1 Introduction

The term specifies the approaches and the underlying theory of how research should be conducted(Quinlan, 2011). In this chapter research methodology will be examined and critically evaluated regarding its application to achieve the research goal of this dissertation. This will be carried out a structured approach, grounded on Quinlan’s (2011) methodological pyramid (see Appendix 6.7), consisting of fundamental philosophies on the bottom, research methodologies in the middle and data collection on top. However, one could argue that by over-simplifying a complex topic to only three major aspects, a large amount of information might be overlooked. Therefore, the author combined this framework with Saunders, Lewis, & Thornhill’s (2006) concept of the research onion (see Appendix 6.8). Arguably, it might represent the other extreme view, however this time through over-complexification and including too many aspects into one scheme. As a result, the house of research methodology was created.

Abbildung in dieser Leseprobe nicht enthalten

Figure3The house of research methodology Visualization developed by author

3.2 Approach

The first layer of the house of research methodology are the different approaches to execute research. Shown in the illustration below are the two major approaches, the deductive and the inductive approach. Introducing and contrasting those two very briefly, one could argue that the former is for testing a theory with the mantra data follows theory, and that the latter is for building theory, consequently with the mantra theory follows data(Bryman & Bell, 2015).

Abbildung in dieser Leseprobe nicht enthalten

Figure4Process of Deduction and Induction Illustration developed by author based on Min (2016)

The research for this dissertation will be conducted by means of the inductive approach, because the author aims at building theory about the effects of big data on the SC, by observing and interviewing in order to identify patterns and formulate a sound thesis. Moreover, it is suited for collecting qualitative data by performing interviews, not only to gather information about the facts regarding the mega trend as well as the SC, but also to gain knowledge about the value humans attach to this topic. The provided flexibility is beneficial for the anticipated research method, allowing adaptions along the research process to readjust the scope and reach a higher level of data granularity(Oates, Kelley, & Barbusinski, 2002). Crucial to reaching the research goal is the realisation of what will be discussed in the following chapters, which are the researcher’s role, motivation and expectations of him. Induction endorses the circumstances that the author is part of the research process due to his employment relationship and that this amongst other things biases the research towards delivering predominantly practical or pragmatical results with a generalizable character.

3.3 Design

This chapter covers the universal plan of how the research will be conducted, for which the foundation was laid by the prior examined theory and especially the research question(University of Southern California, 2017).

The two approaches, which are relevant to this scholarly thesis identify to be of explanatory and exploratory kind. In fact, a synthesis of those two will be applied as the research design, since both feature characteristics that correspond with the research aim. The prior, explanatory studies, look for casual relationships between variables. Applied to our context this might imply relations between SC processes and data analysis processes or it might address the core of big data, which are the connections and correlations between data leading to unforeseen insights and new possibilities. The latter, exploratory studies, aim at clarifying a problem, gaining a better understanding and creating new knowledge by either examining literature or performing interviews. It is argued that its major advantages are the flexibility and adaptability to change. This is very useful, especially when carrying out interviews, since new and unexpected input can influence the direction of the research. This development allows the research to develop from initially a rather broad focus to a continuously narrower scope.

Applying a design, which is created by a merger of the two introduced studies will help achieve the research goal. This is due to the fact that the author is not limited to the strategies and methods allocated to one of the designs but can rely on a variety of advantages and ideas gained from this unique scheme.

3.4 Strategy Role

The strategy of the case study is exceptionally suited for explanatory and exploratory studies, based on its frequency of use (Saunders, Lewis, & Thornhill, 2006) and especially its popularity in business research(Bryman & Bell, 2015). It focuses on researching a certain phenomenon within its real-life context by relying on multiple sources for information and is targeted at generating a meticulous understanding of the subject, in this case the identification of the buzzword big data and its effects.

Additionally, one needs to consider the researchers position, an issue raised before. Being employed by an organization, in this case as a part-time student, can uncover major advantages and disadvantages. For instance, it may ease the hurdles of accessing data but also influence the researcher’s goals and motives. The opportunity of gathering more and better information is opposed by the problem of lacking seniority and time. This leads to the conclusion that being a practitioner researcher creates a certain frame for the implementation of a research strategy.

3.5 Methods

It needs to be recognized that one can use multiple methods either in terms of qualitative or quantitative data or mixed methods by collecting both(King & Horrocks, 2010). In this case, the environment and multiple factors constrain the dissertation to focus on qualitative data, in other words non-numeric information accumulated through techniques such as interviews. The research question is not focussed on delivering a numerical result but a qualitative evaluation. In addition to that can the reviewed literature and the accessible information in form of interviews be classified as qualitative. The detailed concept for the interviews will follow in the according chapter.

3.6 Credibility & Limitations

The credibility of academic work is highly important since it decides upon the value added through the dissertation which is self-evidently also a marking criterion on the part of the institution. Literature contemplates five major issues in this context which are displayed in the table below and will be addressed thoroughly in this dissertation.

Table1Issue regarding credibility & limitation Table created by author based on Bryman & Bell (2015)

Abbildung in dieser Leseprobe nicht enthalten

3.7 Ethics

In the process of conducting research, one encounters ethical issues at several occasions. In this context they are mostly associated with the behaviour one displays towards the direct or indirect subjects and stakeholders of the work, especially in relation to their rights. Ethics are described as a certain behaviour defined through moral standards, norms and principles, which direct moral decisions regarding our own behaviour and the relationship we maintain with others(Blumberg, Cooper, & Schindler, 2005). Research ethics therefore are the application of moral frameworks to every step of the research process. For this and for many other reasons the research process is fully in compliance not only with the standards of Northumbria University but also of the organisation under investigation, because a dissertation based on maleficent research cannot be considered credible and value-adding.

Furthermore, throughout the research process, the researcher has remained mindful of the four key areas of potential risk that are commonly associated with social science research:

1. Avoiding harm to all involved in or potentially affected by the research
2. Ensuring the anonymity of all participants/respondents
3. Gaining informed consent from all participants / respondents (see Appendix 6.9)
4. Avoiding deception

Accordingly, every effort has been made, with close guidance from the academic research supervisor, to eliminate, or as a minimum ameliorate, these potential risks throughout the processes of research design, data generation, data analysis and dissemination. All data has been stored securely, locked or encrypted, throughout the course of the research and will be destroyed upon completion

Specifically, the following actions were taken:

Table2Potential risks and preventive measures (Northumbria University, 2016)

Abbildung in dieser Leseprobe nicht enthalten

[...]