THE USE OF BIG DATA IN STRATEGY MAKING

THE USE OF BIG DATA IN

STRATEGY MAKING

A FRAMEWORK OF SUCCESS FACTORS AND BARRIERS

S2813743 | b.mooiweer@student.rug.nl UNIVERSITY OF GRONINGEN

FACULTY OF ECONOMICS AND BUSINESS MSc BA - SMALL BUSINESS AND ENTREPRENEURSHIP

SUPERVISOR: Dr. Ir. J. KRAAIJENBRINK CO-ASSESSOR: Prof. Dr. P.S. ZWART

JUNE 16TH, 2016 | WORD COUNT: 12.834

ABSTRACT

INTRODUCTION

In recent years, management has an increasing emphasis on big data, and due to its high operational and strategic potential, big data has recently also become the focus of academic and corporate investigation (George, Haas, & Pentland, 2014). The extant literature identifies big data as the “next big thing in innovation” (Gobble, 2013); the next “management revolution” (McAfee & Brynjolfsson, 2012); and that the big data is capable of changing competition by “transforming processes, altering corporate ecosystems, and facilitating innovation” (Brown, Chul, & Manyika, 2011).

Although the big data era has only just emerged, it can be a critical tool for realizing improvements, particularly in the manufacturing industry (McKinsey & Company, 2014). The manufacturing industry has been the backbone of many developed economies and was an early and intensive user of data since the dawn of the computer era (Dutta & Bose, 2015). Hence, big data have become an important factor of manufacturers today and an important factor of the increasing intensity with which enterprises are gathering information (Nedelcu, 2013). Within this paper, the industry of particular interest is part of the manufacturing industry: the metal manufacturing industry. Companies in this industry also employed big data for years (Sowar & Growley, 2011). Despite this experience, this sub-industry still has many challenges and opportunities. According to Sowar and Growley (2011) steel manufacturers have the opportunity to expand the use of analytics in managing their businesses and for strategic decision-making. Moreover they stated that increasingly, new analytics tools are giving management new insights and levers to address these continuing challenges. Therefore, the (metal) manufacturing industry is an particularly interesting industry to analyze the use of big data, and to give a deeper insight into cause and effect of big data and help to optimize industrial processes.

success remain. On top of that Davenport (2014) stressed the importance of organizational barriers to big data, such as resistance from key stakeholders and a lack of experience with analytics.

Thus, regarding both the field of big data and strategy there is a relatively wide range of literature available concerning the relationship between these issues. Authors stress the use of big data but, however, do not agree whether big data leads to opportunities or limitations. Moreover, and more important, the authors give an indication of possible success factors and barriers but are unable to reach consensus as regards to the precise success factors and barriers related to the opportunities and limitations of using big data in strategy making. Additionally, these barriers and success factors are focused on industries in general without focusing on the metal manufacturing industry or any other industry in particular. However, the (big data) characteristics of industries strongly differ (Manyika et al., 2011). Hence, literature fails to express the precise success factors and barriers that are distinctive for strategy making in the metal manufacturing industry.

This paper will contribute to bridge the existing knowledge gap by drawing upon and combining prior “big data” studies, and especially contribute the literature by clarifying both the success factors and barriers of using for using big data in strategy making in the metal manufacturing industry. On top of that, due to its high operational and strategic potential big data - and the use of it - becomes more and more a topic of great interest to managers. Thus, issues concerning the consequences of big data need research beyond the limelight of the hype (Constantiou & Kallinikos, 2015). Indicating success factors and barriers, will help managers to use big data in such way that it is beneficial for the company’s strategy making and performance. Therefore, the research question of this paper is:

“WHAT ARE THE SUCCESS FACTORS AND BARRIERS FOR USING BIG DATA IN STRATEGY MAKING, WITHIN THE METAL MANUFACTURING INDUSTRY?”

LITERATURE REVIEW

In this chapter, the theoretical background of this study is discussed. Background information is provided regarding big data in general, big data in industries, big data in the manufacturing industry, strategy, big data used in strategy making, and success factors and barriers for using big data in strategy making.

BIG DATA

According to Moorthy et al. (2015), in 2003 F. Diebold was one of the first to use the term “big data” by stating: “Big data refers to the explosion in the quantity (and quality) of available and potentially relevant

data, largely the result of recent and unprecedented advancements in data recording and storage technology” (Moorthy et al., 2015, p. 75). Regarding the composition of big data, Russom (2011) stated

that big data can be an eclectic mix of structured, unstructured, semi-structured, and streaming data. To define big data, several authors use the “3Vs”: Volume, Velocity, and Variety (Kwon and Sim, 2013; McAfee and Brynjolfsson, 2012; Russom, 2011). According to Russom (2011) “Volume” is the primary attribute of big data, and refers to the amount of data that either consume storage or entail a number of records; “Velocity” can be described as the frequency or the speed of data generation and/or frequency of data delivery (Russom, 2011); the term “Variety” is used to highlight the fact that data are generated from a large variety of sources and formats, and contain multidimensional data fields including structured and unstructured data (Russom, 2011). In addition to the 3V’s, IDC (2012) and Oracle (2012) include “Value”. The “Value” component must be integrated in order to stress the importance of extracting economic benefits from the available big data (Wamba, Akter, Edwards, Chopin, & Gnanzou, 2015). White (2012) suggested that even a fifth dimension “Veracity” should be added to prior definitions in order to highlight the importance of quality data and the level of trust in various data sources.

S

Structured data is information, usually text files, displayed in titled columns and rows, which can easily be ordered and processed by data mining tools (Sherpa Software, 2016). This definition corresponds with the definition of Kaisler, Armour, Espinosa, and Money (2013), as they state that structured data refers to data that is well defined and is (often in tables) stored in relational databases. As regards to unstructured data, Gartner (2013) defines unstructured data as content that does not conform to a specific, pre-defined data model. It tends to be the human-generated and people-oriented content that does not fit neatly into database tables. Within the enterprise unstructured content takes many forms, such as business documents (reports, presentations, spreadsheets etc.), email and web content. According to Kaisler et al. (2013), unstructured data suitable for analytics can impede end-to-end processing performance.

I

Internal data is the data that is created inside the four walls of a business (Ernst & Young Insights, 2014). Most early big data efforts are targeted at sourcing and analyzing internal data. Research of Schroeck et al. (2012) indicated that more than half of the respondents reported internal data as the primary source of big data within their organizations. Furthermore, Schroeck et al. (2012) described internal data as the most mature, well-understood and unique (external data is also available for others) data available to organizations. This is due to the fact that this types of data has been collected, integrated, structured and standardized through years of work, and by applying analytics, internal data can provide valuable insights. External data refers to data generated from outside the four walls of a business (Ernst & Young Insights, 2014). External data is often needed to incorporate more information (Guillen, Gustafsson, & Nielsen, 2008). According to Segelod and Jordan (2004) the use of external data sources is increasing. It varies by industry and geographically but the trend is obvious and worldwide (Roberts, 1995; Hagedoorn, 2002). It is generally assumed that small firms are more dependent on external knowledge acquisition, than are large companies, which have access to a greater variety of in-house knowledge (Rothwell & Dodgson, 1991). External sources of supplier data can furnish information on suppliers’ technical capabilities, financial health, quality management, delivery reliability, weather and political risk, market reputation, and commercial practices (Davenport & Dyché, 2013).

TYPES OF BIG DATA

TABLE 1.TYPES OF BIG DATA (KAPOW,2015)

Type Description Velocity (V1) Variety

(V2) Volume (V3) Internal External Structured (10) Unstructured (1) 1. Archived data

Scanned documents, statements, insurance forms, customer correspondence, etc.

V1: Low V2: Low V3: Low

Internal 2 = Unstructured

2. Docs XLS, PDF, email, WORD, etc.

V1: Low V2: Medium

V3: Low

Both 4 = Unstructured

3. Media Images, videos, audio, flash, live streams podcasts, etc.

V1: High V2: Medium V3: High Both 5 = Between 4. Business apps

Project management, marketing, automation, productivity, CRM, etc. V1: Medium V2: Low/Medium V3: Medium Both 8 = Structured 5. Public web Government, weather,

competition, health care service, public finance, etc.

V1: Medium V2: High V3: High

External 5 = Between

6. Social

media Twitter, Facebook, LinkedIn etc.

V1: High V2: Low/Medium V3: Medium/High Both 3 = Unstructured 7. Data storage

SQL, NoSQL, Hadoop, file systems, etc. V1: Medium V2: Medium V3: Medium/High Internal 9 = Structured 8. Machine log data

Events log, server data, apps log, audit logs, business process logs, etc. V1: High V2: High V3: High Both 10 = Structured 9. Sensor data

Medical devices, cars sensors, road cameras, assembly lines,

V1: High V2: High V3: High

BIG DATA IN DIFFERENT INDUSTRIES

As regards to the use of big data, Manyika et al. (2011) indicated some obvious differences across industries. For example, the computer and electronic products, and information sectors are sectors that already reap many benefits from big data in terms of productivity and revenue growth. In contrast, local service industries have experienced a sufficient lower level of growth. Due to differences across industries, the barriers to capture value from the use of big data are structurally higher for some than for others. Subsequently, the public industry (e.g. education) may have relatively higher degrees of barriers because of a lack of data-driven mind-set and available data. For the opposite reason, industries such as retail and manufacturing are expected to face lower levels of barriers to overcome. (Manyika et al., 2011)

BIG DATA IN THE MANUFACTURING INDUSTRY

Manufacturing is the process of converting raw material, components, or parts into finished good that meet a customer’s expectations or specifications. Manufacturing commonly employs a man-machine setup with division of labor in a large-scale production (Business Dictionary, 2016). According to Economy Watch (2010), the manufacturing industry refers to companies that involve in the manufacturing and processing of items and indulge in either creation of new commodities or in value addition. The final products can serve as a finished good for sale to customers or as intermediate goods used for production.

As indicated before, the manufacturing industry is a collection of many different industries. Within this paper, the sub-industry of particular interest is the metal manufacturing industry. According to Rabobank (2016), the metal manufacturing industry includes all forms of iron, aluminum and steel manufacturing, as well as forging, engraving, coating and stamping. Companies in the metal manufacturing industry are active as a supplier of operations, components and semi-finished products to other companies. These are both producers and suppliers of finished products (Rabobank, 2016). Companies in this industry employed advanced analyses for years; from process simulation to laboratory management systems to computerized maintenance management systems integrated with real-time production data (Sowar & Growley, 2011). Despite the multiple years of using big data in the metal manufacturing industry, this sub-industry still has many challenges and opportunities. According to Sowar and Growley (2011) steel companies have the opportunity to expand the use of analytics in managing their businesses and for strategic decision-making. Moreover they stated that increasingly, new analytics tools are giving management new insights and levers to address these continuing challenges. In addition, GE Intelligent Platforms (2012) stress the necessity that business and IT leaders have to ask themselves whether their industrial enterprise is maximizing the full potential value of their process data and using that insight to drive real-time improvements.

STRATEGY MAKING

phenomena that management is in charge of the process, but the emergent strategy makes the reference to the probably almost universal experience of strategy practitioners, that there are so many things that can intervene (Sminia, 2009). In other words, the strategy maker may formulate a strategy through a conscious process before he makes specific decisions, or a strategy may form gradually, perhaps unintentionally, as he makes his decisions one by one. Research on strategy formation focuses on a tangible phenomenon - the decision stream - and strategies become observed patterns in such streams (Mintzberg, 1978).

BIG DATA AND STRATEGY MAKING

Concerning the relationship between strategy making and big data, a number of important observations can be made. In the first place, due to disruptive characteristics of development associated with big data, the underlying assumption that organizational routines and capabilities built on cognitive patterns solidify the firm’s ability to adjust to the environment becomes problematic (Gavetti & Rivkin, 2007). Second, due to this volatile nature of the environment, long-term forecasts may be less relevant. Therefore the firm’s search mechanisms cannot be based on the application of a rational approach as suggested by the positioning school in strategy (Porter, 1996; Gavetti and Rivkin, 2007). Due to this discrepancy between the relatively homogeneous and structured assumptions of existing strategy tools and the heterogeneous nature of big data makes the concept difficult to interpret, especially when looking for new business opportunities (Constantiou & Kallinikos, 2015).

Despite the difficulties and discrepancies that can arise when using big in strategy making, big data is an important source to use in strategy making. Regarding the use of big data, it is argued that big data can either support a manager in making a decision or it can automate decision-making. On average is appears that big data is used for decision support 58% of the time, and 29% of the time it is used for decision automation (Economist Intelligence Unit, 2012). According to Provost and Fawcett (2013), the process of using big data to support strategy making refers to the practice of basing decisions on the analysis of data rather than purely on intuition. The role of data is a facet of the decision-making process, but it can be a critical one (The Economist, 2013). Moreover, McAfee and Brynjolfsson (2012) argue that due to big data, managers can measure, and hence know, radically more about their businesses, and directly translate that knowledge into improved decision making and performance. In particular, research of McAfee and Brynjolfsson (2012) indicated that companies in the top third of their industry in the use of data-driven decision making were (on average), 5% more productive and 6% more pro table than their competitors. Thus, using big data leads to better predictions, and better predictions yield better decisions.

ways to compete and win. They are transforming themselves to take advantage of the vast array of information that is available to improve decision-making and performance throughout the enterprise. Moreover, Woerner and Wixom (2015) described that big data offers rich, exciting opportunities to leverage and extend a company’s business strategy toolbox.

Contradicting to this opportunities-view of using big data in strategy making, authors also stress the limitations and possible barriers of big data. Woerner and Wixom (2015) mention the existence of these limitations by stating that they believe that studying how companies actually use volatile, heterogeneous, unstructured data to improve company choices and operations identifies actionable practices that will help companies overcome the limitations of big data. Furthermore, Constantiou and Kallinikos (2015) state that to a certain degree, individuals, communities and organizations will be able to draw on Big Data to improve their choices and operations, including strategy making and strategy implementation. However, they also state that the development regarding big data will keep on redefining the relationships and conditions between organizations and their environments. Therefore, issues concerning the consequences of big data need research beyond the limelight of the hype. On top of that, McAfee and Brynjolfson (2012) believe that, as with any other major change in business, the challenges of becoming a big data enabled organization can be enormous and therefore many barriers to success remain. Moreover, Davenport (2014) stressed the importance of organizational barriers to big data, such as resistance from key stakeholders and a lack of experience with analytics. Therefore, it is also important to find out what threats and barriers companies can face while using big data in strategy making.

SUCCESS FACTORS AND BARRIERS FOR USING BIG DATA IN STRATEGY MAKING

Authors indicate the existence of opportunities and challenges for using big data in strategy making. Therefore this part of the literature review will be focused on indicating the success factors and barriers associated with using big data in strategy making. Definition-wise, the success factors can be defined, as those activities and behaviors, necessary to ensure a successful implementation Wong (2009). Alazmi & Zairi (2003) define success factors as a restricted number of areas in which satisfactory results are guaranteed for a successful competitive performance. On the other hand, barriers can be described as the factors that have a negative effect on the success of implementation (Sigh and Kant, 2008). Within this paper the definition of Karabag (2010) is followed, by defining barriers as the factors that not only restrain a successful implementation but also prevent it.

SUCCESS FACTORS

to the heterogeneous nature of big data, big data is diverse and combines many different types of information. The prerequisite for analytical work is that data must be somewhat common across the enterprise, of high quality, and accessible in a warehouse or analytical mart. (2) Enterprise - A big data initiative should be unified across the entire organization. The most successful analytical competitors take an enterprise approach and therefore they don’t have disconnected silos of data, technology, or expertise. Over the longer term, analytical activity should not be ad hoc, but rather embedded in key business processes and systems for real-time application. (3) Leadership - People leading big data initiatives should be passionate and committed to the cause. They should not be afraid to experiment and make informed investments. With leaders who strongly support using big data in decision-making, companies can make rapid progress. (4) Targets - It is important to find specific targets where big data can make a difference to the business. It is impossible to focus on all areas of the business. The target area should support a key strategic focus of the business. (5) Technology - New architectures and options have emerged to support big data initiatives. For stimulating the effective use of big data in strategy making, organizations must consider new technological options. (6) Analytics - Deriving meaningful analytics from big data requires intervention by data scientists. In large companies, data scientists may have different roles and work in different functional areas, but everyone should be working toward the same goals. To make analytics work, analysts must prove their ability to communicate to leaders and link analytics to key decisions and impact on the bottom line. (Davenport, 2014)

effectively exploit them. (5) Company culture – To succeed, a companies data strategy has to fit the culture and thereby a organization must not asks itself “What do we think?” but “What do we know?” This requires a move away from acting solely on hunches and instinct.

When comparing the two frameworks of McAfee and Brynjolfson (2012) and Davenport (2014), it appears that there are some clear points of overlap. However on other perspectives, the frameworks identify different/additional success factors. Therefore table 2 has been conducted in order to illustrate the similarities among and differences between both frameworks.

TABLE 2.SUCCESS FACTOR GROUPING VARIABLES

Success factor

grouping variables McAfee and Brynjolfson (2012) Davenport (2014)

1. Leadership

• (1) Leadership - Companies succeed because they have leadership that set clear goals, define what success looks like, and ask the right questions.

• (3) Leadership - People leading big data initiatives should be passionate and committed to the cause.

2. Talent

• (2) Talent - Some of the most crucial complements of big data are data scientists and other professionals.

• (6) Analytics - Deriving meaningful analytics from big data usually requires intervention by data scientists.

3. Technology

• (3) Technology - The tools available to handle the volume, velocity, and variety of big data have improved greatly in recent years.

• (1) Data - To analyze, data must be somewhat common across the enterprise, of high quality, and accessible.

• (5) Technology - New architectures and options have emerged to support big data initiatives.

4. Decision-making

• (4) Decision-making - An effective organization puts info and the relevant decision rights in the same location.

• -

5. Targets • -

• (4) Targets - It is important to find specific targets where big data can make a difference to the business.

6. Organizational

• (5) Company culture - To succeed, a companies data strategy has to fit the culture and “What they know”.

BARRIERS

In contrast to McAfee and Brynjolfson (2012) and Davenport (2014), Capgemini (2012) did not indicate success factors but indicated the biggest impediments to using big data for decision-making. (1)

Organizational silos - The organizational silos that separate different types of employees often cause

problems. When employees interact poorly with people outside of their “silo”, it becomes difficult to do the work of the business. (2) Shortage of skilled people - To analyze the data properly, companies struggle to make effective use of unstructured data. Therefore, companies are seeking skilled analysts. However, it is unlikely that supply will meet demand soon, which will cause a “war for talent” as firms try and outbid each other for top-right data analysts. (3) Time needed to analyze large data sets - Executives indicate that they did not experienced a slowing of decision-making due to having to process large quantities of data. However, it is not getting faster either and so it still takes a considerable amount of time. (4) Not enough

data to support decision-making - 40% of the executives believe that the decisions they have made in the

past three years would have been significantly better if there was more data to support decision-making. (5) Big data management is not viewed strategically at senior levels - Capgemini (2012) stated at 55% of the respondents indicated this barrier. (6) High cost of storing large data sets, and (7) Data sets are too

complex to collect and store. Capgemini (2011)

In addition to the abovementioned impediments, Russom (2011) indicated the most important barriers to implementing big data analytics. These barriers are respectably: (1) Inadequate staffing or skills

for big data analytics, (2) Costs, (3) Lack of business sponsorship, (4) Difficulty of architecting big data analytic system, (5) Current database software lacks in-database analytics, (6) Lack of compelling business case, and (7) Scalability problems with big data. (Russom, 2011)

Moreover, the report of Accenture (2014) also illustrated multiple main challenges to implementing big data in companies. The top seven of these challenges include: (1) Security, (2) Budget, (3) Lack of talent to implement big data, (4) Lack of talent to run big data and analytics on an ongoing

basis, (5) Integration with existing systems, (6) Procurement limitations on big data vendors, and (7) Enterprise not ready for big data. (Accenture, 2014)

TABLE 3.BARRIER GROUPING VARIABLES

Barrier grouping

variables Gapgemini (2014) Russom (2011) Accenture (2014)

1. Lack of talent and skills

• (2) Shortage of skilled people

• (1) Inadequate staffing or skills for big data analytics

• (3) Lack of talent to implement big data • (4) Lack of talent to

run big data and analytics

2. Financial

barriers • (6) High cost of storing • (2) Cost, overall • (2) Budget

3. Difficulties in executing analyzes

• (3) Time it takes to analyze large data sets

• (4) Difficulty of architecting big data analytic system • (5) Current database software

lacks in-database analytics

• - 4. Organizational barriers • (1) Organizational silos • (5) Big data management is not viewed strategically at senior levels • (3) Lack of business sponsorship • (6) Lack of compelling business case • (7) Enterprise not ready for big data

5. Limitations of database and systems

• (4) Not enough data to support decision-making • - • (5) Integration with existing systems • (6) Procurement limitations on big data vendors 6. Complexity and scalability issues • (7) Too complex to collect and store

• (7) Scalability problems with

big data • -

STUDY PURPOSE

METHODOLOGY

The literature review shows the relationship between big data and strategy. However, this relationship received little empirical scrutiny. Only to a small extent authors described the relationship and indicated some possible implications of using big data in strategy making. Moreover, inconsistencies arise when comparing the frameworks and perspectives of authors. Therefore existing literature fails to appoint the precise success factors and barriers for using big data in strategy making. By re-exploring this first cycle of exploration of the relationship between big data and strategy making, a contribution to the literature can be provided. Thus, theory development is needed. According to Yin (2014), exploratory research is useful when a specific field of research is underdeveloped, which is the case in this study.

DATA COLLECTION AND SAMPLE SELECTION

Due to the relatively limited literature available, the observation of the phenomenon is done by primary data (interviews). Based on the literature review, a basic framework of potential success factors and barriers is formed. Subsequently, the interview data provide insights in the real-life framework and provides input to adjust and redefine the framework. As regards to the collection of qualitative data, semi-structured interviews are used. According to Dicicco-Bloom and Crabtree (2006) a semi-semi-structured interview is frequently the only source in a qualitative study. The method of semi-structured interviews include a set of predetermined open questions, but also gives the interviewer the freedom to go deeper into the subject by asking questions based on the interviewee’s answers (DiCicco-Bloom & Crabtree, 2006).

SAMPLE SELECTION

Within the literature review, diverse mechanisms, roles, and factors concerning the use of big data in strategy making are discussed. Following this line of thought, semi-structured interviews are conducted with two different groups of participants. In the first place the CEOs of metal manufacturers will be interviewed. The second group of participants consists of experts and consultant in the field of big data, such as companies that offer solutions to deal with big data. The two groups of participants represent a wide variation regarding their experience with big data, strategy making, as well as factors and barriers related to the topic, while allowing for the recognition of common patterns within and between groups.

In order to address participants, the method of snowball sampling is used. The reason of this is that snowball sampling can be applied as an informal method to reach a target population. The aim of this study is primarily explorative, qualitative and descriptive and therefore snowball sampling offers practical advantages (Hendricks, Blanken and Adriaans, 1992). Snowball sampling may simply be defined as: “A

technique for finding research subjects. One subject gives the researcher the name of another subject, who in turn provides the name of a third, and so on” (Vogt & Johnson, 2011, p. 368). Snowball sampling can

the social networks of identified respondents to provide a researcher with an ever-expanding set of potential contacts (Thomson, 1997). In addition, snowball sampling can produce in-depth results and can produce these relatively quickly. Obviously, this method also has its limitation (see limitations part).

After identifying the companies, the entry to the field is established in the first place by making telephone calls. During this call, companies are explained about the research project and are asked whether they are interested or not. Subsequently, emails with a formal explanation about the research project are sent, indicating the purpose of the study, and including an invitation to participate. These steps are taken as it is considered important that participants understand and accept the aims of the study and hence, willingly imparted information (Foddy, 1993). Moreover, to gain their trust and confidence in terms of revealing information, participants’ anonymity is promised, and a copy of the results is promised.

SAMPLE SIZE

Regarding the sample size of this research, the issue of “how many interviews is enough?” arises. Focusing on this issue, study of Baker & Edwards (2012) came to the conclusion that “it depends”. This can be related to the characteristics of qualitative research, as this type of research relies on subjective evaluations regarding the quality of information obtained from interviews in building a convincing analytical narrative (Mason, 2012). When the researcher determines additional interviews no longer reveal fresh insights, theoretical data saturation can be said to have occurred (Bryman, 2012; Gerson & Horowitz, 2002). It is at this stage that the researcher determines enough interviews have been conducted. Unfortunately, data saturation can only be known after the conduction and analyzes of interviews.

DATA ANALYSIS

To analyze the data the transcripts were processed within case first. The goal of a within-case study is to describe, understand and explain the phenomena researched in each single case (Miles, Huberman, & Saldaña, 2013). Each single case was put in a tabular display with the same codes as the interview guide in order to make a clear overview for the researcher. After that a cross-case analysis was performed. The advantage is that multiple cases can be compared, which adds to the generalizability of the study (Miles et al., 2013). To be able to analyze the rankings - regarding the importance of the success factors and barriers factors - in consistent and systematic manner, four groups of factors are indicated: key factors, additional factors, non-relevant factors, and regular factors. The groups are divided based on the following criteria:

1. Success factors and barriers are considered as “key factors” when:

• At least 50% of participants consider the success factor or barrier as one of the top 3 factors.

2. Success factors and barriers are considered as “additional factors” when:

• At least 50% of participants mention the success factor or barrier as an additional factor.

3. Success factors and barriers are considered as “non relevant factors” when:

• At least 50% of participants mention the success factor or barrier as not relevant.

4. When success factors or barriers do not fit any of the abovementioned criteria, it must be considered as “regular factors”.

QUALITY CRITERIA

SAMPLE

COMPANY PROFILES

Table 4 provides an overview of the company profiles. In the first place, four companies in the manufacturing industry in The Netherlands participated. The second group of participating companies consisted of four that provide software solutions for companies in the manufacturing industry in The Netherlands, and therefore must be considered as experts. The dates of the establishment ranged between 1978 and 2006. The sizes of the companies were determined based on the number of employees only, which counted between 10 and 89 employees. Additionally, the geographic area indicates that three companies are located in Flevoland, while the others are divided among several geographic areas.

TABLE 4.COMPANY PROFILES

Company Date of establishment

Geographic area

Employee

count Core activities

Case A 1980 Flevoland 89

Company specialized in drawing, producing and assembling steel constructions to serve industry and utility. The final products are buildings, skeletons, etc.

Case B 1981 Flevoland 65

Supplier of semi-finished products for machine manufacturers. Varying from simple metal plates to complex products like sub-assemblies.

Case C 1978 Flevoland 80

Supplier of semi-finished and complete metal products for mechanical engineering, equipment manufacturers, agricultural machinery and high-tech industry.

Case D 1956 Drenthe 38

Supplier of measuring solutions for the oil, gas, and nuclear industry. For these products, Case D conducts all actions related to the metal part of the products.

Case E 2006 Utrecht 21

Business Intelligence (BI) specialist and provider of solutions for organizations to access data in a quick and simple way, by creating useful dashboards.

Case F 1985

Noord-Brabant 10

Company that develops software solutions for steel construction companies. In order to offer solutions, Case F developed an own-made ERP system.

Case G 1982 Gelderland 85

ERP-software specialist in customer-driven manufacturing industry. What we do is create, build, sell, implement and maintain software.

PARTICIPANTS PROFILES

The profile and personal traits of the study participants are summarized in table 5. The participants were all males aged between 24 and 50 years. The level of education was not diverse because the participants had a Bachelor degree or a Master degree. Due to these degrees, the level of education can be considered to be relatively high. Regarding the experience of participants in the metal manufacturing industry, it appeared that it varies between 2,5 to 25 years of experience.

TABLE 5.PARTICIPANT PROFILES

Company Gender Age Education Role/function Experience

Case A Male 35 Master of Business Administration

(Operations & Supply Chain Management) General Manager 3 years

Case B Male 50 Bachelor of Science

(Mechanical Engineering) Director 25 years

Case C Male 47 Bachelor of Science

(Mechanical Engineering) Director 9 years

Case D Male 39 Bachelor of Commerce

(Commercial Economics Director 16 years

Case E Male 24 Bachelor of Business Administration

(Business Management) BI consultant 4 years

Case F Male 40 Bachelor of Business Administration (Business Informatics)

Chief Technical

Officer 17 years Case G Male 38 Bachelor of Engineering Product Manager 14 years

Case H Male 30 Bachelor of Engineering Business

Consultant 8 years

INTERVIEW GUIDE

table 6). By using this flow of the interview, the topics are sequenced from broad to narrow. This method of “funnel sequence” also goes for the questions of every category. By doing so the possibility that answers are affected by previous questions is reduced. Concerning the analyses of the interviews, table 6 provides some insights. Within this table, the steps of defining the information need/conceptual variables and the translation towards indicators and variables are indicated.

TABLE 6.OPERATIONALIZATION OF THE RESEARCH QUESTIONS

Name Set of objects Set of

values Indicators / Variables Question 1. Definition and

components of big data

How is big data defined and of which parts does

it exist

Nominal

• The first ideas about big data • Which of the 5V’s are appointed

(codebook X1 - X5) • What is part of big data

C1, C2 and C3

2. Importance, role and performance of big data (in general)

Level of importance, the role big data plays and

in which way big data influences performance

Nominal

• The importance of big data • The role of big data: internal vs.

external focus (codebook Y1 - Y2) • Influence on performance: now

and in the future

C4, C5 and C6

3. Attitude and plans about using big data in strategy

making

What are the attitudes and plans about using big data in strategy

making

Nominal

• Big data: problem or opportunity • (Increased) use of big data in

strategy making

D1 and D3

4. Use and responsibilities of big data in strategy

making

In what way is big data used in strategy making

and who is responsible

Nominal

• How is big data used for strategy making: support vs. automation (codebook Z1 - Z2)

• Responsible for big data strategy

D2 and D4

5. Success factors of big data in strategy

making

What are the success factors of using big data

in strategy making

Nominal / Ordinal

ranking

• Success factors indicated • Agree with framework, or add

and/or delete factors. • Ranking of success factors

E1, E2 and E3

6. Barriers of big data in strategy

making

What are the barriers of using big data in

strategy making

Nominal / Ordinal ranking

• Barriers indicated

• Agree with framework, or add and/or delete factors.

• Ranking of barriers

RESULTS

The purpose of this study is to provide an understanding for how big data influence the strategy making of companies within the metal manufacturing industry. In the first place the chapter is focused on indicating the participants’ view regarding the definition, components, importance, and the influence on performance of big data. In addition, an orientation towards the use of big data in strategy making and the responsibilities of the big data strategy is provided. Finally, and most important, the last part of this chapter is committed to the success factors and barriers of using big data in strategy making.

BIG DATA

DEFINITION OF BIG DATA

In order to analyze the participants’ definitions of big data, table 7 indicated similarities and differences. TABLE 7.PARTICIPANTS’ DEFINITION OF BIG DATA

Company Five V’s Definition

Case A Variety | Value | Veracity |

Data is widely available but is on itself not very useful, so you have to make it useful. By making it “Big”, companies do not only focus on their own part of the supply chain, but doing “more” with it.

Case B Volume | Value The total package of data that can be processed and used.

Case C Variety | Value | Veracity

The “vernetzung” between the diverse data systems and the advantage generalized by it.

Case D Variety | Value | Volume

Collecting the available information, and make it available throughout the organization. This can be very broad and much, so usability is important

Case E Volume | Variety | Value | Veracity

Many lines of information/data, more than a billion, that are measured in quantifiable tables. This data is mounted it from multiple sources. Data can come from anywhere, and by connecting this; new and valuable insights can emerge.

Case F Volume | Velocity Fast accessible data storage

Case G Volume | Variety | Value | Veracity

The combination of high volumes of data, from multiple sources to make it useful

As table 7 shows, a strong similarity is that most participants make the translation towards the use and advantage (Value) of big data. For example, participant A defined big data as: “Data is widely available

but is on itself not very useful, so you have to make it useful”. In addition, participant G also stressed that

the eventual goal of big data is to make it useful (value) as he defined big data as: “The combination of

high volumes of data, from multiple sources to make it useful”. However, also clear differences arise. On

the one hand participant C illustrated the importance of “veracity” by defining big data as: “The

‘vernetzung’ between the diverse data systems and the advantage generalized by it”, while participant F is

more focused on “velocity”, as he defined big data as: “Fast accessible data storage”. With respect to the five V’s, it appears that none of the participants defined big data in such terms that all five V’s are included. However, due to the differences in definitions, in total all V’s are mentioned at least once. Interestingly, the V’s of “Value” and “Veracity” are also appointed frequently. This feeds the suggestion that only using the “3Vs” would not be sufficient in this specific industry. Hence, to define big data in the metal manufacturing industry all five V’s must be included. Adding “Value” confirms the theories of IDC (2012) and Oracle (2012), while the addition of “Veracity” is in line with the theory of White (2012).

COMPONENTS OF BIG DATA

Regarding the components of big data: archived data, docs, media, business apps, public web, social media, data storage, machine log data, and sensor data (see table 1), it is striking that although none of the participants’ big data definitions can be seamlessly interchanged with the definition of others, each participant agrees that the submitted components of big data together form “big data”. This unanimously agreement of participants contributes to the suggestion that everyone implicitly knows what big data is, but when asked to define big data it appears to be difficult to give a comprehensive and explicit definition.

IMPORTANCE, ROLE AND INFLUENCE ON PERFORMANCE OF BIG DATA

As regards to the importance of big data to companies in the metal manufacturing industry, the answer is simple: it is an important component that already has improved the overall performance of companies in the metal manufacturing industry. This positive effect on performance, however, does not mean that all companies already reap the full benefits of it. Among others, participant D clearly indicated this issue by stating: “The data that we have analyzed (which is not all data), has led to increases in efficiency and

performance”, but he also stressed that: “At this moment, however, we still are at the early stages of using big data the right way. The desire is there, but it is still a long way”. In line with this, participant H also

stressed this process as he stated: “It IS very important. But in this industry it is, in some way, an

underdeveloped area. Companies in this industry do not reap the full benefits of it (yet).”

participant E indicated: “Especially business applications, machine log data and sensor data are

important components of big data in the metal manufacturing industry”. On the contrary, (only)

participant C stresses the additional use of external data sources: “Not only the internal side is important.

Within our supply chain we strongly focus on using all (external) data available to use it in such way that processes can be executed in a more efficient way”. In order to explain the strong internal focus,

participants A and F connect the internal use with the conservativeness of the industry. According to participant A is, due to conservativeness, data like customer data is not really measured. Participant F, on the other hand, connects the conservativeness with companies’ preference to investing in machines instead of software. These explanations are not in line with what Rothwell and Dodgson assumed in 1991, as they primarily connected an internal data focus with the size of companies. Due to lack of evidence, it is not possible to confirm or reject both explanations. However, based on this research conservativeness might be an additional factor in explaining why companies have an internal data focus.

Moreover, as regards to the impact of big data in the future, the participants were unanimous. All companies indicated that they believe that there is a positive relationship between big data and performance. Actually, multiple participants believe that this positive relationship will even become more important, as they declared that: “The effect of big data will become bigger” and “Big data will become

more and more important”. In addition, participant F stressed that this positive relationship depends on

the attitude of a company, as he stated: ”Companies that consider big data as an important component

will absolutely use big data to improve the overall performance. Companies with limited focus in this area will not reap the full benefits of big data”.

BIG DATA IN STRATEGY MAKING

USE OF BIG DATA IN STRATEGY MAKING

Focusing on the use of big data in strategy making and the responsibilities associated with it, the participants are consentient again. Concerning the use of big data, all participants highlighted – as showed in table 8 – that big data is solely used to support decision-making. None of the participants indicated that big data is used to automate this process (yet). Despite the similarities among participants, when looking at the details of using big data in strategy making, some differences arise. For example, participant C uses big data with regard to innovations, as he stated: “we use big data to base all our innovations on”. Others are strongly focused on the project administration part, as participant F indicated that data is essential to “keep track of projects and, eventually, essential for owners to determine the direction for the company.”

percentages of 58% for decision support, while 29% of the time it is used for decision automation. Hence, there is a clear difference between both studies. Important to mention, is that this difference does not automatically mean that the results of both studies are inconsistent, however the scope of this research is too limited to formulate grounded statements.

TABLE 8.HOW IS BIG DATA USED IN STRATEGY MAKING AT YOUR ORGANIZATION? Company How used in

strategy making Quote

Case A Support

People have their beliefs and make decisions based on “their” information. Therefore, we try to show the information in a uniform format and dashboard, so that people can use the right information, which can be easily understood.

Case B Support

For the future, I believe that the automation of decision-making will become more relevant for our industry. However, it is a long way before we get there. So at this moment, big data is primarily used to support decision-making.

Case C Support

Big data is essential for us to formulate a proper strategy. Moreover, within our supply chain we strongly focus on using all data available to use it in such way that processes can be executed in a more efficient way.

Case D Support

Strategy making coincides with the analyses of information. Firstly, based on data we choose a direction and, secondly, data enables us to check whether the strategy is working. So we use big data to support our strategy making.

Case E Support

When thinking about strategy making, I think about “predictive analyses”; what are the predictions and how do I (automatically) adjust my policy based on that knowledge. Nowadays, this industry is in the stage of “descriptive analysis”.

Case F Support

Project managers are dependent on big data, as they need to keep track of the project administration. Secondly, it is essential for owners to collect all (project) information to determine the correct direction for the company.

Case G Support

Big data is used as the basis for the decision making process. This basis consists of all kinds of data sources, and these components together are the basis of strategy making

ATTITUDES AND FUTURE PLANS

Regarding the attitudes of participants related to using big data in strategy making, the managers of metal manufacturers indicate that they all have plans to increase the use of big data in strategy making. Exemplary of these plans are statements of participants B and A, as they respectively state: “Absolutely.

But that is also a must” and “We are planning to increase the use of big data, whereas changing the culture (and people) is one of the most crucial factors”. Surprisingly, the experts seem to be more

skeptical about the future plans of metal manufacturers. For example, expert E stated: “I think that the

major focus of companies in the metal manufacturing industry is first to make sure that all processes are correct”, while expert H indicated the importance of awareness: “Important is that companies must be aware of it before they can reap the full benefits of big data.”

When, subsequently, focusing on whether big data is considered as a problem or an opportunity, the outcomes are relatively equal. Most participants identified that big data is an opportunity. Experts F, G, and H highlight an addition on this view: the presence of awareness. For example, participant F stated:

“Some companies see big data as an important opportunity. However, the really conservative companies may consider big data as a problem”, and participant G indicated: “They have to be aware of the added value of it. When they see it, they will increasingly use big data in strategy making”. Additionally,

participant B, C, D, and G suggest that this “opportunity” that most participants indicated, may be considered as a “must”. Participant B indicated this “must” as he believes that companies in Europe have no other choice than using big data to formulate a proper strategy, to prevent a domination of nations like China. Participant C followed this line of reasoning as he stated: “When you ignore it you won’t be

successful, so in a sense it is a forced opportunity." In addition, participant D indicated: “This desire is strengthened by the environmental change: the world is getting ‘smaller’ and competition increases.”

BIG DATA IN STRATEGY MAKING - A FRAMEWORK OF SUCCESS FACTORS AND BARRIERS

SUCCESS FACTORS OF USING BIG DATA IN STRATEGY MAKING

In order to identify the factors of which participants think about when asked about the success factors, they were asked about what they consider to be the most important success factors for using big data in strategy making. In table 9, the participants’ answers are collected and, notably, showed a relative wide range of success factors. When analyzing this broad range, all mentioned success factors could, roughly, be distributed among three groups. In the first place multiple participants mention components being part of the group “people”, by stressing the importance of mentality, creativity, talent, and support of people. Secondly, participants appoint components such as vision, ambition, dreams, and direction as relevant success factors. This group can be considered as the “future/leadership” group. Finally, the group “data” can be formed, as participants stress the importance of design, quality (no garbage in, no garbage out), correctness, availability (low thresholds) and structure of data.

TABLE 9.WHAT ARE THE SUCCESS FACTORS TO USE BIG DATA IN STRATEGY MAKING?

Company Success factors to use big data in strategy making

Case A Mentality and creativity of people – no focus on individual interest, but the bigger picture and being able to connect the knowledge and information.

Case B Talent - the correct use and seek for talented and young people, in order to promote innovation and progress. Additionally, talent must be offered “space” to initiate innovations.

Case C Vision/Targets | Ambition/Passion | Dare to dream

Case D The most important success factor is that people support big data. That people believe and experience that there is an added value in big data. So the human aspects.

Case E

(1) Support among employees - Everybody needs to be convinced that they need the use of big data to come to the right strategy. (2) Design - When the big data is not presented the right way, people will not use it sufficiently. (3) Correctness of data - From day one, data needs to be correct.

Case F It is important to have a clear sense of direction for the future of the company. When this is lacking, companies may use big data the wrong way.

Case G The structure of data | Provide insights | Quality

Despite the broad range of success factors discussed above, the participants were relatively consentient about the framework of success factor developed in the literature review (table 2 and 10). That is to say, to a large extend do all participants agree with the pre-defined list of six success factors. However, some participants would carry out small changes. Participant B indicated that he considers “decision-making” as part of “leadership”, while participant D does not agree that the component “talent” is a major success factor. On top of that, participant F believes that “talent” and “leadership” often relates to the same person and considers the two factors as one. Finally, participant H argues that “decision-making” and “targets” are part of the “organizational” success factor (all success factors that are considered as part of other success factors are, in table 9, rated with the highest ranking: 7). Additionally, participants A, B, D, G, and H stressed the need of additional success factors, which are grouped into the factor: “communication / human aspects”. Participant A indicated this need by stating: “What I miss are the culture, mentality and

communication, so the “human aspects” of a company”. On top of that, participant B mentioned: ”For sure, communication is part of every success factor, but due to the importance I consider it as a success factor on itself.

As regards to the ranking of the success factors, again diverse findings arise (see table 10). Although the participants largely agree with the six predefined success factors, when analyzing the rankings a wide variety of results comes forward. Due to this variety, some success factors get the highest ranking from participants, while other participants consider this success factors to be the least important. This causes, in general, that the scores per success factor do not strongly differ. When analyzing the number of times that factors are considered as a top 3 success factor, it becomes clear that the factors: “Talent” and “Decision-making” are only by 12,5% of the participants ranked as a top 3 factor.

TABLE 10. THE MOST IMPORTANT SUCCESS FACTORS

Success factors Case A Case B Case C Case D Case E Case F Case G Case H Top 3

BARRIERS OF USING BIG DATA IN STRATEGY MAKING

After analyzing the success factors, the barriers of using big data for strategy making are discussed. To do so, the same route as the success factors is used. Therefore, in the first place an open question regarding the most important barriers is addressed. When studying the barriers mentioned (table 11), again the broad range of answers can could, roughly, be distributed among three “groups”. Firstly, the “people” group appears to be also relevant regarding the barriers. Mentioned components of this group are: mentality and culture, age, ignorance, knowledge, capacity, and communication. In the second place the group of “future/leadership” consists of barriers such as a lack of priorities, lack of leadership, lack of management awareness and conservative leadership. Finally, the barriers:capacity, costs, conflicting interests, inability to change, wrong data (no garbage in, no garbage out), limitations of database, limitations of tools and dashboards, and limitations of project managers indicate the relevance of the barrier-group “organization”. TABLE 11.WHAT ARE THE BARRIERS TO USE BIG DATA IN STRATEGY MAKING?

Company Barriers to use big data in strategy making Case A The mentality and culture: Human aspects

Case B Age (being able and dedicated to change) | Knowledge Case C Capacity | Priorities

Case D

(1) Not having the proper tooling / dashboards, and due to that big data is not available in the right place and way. (2) Mindset of people, people are thinking in ways of “now” instead of “tomorrow”. So again the human aspect is important

Case E

(1) Costs - especially smaller companies face this barrier, (2) Conflicting interests within the organization, (3) External project managers (without decision-making rights), (4) Communication, (6) Limitations of servers/database, (7) Lack of leadership

Case F

(1) Leadership – many companies still have “conservative leaders” that do not give big data the priority is deserves. (2) Organization – sometimes departments are solely focused on their own goals and forget the bigger picture.

Case G Wrong data | Inability to change | Finance

Case H No garbage in, no garbage out | Ignorance / lack of clarity | Management awareness | Capacity / Time

but part of the “lack of talent and skills” barrier. In addition, participant H does not consider “financial barriers” as relevant barriers, while the barrier “difficulties in executing analyzes” are, according to him, part of the talent and skills barrier (barriers that are considered as part of other barriers are, in table 12, rated with the highest ranking: 9).

In the case of the seven barriers, the participants did not propose major changes. In contrast to this, multiple participants indicated the need for additional barriers. These barriers can be defined into three factors: (1) lack of communication / human aspects, (2) lack of leadership and targets, and (3) lack of time, capacity and priority. The additional barrier “lack of communication / human aspects” is indicated by participant A, B, D, E, and H. Participant A indicated the need for the barrier by stating: “Again, what I

miss are the culture, mentality and communication, so the ‘human aspects’ of a company”. Four

participants (A, B, E and F) stressed the need for barrier “lack of leadership and targets”. Thirdly, participant C identifies the additional barrier “lack of time, capacity and priority”, as he considers it as a core barrier.

When subsequently asked to rank the “new” list of barriers, it appears that participants have different views (table 12). However, in contrast to the success factors, the barriers show more variety as regards to the frequency that barriers are indicated as a “top 3” barrier. For example, the barrier “lack of

communication / human aspects” is considered a top 3 factor, 62,5% of time, while the barrier

“complexity and scalability issues” are not considered as a top 3 factor even once. TABLE 12. THE MOST IMPORTANT BARRIERS

Barriers Case A Case B Case C Case D Case E Case F Case G Case H Top 3

Lack of talent and skills 3 4 5 8 4 3 7 6 25%

Financial barriers 5 7 7 5 7 7 3 (9) 12,5%

Difficulties in executing analyzes 6 6 6 3 5 5 4 (9) 12,5%

Organizational barriers 4 2 4 4 2 1 1 2 62,5%

Limitations of database and

systems 9 5 3 2 8 8 2 5 37,5%

Complexity and scalability issues 8 (9) 8 6 (9) 4 5 4 0%

Security issues 7 8 2 7 3 6 6 3 37,5%

Lack of communication / human

aspects 2 1 1 1 1 62,5%

Lack of leadership and targets 1 3 6 2 37,5%

Lack of time, capacity and

USING BIG DATA IN STRATEGY MAKING - A FRAMEWORK OF SUCCESS FACTORS AND BARRIERS

In order to give an interpretation of the rankings of success factors and barriers, the criteria as described in the methodology part are used. When analyzing the success factors based on the criteria, it appears that the success factors “communication / human aspects” met the minimum criteria of 50% and is added to the list. This success factor is also key success factors since it meets the requirements (>50% top 3). Regarding the other factors it appears that the success factors “leadership”, “targets”, and “organizational” are also considered as key success factors.

With respect to the barriers, the aforementioned criteria indicate that the barriers show a different overview compared to the success factors. Based on the criteria, the barriers: “lack of communication / human aspects” and “lack of leadership and targets” must both be included to the list of barriers as they meet the 50% requirement. On top of that, the barrier “lack of communication / human aspects” must, together with the “organizational barriers”, also be seen as a key barrier. On the contrary the other barrier groups must, based on the criteria, considered as regular barriers. This indicates that with respect to the barriers, two groups are more important than the other groups. However, being a regular success factor or barrier does not mean that the factor is negligible. Even stronger, in many cases participants indicated that all barriers and success factors are interrelated with each other and that without one problems would arise. When combining the “new” rankings of success factors and barriers, the classification as illustrated in table 13 can be formed:

TABLE 13.A FRAMEWORK OF SUCCESS FACTORS AND BARRIERS

Success factors Barriers

Key success factors • Leadership • Targets • Organizational

• Communication / human aspects

Key barriers

• Organizational barriers

• Lack of communication / human aspects

Regular success factors • Talent

• Technology • Decision-making

Regular barriers

• Lack of talent and skills

• Difficulties in executing analyzes • Security issues

• Financial barriers

• Limitations of database and systems • Complexity and scalability issues

CONCLUSION AND DISCUSSION

In this paper, the gap with respect to how big data is used for strategy making in the metal manufacturing industry is appointed. By doing so, the research question: “WHAT ARE THE SUCCESS FACTORS AND BARRIERS FOR USING BIG DATA IN STRATEGY MAKING, WITHIN THE METAL MANUFACTURING INDUSTRY?”can be answered. As illustrated in table 12, the success factors forusing big data in strategy making, within the metal manufacturing industry are: (1) Leadership, (2) Technology, (3) Targets, (4)

Organizational, (5) Communication / human aspects, (6) Talent, and (7) Decision-making. The barriers

for using big data in strategy making, within the metal manufacturing industry are: (1) Organizational

barriers, (2) Lack of communication / human aspects, (3) Lack of leadership and targets, (4) Lack of talent and skills, (5) Difficulties in executing analyzes, (6) Security issues, (7) Financial barriers, (8) Limitations of database and systems, and (9) Complexity and scalability issues.

Based on the abovementioned groups of success factors and barriers and other findings, multiple eye-openers can be appointed. In the first place, using big data in strategy making is not static but a dynamic process, affected by - among others - a diverse range of interrelated success factors and barriers. This dynamic and interrelated character contributes to the suggestion that, although some factors are indicated to be more important than other, none of the factors must be neglected but must be consciously threated in such way that a successful use of big data in strategy making is achieved. Secondly, the presence of the factor “communication / human aspects” in both groups of success factors and barriers is interesting since existing literature completely ignored this factor. In fact, this research even considers this factor as a key success factor and key barrier, indicating the importance of “communication / human aspects”. Therefore, this study stress that the influence of such “soft skills” must not be underestimated and that (future) literature should be aware of this importance.