Outsourcing Big Data Analytics to Data
Analytics Providers: from Insights to
Big Data-driven Innovation
Exploring a path from Big Data insights to business value
Roxanne Markus
UvA: 6296823/VU:2566833 MSc. Entrepreneurship
Final Master Thesis Entrepreneurship
University of Amsterdam & Vrije Universiteit Supervisor: Dr. Wietze van der Aa
S
TATEMENT OF
O
RIGINALITY
This document is written by Roxanne Markus who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Acknowledgements
This thesis is part of the MSc. Entrepreneurship joint program from the Universiteit van Amsterdam and the Vrije Universiteit. During this Master’s program I was honored to meet many inspiring people who made my life as student so much greater. One of which was of course my supervisor Dr. Wietze van der Aa, who helped me to be critical of my own work.
I would also like to thank the people from Philips for their constant support and willingness to help me improve my work. In particular, I would like to thank my supervisor Pieter Custers and team member Chen Chen, who provided me with the information, contacts and access to databases to conduct my study. Also my fellow interns without whom my time as an intern has become an unforgettable one.
Of course I would also like to thank also all the participants to this study who dedicated their time to provide me with their insights and views on my subject matter.
Contents
STATEMENT OF ORIGINALITY ... 2
ACKNOWLEDGEMENTS ... 3
ABSTRACT ... 5
1.
INTRODUCTION ... 6
1.1. Research Question ... 8 1.2. Structure ... 92.
THEORETICAL FRAMEWORK ... 10
2.1. Big Data Analytics & Big Data-driven Decision Making ... 10
2.2. The implementation of Big Data Analytics ... 14
2.3. Big data analytics use cases ... 20
2.4. Conclusion to this chapter ... 24
3.
METHODOLOGY ... 26
3.1. Research Methods ... 26
3.2. Research approach... 28
3.3. Reliability and Validity ... 32
3.4. Limitations ... 32
4.
ANALYSIS & DISCUSSION ... 33
4.1. Organizational culture challenges ... 33
4.2. Solving the organizational culture challenges... 46
5.
CONCLUSION ... 51
5.1. Theoretical implications ... 53
5.2. Practical implications ... 54
5.3. Limitations and suggestions for future research ... 55
REFERENCES ... 56
Abstract
Big data have the potential to provide for vast competitive advantage if organizations know how to derive value from them. An increasing amount of organizations is seeking for ways in which optimal benefit can be yielded from their big data and thus they explore the many ways in which they can implement big data analytics. Big data analytics implementation through partnerships is an attractive strategy for organizations as they do not have to invest in in-house resources and capabilities. Moreover, data illiterate organizations can even learn from their data analytics providers and in so doing elevate the insights that these partners can provide them. In addition, big data analytics is only beneficial for an organization if the right goals are set. One recently popular use case is big data-driven innovation, which uses big data insights for less risky and more secured innovation ideas. As innovation perceived to be essential for the survival of firms in the current market and big data analytics implementation through outsourcing can lead to additional insights, using big data partnerships to create big data-driven innovation seems to be an interesting implementation strategy for organizations. However, when firms pursue this path, they still need to overcome some challenges pertaining to their organizational culture, in order for this approach to be successful. After all, big data analytics practices thrive in a culture in which insights are cultivated and decisions are made based on them. The current article therefore explores the organizational cultural challenges that organizations face when implementing big data analytics for the purpose of big data-driven innovation. Through a systematic content analysis of 5 interviews with data managers and 61 big data related articles this study has found that especially the challenges related to adaptation, external and internal communication, top management support and data-driven decision making need to be solved. In addition this study has found that these issues can be resolved by investing in open innovation and education. The current work has therefore contributed to the existing literature about big data analytics implementation and big data-driven innovation and has provided
practitioners with the knowledge that is necessary to make decisions about big data analytics implementation.
1. Introduction
Today’s economy is currently facing a data revolution: detailed data about processes, operations, customers, partners and competitors have become more available and are even growing at an exponential rate. The collection of these data, which are characterized by their large volume, variety of forms and great velocity are often termed 'Big Data' (De Mauro et al., 2015). Because of the emergence of Big Data, decisions do no longer have to be based on a “gut-feeling”, rather, they can be driven by data-based insights. Brynjolfsson, Hitt & Kim (2011)have found that performance of firms that emphasize data-based decision making and business analytics is 5-6% higher than would have been expected given their other investment and information technology usage. That is, data-driven decision making has the potential to drive firm performance. Unfortunately, the problem for many organizations is that their business operations are not suitable for the management, let alone the analysis of big data (Lavalle et al., 2011). If firms succeed to effectively manage and extract valuable insights from their big data, they can create substantial competitive advantage in relation to those firms that do not have these capabilities (Davenport, 2006). Tan et al. (2015) even state that the key factor to gaining competitive advantage in today’s rapidly changing business environment is the ability to extract big data to gain helpful business insights. It is therefore increasingly important for an organization’s success that it has the right data analytics capabilities in order to compete in the current market (Cavanillas, Curry & Wahlster, 2016).
For this reason, many organizations are seeking for the best ways in which they can invest in the implementation of big data analytics capabilities for the generation of increased business value (Lavalle et al., 2011). Big data analytics capabilities can be implemented in several ways: they can either be developed in-house, or they can be outsourced to a data analytics provider. Companies could expand their workforce by hiring data analysts, data architects and by forming a center of excellence devoted to data analytics alone. Data analytics is a time-consuming activity of which the results cannot easily be predicted. Mining through terabytes of data in order to find those insights that can lead to increased business value can at times be perceived as searching for a needle in a
haystack of which even the location of the barn, farm, or country is unknown. Another logical strategy is to create a partnership with data analytics firms who have the right technology, skills and experience to make sense out of data. The partnership is a time- and often a cost-efficient means for a company to adopt data analytics capabilities, while still being able to focus on core business activities (Lee, 2001; Chen, Chiang & Storey, 2012). In addition, firms that choose partnering with a data analytics company as a means to implement their own big data analytics activities could enhance their big data analytics capabilities by drawing on and learning from the expertise of data analytics firms. Because of these potential synergy advantages combined with little required investment in resources and capabilities, forming a data analytics partnership seems to be an attractive strategy to implement big data analytics implementation. However, what these
partnerships do not ensure is that the results of big data analytics will be used for the right big data use cases and that the actual usage of big data driven-insights becomes part of the organizations culture.
When it comes to the potential big data analytics use cases, the existing body of
management literature of big data analytics can be divided into two main categories: those articles that stress the results and implications of business analytics and those that are aimed towards marketing analytics. The former category of articles about business analytics, is concerned with how big data analytics can be used for the optimization of business (Chen, Chiang & Storey, 2012), while the latter is targeted towards understanding how customer data can reveal future customer behavior (Sagiroglu & Sinanc, 2013). According to Davenport (2006), the reason why these two use cases receive so much attention is because their results can easily be measured by both scholars an practitioners, thus validating the importance of big data analytics in firms. However, big data’s potential reaches further than increasing marketing effectiveness or improving operational
efficiency: the insights gained from business analytics and market analytics practices provide a firm with potentially valuable business and market knowledge, which can be exploited for the creation of innovative products and services. Studies have shown that innovation is an important enabler for
firm performance (Koellinger, 2008; Kemp et al., 2003). The use of (big) data analytics for innovation, then, could even enhance the success rate of a firm’s product innovation activities, because of its foundations on fact-based decisions. (Big) data-driven innovation is therefore a research area that is now gaining increasing attention by scholars and data practitioners, but the concept is still in its infancy and rigorous studies need to be conducted in order to validate its purpose (Jetzek, Avital & Bjorn-Andersen, 2014). Knowing what could potentially cause data-driven innovation of products and services could have tremendous impact on a firm’s success.
Combined with the strategy to outsource big data analytics capabilities, using big data insights for the purpose of innovation may form a beneficial implementation strategy. However, as mentioned before, when resources, capabilities and purpose have been ensured, it is then the task for the organization as a whole to adopt this new system and use it for the right purposes. Especially when a company has outsourced their big data analytics activities, the cultural adoption of big data-driven decision-making could pose as an important challenge, since big data analytics is then not rooted in the organizations culture. Therefore, this study aims to provide a deeper understanding of these organizational cultural challenges that firms need to overcome, in the implementation phase of big data analytics capability acquisition through a partnership for the purpose of big data-driven innovation.
1.1. Research Question
The question that of this study is: what organizational culture challenges would firms need to
overcome in the implementation phase of big data analytics when partnering with data analytics providers so that insights can be used for data-driven innovation ?
Because innovation can be perceived as a broad concept, attention will only be given to the
production of innovative ideas that can eventually result in innovative products and services. That is, this thesis will be about those ideas that are novel and lead to the specific solution of a market need. By answering this question, not only will the organizational culture challenges be identified, but also
the success factors that lead to the solutions to these challenges will be suggested in the forthcoming study. The aim of this study is to provide practitioners with the knowledge necessary to make big data analytics implementation decisions . The results of this study will also help them to identify certain cultural challenges before they occur and will enable them to create strategies for a smooth integration of big data analytics in the organizational culture so that the results of its practices may be used for the innovation of products and services. By doing so, this study also contributes to the current body of literature about big data analytics implementation and expands the knowledge that currently exists about drivers of big data-driven innovation.
1.2. Structure
In order to contextualize this study, the following chapter will give the definitions and descriptions of the theory on which this study builds. Chapter 2, will therefore discuss the concepts of big data analytics, the implementation of which and big data driven use cases. Furthermore, qualitative methods have been chosen to explore the research question. Content analysis of secondary data and semi-structured interviews lead to the results. Why these methods have been selected and how they have been utilized will be explained in chapter 3. Chapter 4 presents the results of the content analysis and therefore the various challenges and their implied solutions. The current paper ends with a conclusion of these results and suggestions for future research in chapter 5.
2. Theoretical Framework
Big data analytics is a fairly new concept which has recently been introduced in literature. It’s
underlying foundation give a deeper understanding of how big data analytics may provide benefits to current businesses and implies what organizations need to change in order to yield the best results of its practices. Therefore, this chapter will provide an introduction to previous works about big data analytics and big data-driven decision-making. Moreover, big data analytics can be implemented in several ways, this chapter will discuss recent papers about the various strategies organizations can adopt to implement data analytics capabilities. Lastly, this chapter will outline main big data use cases and will dive deeper into big data-driven innovation in particular. In the concluding section, a conceptual framework will be presented, that connects the literature to be discussed and that will provide for the basis upon which this study will be conducted.
2.1. Big Data Analytics & Big Data-driven Decision Making
The current era is often named the ‘the Digital era’ because of the fact that many analog operations have been digitized in the past two decades (McAfee & Brynjolfsson, 2012). For organizations, this means that where customer insights were first discovered through interviews with customers or by publishing surveys, they are now captured from digital technologies like the Internet. These recent technological developments have enabled firms to generate and store data such as online customer data, social media data, point-of-sales-data, personnel data, sensor-generated data, log data together with data from machines, enterprise resource systems and et cetera. The combination of these data sets result in one that is so complex that it cannot be processed by traditional data management systems. Laney (2001) suggested that the complexity of the resulting big data lies in its three main dimensions: volume, variety and velocity. First, he speaks of the large Volume of data, which is a result of data sets being stored almost continuously, and from a large variety of sources. Another characteristic of big data is its Variety, which applies to the variety of the data and
big data is characterized by Velocity, which is related to the increasing rate with which big data is produced. Ever since his adoption of the Three V’s, this concept has emerged as a common framework to describe big data (Gandomi & Haider, 2015; Kwon et al., 2014; Chen et al., 2012). Other authors have drawn upon the works of Laney by even adding other V’s, such as Veracity, which is the aspect of big data related to the meaningful usability or quality of big data and its Value (Gandomi & Haider, 2015). According to Prescott (2016), big data also refers to “the large amount of data that is continuously being collected, stored, and managed, along with the evolving IT
technologies that make this possible and analytic techniques used for gaining and understanding of the data.” Due to its high operational and strategic potential, big data has the potential to generate business value and allows enhanced visibility of firm operations and improved performance
measurement systems (Wamba et al., 2015). Realizing its potential, firms grasp the necessity of big data in the current Digital era as a tool for competitive advantage What is left for firms is now to find and execute the strategies with which maximum big data profit can be reached.
These strategies are necessary to develop, because big data in isolation, is worthless. That is, without processing, big data alone cannot lead to competitive advantage. According to Gandomi & Haider (2015) the “potential value of [big data] is only unlocked when they are leveraged to drive decision-making”. Big data are data that are so large and complex that they require unique
techniques and methods for their storage, management and analysis. Big data analytics refers to the application of advanced analytics techniques to big data sets for the acquisition of business and market intelligence (Gandomi & Haider, 2015). Big data analytics technologies and techniques have enabled firms to analyze large scale and complex data for various applications intended to augment firm performance in various dimensions (Kwon et al., 2014). Examples of those technologies and techniques are database and data mining tools, predictive modeling and web analytics that allow organizations to use big data for the prediction of future trends and track, for example, online customer behavior. Through these big data analytics tools, a firm can, for instance, improve the monitoring of the acceptance of its products and services in the marketplace and the understanding
its business environment (Davenport, 2012). These insights can then be used to make well-informed, data-driven decisions that can to lead to more personalized products, which could then lead to an increased amount of product sales. This way, Wamba et al. (2015) argue, retailers have achieved up to 15 to 20% increase in ROI by putting big data into analytics. More about big data use cases will be discussed in subchapter 2.3.
2.1.1. Big data processes
Although the term ‘big data’ flourished in academic research and in business since the beginning of the 2000s, the practice of deriving value out of data to drive evidence-based decisions is not a new one. As early as the 1970s, business journals began publishing articles related to decisions support systems, which were defined as follows:
“[A decision-support system] is an interactive computer-based system or subsystem intended to help decision makers use communications technologies, data, documents, knowledge and/or models to identify and solve problems, complete decision process tasks, and make decisions” (Gory and Scott Morton, 1989).
Making well-informed decisions to reduce risk and maximize potential value was at the core of early decisions support systems. With these systems, companies focused on making their most complex decisions with help from powerful information systems that arose during that time (Hedgebeth, 2007). Firms that adopt big data analytics continue the practices of these decisions systems by maximizing value (for example, by decreasing operational costs), yet at a larger scale. According to practitioners, big data has the ability to improve a firm’s decision-making process by advanced visualization of a firm’s operations and improved performance measurement mechanisms (McAfee & Brynjolfsson, 2012). The only difference, then, between early decision support systems and big data analytics by these firms is that the volume, variety and velocity of the data has been increased since the emergence of the Digital Era.
According to Gandomi & Haider (2015), big data analytics is part of a five-stage process of extracting insights from big data. As shown in figure 1, the first three stages are involved with the
management of data, which is its acquisition, processing and visualization. The final two stages of the big data process entail big data analytics, which can be broken down into the modeling, analysis and interpretation of the data.
Figure 1- Big data Processes (Gandomi & Haider, 2015)
Perceiving big data analytics as a sub-process in the overall process of insight extraction from big data is an essential realization for a firm to make. The performance of a firm in every stage of the process affects the eventual insights that are to be made. For example, if a firm invests in the right people, tools and methods to interpret the data, but fails to extract and clean the right type of data, insights may be drawn upon faulty or incomplete data. This would result in wrong decisions being made. It could also be that the input quality of data may be so low that the analysis of the big data may not lead to any important insights. Thus, thorough understanding of data management and analytics needs to be developed in order for big data analytics processes to lead to significant business value. Hence, it is important for firms to understand which challenges they will face in the big data process.
2.1.2. Trend in big data analytics literature
Similar to early works about big data analytics, extant literature about decisions support systems is aimed towards exploring the techniques, tools and methods that could be used to acquire insights (Rao et al., 2006). However, a recent trend in academic literature about big data analytics is that research focus has shifted from the technical side to a strategic side. Fewer academics aim their studies on understanding how analytics can best be performed but more on what analytics can be
used for. Not only in literature but also the practitioners of big data analytics have witnessed this shift. In the early 2000s, CPU and IT departments faced dramatic challenges and technical issues concerning the exponentially increasing quantity of potentially useful data. First and foremost, there was trouble finding the most efficient ways to store and manage big data (Kohavi et al., 2002; Russom, 2011). Nevertheless, now that widespread big data analytics systems are in place and technical problems have been resolved, firms tend to use big data, not only for a way to decrease operational cost and risk but also as a means to differentiate themselves and gain significant competitive advantage (Davenport, 2006).
This shift in focus signifies that scholars and practitioners are now mostly concerned with the question of how to successfully implement big data analytics so that big data will generate
substantial business value (Bean, 2016). The issues related to this question require an analysis of the various big data analytics adoption strategies that a firm can perform together with an exploration of the ways in which firms can profit from their data-driven insights. The following two sections will provide an overview of these two issues.
2.2. The implementation of Big Data Analytics
In order for a firm to benefit from the big data it has accumulated through its sensors, web presence, customer or store operations, it must adopt the right capabilities with which big data analytics can be performed. According to Gupta and George (2016), firms must therefore develop or acquire what they term big data analytics capabilities, which are a firm’s ability to assemble, integrate, and deploy its big-data resources. According to Kwon et al. (2014), because big data analytics involves database searching, mining and analysis, it can be seen as an innovative IT capability that can improve firm performance. By successfully adopting big data analytics capabilities, a firm will be able to use the big data it has generated for the purpose of increasing business value. Therefore, big data analytics can be perceived as an essential source of competitive advantage (Brown et al., 2011). Viewing big data analytics as an IT capability implies that the implementation of such a capability must be approached
like other innovative IT systems and successful big data analytics implementation depends on resource and capability factors.
2.2.1. Successfully implementing big data analytics
Traditional tools and techniques for the analysis of ‘little data’ are often not advanced enough for the management, storage or analysis of big data (Chen et al., 2012). This is because these traditional data, such as the logs of an organization’s personnel can often be mapped in simple data
spreadsheet programs. In contrast, big data sets such as the clickstream activities of customers on an organization’s website, would be too large and complex to be processed in a simple data
spreadsheet program and need to be processed by a skilled team of data analytics professionals. Because big data is so different from the traditional data sets that organizations have formerly acquired, firms often struggle with finding the best ways to acquire the resources and capabilities that are necessary for its successful implementation. Success, in this respect, means that the big data analytics system is correctly set in place to achieve the desired big data results. These results are, of course, specific to an organization’s vision and goals, but can be related to three big data analytics use cases, which will be described in the next subchapter.
For the successful implementation of big data analytics into any business, firms need to address certain challenges that the difference between traditional data and big data brings about. Before discussing these challenges, it is important to emphasize that adoption of big data analytics capabilities is a relatively new phenomenon. Consequently, not much literature is focused on this topic and insufficient academic literature exists to draw upon. Fortunately, there are other, older innovative IT systems, that have gone through similar roads as the big data analytics system. Understanding how a system with similar features as big data analytics was adopted by firms in the past and learning from the challenges it faced, will provide managers who desire to adopt big data analytics with an important foundation of knowledge. An IT technology that fits this description is the Enterprise Resource Planning (ERP) system. This is an IT solution that provides a centralized IT
data analytics, the solution has been designed around the collection and storage of information and is often argued to have been critical to increasing business value of its early adopters (McAdam & Galloway, 2015). Also similar to big data analytics, ERP systems hold the promise of improving processes and decreasing costs (Fui-Hoon et al., 2001). Moreover, ERP systems have been used in literature to discuss the implementation of other IT-related systems which validates the usefulness of exploring this phenomena (King & Burgess, 2008). Thus, exploring what has been said in academic literature about the adoption of ERP capabilities, will provide a good overview of the factors that could lead to success of the adoption of big data analytics capabilities.
Through an investigation of the big data analytics capabilities and ERP systems adoption literature to date, four main critical success factors for the implementation of these systems were identified: resources and capabilities (Erevelles et al. 2016; Kwon, Lee & Shin, 2004; Robey et al., 2002), top management support (LaValle et al., 2011; King & Burgess, 2008; Davenport, 2006), adoption by organizational culture (Erevelles et al., 2016; Kwon, Lee & Shin, 2004; Robey et al., 2002) and a clear objective (Dutta & Bose, 2015; Robey et al., 2002).
Resources and capabilities
As mentioned before, acquiring insights from big data can be perceived as a process consisting of both the management and analysis of big data. Big data management requires physical, human and organizational capital resources that can successfully process, collect, sort and extract insights from big data (Erevelles et al., 2016). This means that firms that wish to adopt big data analytics systems need to have access to the right tools and methods to manage and analyze big data. Moreover, big data analytics requires the expertise and skills of qualified data analytics professionals who can not only use the right tools and methods to acquire, process and visualize the data, but also know how to translate big data into meaningful, business related insights (Chen et al., 2013). Kwom, Lee & Shin (2014) argue that inadequate staffing and skills belong to the main reasons why many big data adoption initiatives are unsuccessful. Firms that believe that employees who manage traditional business or customer data know how to manage big data, will eventually realize that big data
analytics requires other skill sets and expertise. Investing in new, capable and relevant data analytics resources and capabilities is therefore essential for successful big data analytics implementation.
Top management support
Another reason why big data practices sometimes do not pay off is because of lack of business support (Russom, 2011). The incorporation of big data analytics in a company is a very drastic one and support from top management is essential in both enabling employees to exploit the possibilities that big data provide as well as to convince them of the importance of the value of big data analytics. Similar results can be found in ERP literature; Fui-Hoon (2001) argues that management should have a strong commitment to use the system for achieving business aims (which will be discussed
hereafter). The reason for this is because top management support is critical for the adoption of a new system by the whole of an organization. If top management stimulates the use of such a new system, it is more likely to be adopted by employees at other parts of the organization. Top
management needs to clearly communicate its purpose and then translate it into action. This should not only be done in the analytics department, but throughout the whole organization were big-data driven insights will be used. This means that although the need for big data analytics, may come from various parts of the organization (e.g. the marketing department that wants to know what users are saying about a brand or support staff that wants to know how the business is operating over time), the decision to implement big data analytics into the daily activities from a company should be supported by top management.
Adoption by organizational culture
Culture can be defined as ‘a pattern of practices, behaviors and norms organized around a set of shared aims and beliefs’(Kiron & Shockley, 2011). As one can expect, making changes in the way that employees have acted and thought over a long period of time is something that is not easy to achieve. Still, knowing what challenges a firm needs to overcome in order to make its culture more adaptive to big data analytics is essential for the success for the implementation of big data analytics
capabilities. Although the factor of organizational culture adoption of a big data analytics results is an essential part of the implementation of this IT system, this subject has not been thoroughly discussed in literature. It will be useful for practitioners to understand what factors of organizational culture in specific will need to change in order for big data analytics to lead to the desired results.
According to Erevelles et al. (2016), when a firm does not succeed to design its organization around big data and to educate its members concerning proactive use of insights to improve a firm’s capabilities, big data analytics implementation may lead to disappointing results. They therefore also argue that firms should establish an ignorance-based view, which means that firms rely not only on knowledge – or what they do know -, but also on their ignorance – or what they don’t know (i.e., ignorance). An ignorance-based view implies that firms should be moved towards understanding the things they do not know, which facilitates latitude and freedom. This, in turn, is critical for
stimulating creativity within an organization, whereas demanding perfect knowledge may hinder creative activities. Creativity and latitude are factors that could help shape use cases other than decreasing costs and improving operational efficiency, which is something that will be discussed in the next subchapter.
Erevelles et al.’s (2016) study implies that firms need to make sure that the organizational culture is centered around the use of big data insights to make decisions. Data-driven decision making should not only be a task of top management or data scientist, but for everyone in the organization in order for it to lead to the desired results. In their study of 4500 business manager’s use of data analytics, Kiron and Shockley (2011) came to a similar conclusions, the most advanced users of big data typically have a strong data-oriented culture. Top managers are therefore encouraged to assess organizational culture to determine whether their company is open to adopt big data insights into their daily practices and decision-making (Erevelles et al., 2016).
A clear objective
Several studies have indicated that the desired results of big data analytics as well as similar systems such as ERP need to be well articulated before the implementation of this system has taken place (Including: Dutta & Bose, 2015; Robery et al., 2002). According to Dutta & Bose (2015), big data projects should be thoroughly structured and planned in order to ensure smooth execution, successful deployment and adoption by the firm. There should be a detailed and encompassing roadmap that outlines the data modeling and analysis, but also the implementation of the actual results of the big data insights. In other words, in order for big data analytics capabilities adoption to run efficiently, firms should first need to determine the type of insights they wish to acquire and then also provide for a roadmap on how to actively use those insights when they are acquired. The data that is extracted as a result of big data analytics practices should thus have a clear purpose. This will be discussed in the next subchapter
2.2.2. Outsourcing big data analytics
As explained before, acquiring big data analytics resources and capabilities can be a costly, timely, and effortful activity for a firm to engage in. This is the reason why some firms choose to outsource the resources and capabilities to a data analytics firm which is specialized in data analytics (Fogarty & Bell, 2014). These third parties are specialized and experienced in (big) data analytics, storage and management (LaValle et al., 2011) and therefore provide an efficient solution to firms that wish to jump on the big data analytics bandwagon, but do not have the capabilities in-home. Studies about the adoption or implementation of big data analytics capabilities are often related to the technical or strategic factors associated to big data analytics (Gupta & George, 2016; Norris & Baer, 2013), yet little is known about the factors pertaining to the organizational culture challenges of outsourcing big data analytics operations. In their article, Fogarty and Bell (2014) explained that outsourcing big data analytics allows companies to focus on using the data for the improvement of their core business, and can build on the insights that their data analytics partner delivers. They do not have to worry about having the right people and skills. Moreover, the fact that a company invests in big data
analytics implies that top management has committed to the activity and has supported a large investment on its behalf.
It is the task of a data analytics company to provide their customers with rich and complete insights that hides all complexity associated with data, its location and structure, while also providing actionable intelligence. These ventures are therefore often characterized by their ability to help customers make the best out of the new insights they have gathered. According to Fogarty and Bell (2014), smaller big data analytics companies also have the added value that they are more motivated than larger big data analytics companies to help their clients find new insights from data.
In addition, outsourcing big data analytics has the benefit that top management support is assured, because after all, top management plays a critical role in enabling the partnership with an external big data analytics provider. Outsourcing also assures that human and technological
resources and skills necessary for successful big data analytics can be acquired through a professional big data analytics provider of the firm´s choosing. Through this partnership a firm can decide to do business with an analytics provider who has experience in the industry in which that firm competes and knows how to deliver relevant and big data-driven insights. However, a partnership with a big data analytics firm does not ensure that a data illiterate company will have the right organizational culture and will set the right objectives for the implementation to be fully successful. In the following section, several probable big data analytics objectives will be discussed.
2.3. Big data analytics use cases
As mentioned before, this study is focused on the big data analytics efforts of organizations of any industry that have used or are considering outsourcing big data analytics as a means of adopting the big data analytics capabilities. The previous two subchapters discussed how former literature stresses the necessity of big data analytics for today’s firms and the main challenges that firms face when they want to implement big data analytics. Big data analytics implementation is one issue that
can be solved through outsourcing big data analytics capabilities. However, firms must still overcome certain challenges when it comes to organizational culture adoption of big data analytics insights. Another concern for firms that have outsourced their big data analytics capabilities relates to the issue of big data insights usage, the understanding of which is essential for the success of big data analytics activities. According to LaValle et al. (2011),the leading obstacle pertaining to widespread big data analytics adoption is the lack of understanding of how to use analytics to improve the business. Big data analytics partners can manage and analyze the big data, but it is up to their partners to determine what kind of insights they want to gather and, more importantly, what they want to do with those insights. In order for big data analytics to bear fruit, a firm must first
determine what it wishes to accomplish through its use. Therefore, the current sub chapter outlines several use cases for big data analytics which, in essence, are the ways in which a company uses big data analytics insights for the improvement of its business value. Moreover this subchapter sheds light onto one use case in specific, big data-driven innovation, for its relevancy to today’s dynamic markets and its particular use in partnerships with big data analytics providers.
2.3.1. Business and Market insights use cases
Firms often adopt big data analytics with the aim to obtain insights about issues that were unknown or uncertain before or to obtain traditional insights in a new and improved fashion. In the current body of literature about big data analytics, use cases fall into one of two main categories. The first is
business insights, which refers to the results of big data analytics that help businesses optimize
operations and make data-driven strategic decisions (Chen et al, 2012). According to LaValle et al. (2012), organizations that use business insights for decision making perform better than those business that do not. Banerjee et al. (2013) state that business analytics is often about either describing what has happened, diagnosing why it has happened, predicting what is likely to happen or prescribing what a firm should do about it. With business insights, organizations are able to increase their efficiency with regards to internal processes and operations, personnel planning, production, inventory, resources etc. Business insights-related use cases help firms with decision
making in these areas and therefore often result in decrease of operational costs.
Another category of big data use cases in current academic literature is market intelligence, this category includes the insights big data analytics provide about the current and latent needs of customers. Traditionally, market intelligence relied on market surveys, customer interviews and focus groups for a better understanding of consumer behavior and to improve product design (Hedin et al, 2011). However, Hedin et al. (2011) argue ‘with big data analytics key factors for strategic marketing decisions such as customer opinions toward a product, service or company, can be automatically monitored by mining social media data’. Also with the help of big data analytics, market intelligence can now be acquired in different ways and from a larger variety of sources (e.g. social media, web pages, sensors, RFID etc.). Different methods can be applied to discover market intelligence which can lead to more valuable insights about a market and thus provides a better source of data-driven decision making. This way, companies are able to use market intelligence to learn more about their customers and serve them more successfully. Big data analytics use cases that lead to market intelligence therefore lead to more revenue in terms of more products being sold, or could lead to an decrease in costs because of better efficiency of marketing activities.
Although business insights and market intelligence are the two most obvious results of big data analytics today, LaValle et al. (2011) argue that a third, less conventional use case could also lend itself to the increase of business value for big data analytics adopters. This business case of big data/driven innovation will be discussed in the following sub chapter.
2.3.2. Big data-driven innovation
According to LaValle et al., (2011), the velocity and variety of big data lend themselves to the discovery of ways to serve new markets with innovative product and services. That is to say that big data does not only have to result in innovation meant for optimization of business processes and customer insights, but also has the potential to specifically drive product and service innovation. Big data-driven innovation refers to the exploitation of any type of data in the innovation process for the purpose of value creation (Jetzek, Avital & Bjorn-Andersen, 2014). In literature,
several authors have dedicated their works to the notion that innovation is a process that can best be driven by big data. Some authors describe big data as “the next big thing in innovation” (Gobble, 2013), while others call it the “next frontier for innovation, competition and productivity” (Brown et al., 2011). Again some authors agree that big data is a means to unlock organization business value by unleashing new organizational capabilities and value (Davenport et al., 2012). Traditional means of innovation are often driven by what some authors suggest to be the incorrect approach and incorrect measurements of customers and their behaviors. Ulwick (2002) argues that, when conducting market research, companies often expect customers to offer potential solutions. According to critics this is a wrong approach, since customers often do not know what the ultimate product or service to meet their needs should look like or even what it should entail. Also, customers may have a limited frame of reference when suggesting innovative and new ideas. The same author also suggests that these types of customer co-creation may only result in incremental innovations. Therefore, companies rather elicit from customer the desired characteristics of a product and use this information to innovate (Bordoloi & Guerrero, 2008). This way big data insights can be used for the production of innovative ideas that can lead to innovative products and services (Kusiak, 2009).
2.3.2.1. Big Data-driven Innovation as a use case
Big data enable the discovery of megatrends and have the potential to reveal future business opportunities. Moreover, through the use of market intelligence, firms are able to produce products and create services that satisfy those unmet needs for which companies are forced to innovate their current businesses. According to Parmar et al. (2014) data analytics can power innovation because of three drivers: the explosion of big data, better tools for data and cloud-enabled business services. The massive amounts of data that are available and our ever enhancing capacity to integrate, analyze and exploit structured and unstructured data and our ability to make sense out of them has been transformed. In the past, the knowledge that we used to make decisions was only available in physical space, but now that business has become virtual it is even easier to share insights and to
drive innovative new thoughts. Weber (2016) argues that information is at the core of most modern disruptions and that it opens up new opportunities to attack industries from different angles. Therefore, big data as a foundation upon which organizations make data-driven decisions for the creation of innovative ideas seems to be a powerful strategy for firms to engage in.
2.4. Conclusion to this chapter
Big data analytics is becoming an increasingly important means for organizations to obtain
competitive advantage. Although many firms realize that there is much potential in the big data they acquire, they find that they do not have the right resources and capabilities to turn big data analytics into valuable insights. The implementation of big data analytics capabilities is therefore a challenge that many of today’s business and marketing insight mangers are engaged with. However, because the topic is a fairly new one, not much literature has been devoted to the challenges that need to be overcome before firms can successfully adopt big data analytics into their core businesses.
Fortunately, the abundant literature about the implementation of earlier IT systems similar to big data analytics, teach us that organizations must pay attention to top management support, acquiring the right resources and capabilities, making sure the organizational culture adopts big data and having the right vision and goals that lead to big data analytics success. These goals and visions were formerly mostly concerned with the acquisition of business and market intelligence in order to improve current practices. However, another, more promising way for big data analytics to lead to business value is by using it as a source of knowledge for the creation of innovative products and services. Big data-driven innovation is a relatively new concept, but is especially promising in the context of a partnership where different sets of data can be combined.
The current study proposes that when organizations choose to partner with a firm that is specialized in big data analytics this would imply that the firm has support from top management and has acquired the right resources and capabilities. If a firm has then established the vision and goal that big data analytics practices should mostly lead to innovative products and services, what is
then left for an organization, is to overcome the cultural challenges that exist to adopt big data analytics for its big data analytics journey. The conceptual framework below, depicts the premises of this study. The purpose of this study is to explore what organizational culture challenges must be overcome in order for an organization to use big data analytics, acquired through a partnership, for innovation.
Figure 2 – Conceptual Framework
Thus, with the following study it will become clear that the partnership with a data analytics company followed by solving organizational culture challenges that lead to big data-driven
innovation result in the successful implementation of big data analytics practices. Obviously, more factors could lead to the success of big data driven innovation and more factors could lead to successful big data analytics implementation, but one of the factors that is ill-represented in the current literature is the organizational culture of a company as well as big data-driven innovation. For this reason, this study emphasizes on exploring the challenges relating to an organization’s culture.
Succesful big data analytics implementation Big data-driven innovation Solving organizational culture challenges Partnership with data analytics company
3. Methodology
The current study is approached from a qualitative perspective. Qualitative research is especially useful for research topics that are understudied and where a general body of knowledge is yet to be formed (Patton, 1990). Recent literature about big data analytics and big data-driven innovation adopted this research approach with the same intention (Jetzek et al., 2014). Similar to the article of Fogarty & Bell (2014), this study mainly draws on the investigation of business cases as the primary means of data collection. As mentioned before, the aim of this study is to identify the challenges and solutions to the challenges pertaining to the organizational culture change of companies that aspire to implement big data analytics through a partnership for the creation of innovative products and services. In order to do so, it is reasonable to analyze how organizations have implemented big data analytics in the past and to discover what organizational culture challenges they overcome. A combination of secondary data and semi-structured interviews was chosen as a source for this purpose. These sets of data were then analyzed using content analysis method, which is often used for coding and classifying qualitative data with the aim to discover new and relevant insights (Neuendorf, 2002; Kohlbacher, 2006). This explorative method suits the current study, because of the lack of prior knowledge about the research topics at hand. The coming section provides a finer description of the research methods and the approach used to perform this study.
3.1. Research Methods
Semi-structured interviews
This study uses semi-structured interviews for the collection of data. Semi-structured interviews are often used in studies for their ability to answer question that are aimed towards a deeper
understanding of a certain topic (King, 1994). In semi-structured interviews the researcher has some key questions that need to be covered, although their use may vary from interview to interview (Bryman & Bel, 2015). As opposed to structured interviews, where the researcher uses
semi-structured interviews allow for an in-depth discussion about a certain topic to arise. Semi-semi-structured interviews are a suitable approach, especially for this study, where the focus lies on the discovery and exploration of the various organizational culture challenges that may exist. Another reason why semi-structured interviews suit this study is because they could lead to a discussion about the possible solutions to these challenges to arise. Because these challenges and the solutions to these challenges may vary from one company to another, it would serve this study well to look at big data implementation from various perspectives. Multiple viewpoints enable for more challenges to be discovered, which would lead to a richer set of data to draw conclusions from. For these reasons, multiple interviews where held with respondents from a variety of companies. Semi-structured interviews are not tied to a standardized set of questions, which leads to an in-depth conversation about the various research topics. For this study, the result was that a broad understanding of cultural challenges was developed and the semi-structured interviews also made it possible to explore the many ways in which the respondents resolved these challenges.
Even though semi-structured interviews are a good way to discover rich, diverse and in-depth data, they are often critiqued for their subjective nature. The data that results from interviews are dependent on the memory, eloquence and phrasing of the respondent. This is why another, more subjective method for data collection called secondary data collection, was used for this study in addition to the semi-structured interviews.
Secondary data
Although the topic of big data analytics implementation is still in its infancy in contemporary
literature, there is a body of case studies that have been performed around big data analytics, it’s use and it’s adoption by organizations. These case studies are often embedded in articles from journals, newspapers, blogs and online magazines (also known as secondary data sources) and are based on interviews with directors, senior managers and data analyst professionals and on observations within the data analytics industry. These articles are particularly useful for this study because they contain valuable information about historical issues related to the implementation of big data analytics. For
this reason, the current study’s approach is to draw insights from the multiple cases that are found in these articles so as to make use of the already existing body of literature. This method is often termed secondary data analysis, which is a prominent research method for the analysis of historical phenomenon. Although using documentary secondary data as a source of data collection has increasing popularity in academic research, there are a number of issues related to using such data (Saunders, Lewis & Thornhill, 2012). This includes the issue of the quality and evaluation of its usefulness in relation the research questions and objectives. Hence, careful attention has been given to the selection and evaluation of the documents for this study as can be seen in the next subchapter as will be clarified in the next subchapter.
3.2. Research approach
As mentioned earlier, this study used a mixed approach of two complementary research methods. The semi-structured interviews lay a broad foundation of knowledge to be drawn on for the
completion of the secondary data analysis. Finally, the data from both methods were used for coding and analysis, which led to the eventual results.
1. Semi-structured interviews
First, semi-structured interviews were held with the employees in order to provide for a rich context for this study. For these interviews, respondents that are involved with the implementation of big data analytics and innovation in their companies were selected. These respondents were found though personal network and the snowballing method, meaning that the respondents were asked after every interview to name another person that would be suitable to discuss the same topics. A selection was made for respondents from various industries in order to generate findings that are general, rather than industry-specific. The table below depicts the chosen respondents, their function and the names of their companies. Company (B), was eager to be interviewed, but wished for their company name to remain anonymous for the purpose of this study.
The interviews were designed to obtain in-depth insights into big data analytics capabilities acquisition choices and big data-driven innovation. The aim was to discover the main challenges that managers faced. The conversations all lasted between 45 and 75 minutes and were held at the offices of the respondents.
Table 1: List of Respondents
Respondent Respondent function Company
(1) Senior Project Manager (A) IBM
(2) Senior Project Manager (B) Anonymous. Large, multinational market research company.
(3) Regional Manager (C) Cofely
(4) Innovation Manager (D) Philips
(5) R&D Manager (E) Staver
In the interviews, the following four questions were discussed. 1. How big data analytics was adopted in the company
2. What needed to be changed in order to adopt big data analytics 3. What were the challenges pertaining to these changes
4. How were these challenges overcome
The answers to these questions were then noted down and integrated in the qualitative data analysis computer program Atlas.ti for coding. From the interviews a number of recurring topics and themes relating to every one of the four main questions were identified as presented in table 2. For question three and four, the focus of the analysis remained on the cultural challenges of the organization although other challenges were also discussed by the participants. The codes were then used as a basis to perform the secondary data acquisition part of this study.
Table 2: Content analysis codes deducted from semi-structured interviews
Question number Codes
1. How big data
analytics was adopted
- Center of Excellence (CoE)
- Hiring a team of data analytics professionals that work within the organization
- Partnership/outsourcing
- Combination of partnership/outsourcing and internal team of data analytics professionals
2. The necessary
changes
- Top management change
- Cultural change – communicational - Cultural change – practical
- Process changes (financial accounting, objectives, production) 3. Cultural challenges - Changing data insights mindset
- Changing daily practices - Communication
- Outsourcing – adapting to data analytics partner 4. Solutions to
challenges
- Communication - Education
- Not, challenges are still present
2. Secondary data acquisition
After the interviews were completed, secondary data was collected from a selection of articles concerning big data analytics implementation through partnerships, data-driven innovation and organizational culture. In order to access these articles, a news, journal and magazine article
database called LexisNexis was used. This database allows the user to enter search queries and filter them on the basis of type of source, country, date and many more filters. The focus of this study is on big data analytics implementation through partnerships, big data-driven innovation and the organizational culture challenges resulting from the implementation. For this reason the following search queries were used: “Big data analytics capabilities”, “Big data analytics implementation”, “big data analytics outsourcing”, “partnership”, “data driven innovation” and “culture”. Through
LexisNexis a combined number of 998 hits were then found that were related to all or some of the search queries used. These results were then filtered on duplication of articles and the quality of their sources. Furthermore, every article was checked to ensure that the article related to big data
implementation using a partnership and/or big data implementation for the purpose of big data-driven innovation. By making this distinction between articles, two respective clusters of articles were established. Doing so enabled the analysis to be more structured and to look deeper into what organizational culture challenges were related to big data implementation through a partnership, and what organizational culture challenges were related to big data-driven innovation. The quality of the articles in this study is defined by the articles’ objective nature (fact-based, rather than opinion-based) and use of reliable sources (academic literature, statistics, observations or interviews). These relevancy and quality filters resulted in 61 of articles suitable for this research (See Appendix 1).
3. Coding and Analysis
The analysis was conducted through a method named content analysis, which is often used for the coding and classification of qualitative data with the aim to discover important findings (Neuendorf, 2002). With content analysis, two types of coding can be used, one with which categories are determined prior to the analysis (inductive approach), and one in which categories were developed during or after the analysis of the data (deductive approach) (Kohlbacher, 2006). For this study, the secondary data were coded using the codes retrieved from the interviews (see table 2), thus using a deductive approach. Every relevant excerpt or quote from the articles was then clustered under one of the codes from the row 2 of table 2. This resulted in a selection of extracts of both semi-structured interviews and articles related to their respective codes. These pieces of text were then further divided into clusters with recurring themes. The themes that followed were thus deductively
acquired first order sub codes. These sub codes each explained an area in organizational change that was an issue to a number of organizations mentioned in the interviews and in the articles. Thus, the sub codes, together with their accompanied quotes and excerpts led to the results of this study, which will be presented in the following chapter. Although all codes, sub codes and their related quotes and excerpts were interesting and provided for a relevant and useful context of this study,
emphasis was placed on the codes and sub codes of question number 2, 3 and 4 from table 2, because these are directly related to the main research question of this study.
3.3. Reliability and Validity
The issue of reliability is concerned with whether alternative researchers would reveal similar
information given a chosen research method (Easterby-Smith et al., 2008). Therefore, this study used a mixed-method research approach for the collection of data. A mixed-method approach has the advantage of using a phenomenon from a multitude of perspectives. This decreases the risk of research bias to occur. Furthermore, the analysis of the multiple case studies and secondary data were based on content analysis approach, which enables a structured and repeatable way of data coding and analysis. The interviews that were conducted for included a key topics and questions so as to enhance this study’s repeatability. Research validity is obtained through the use of multiple interviews and multiple sources of data collection.
3.4. Limitations
Qualitative research is a reliable and valid approach for the description, discovery and exploration of new phenomena. However, the approach does not lend itself to generalizability. Although limitation has been reduced by the use of a multiple case study approach, this study needs to be conducted and redesigned in a quantitative fashion in order for the results to be applicable for a wide range of practitioners.
4. Analysis & Discussion
Big data analytics implementation through a partnership with a data analytics firm for the purpose of data-driven innovation is a very rare research topic in literature. In order to explore what the
organizational culture challenges are that come with this topic, the knowledge from the employees who have dealt with this issue and the articles that are related to this topic have been consulted in this study. By coding and classifying these interviews and 61 articles (see Appendix 1) related to the implementation of big data analytics, data-driven innovation and the related organizational culture challenges, the results of this study were found. The following section is a presentation of these results together with a discussion about how the results of this study compare to prior literature relating to similar research topics.
4.1. Organizational culture challenges
In the previous chapter it was mentioned that although the articles and semi-structured interviews provided for insights into several aspects of big data analytics implementation, emphasis of this study lies on the organizational culture challenges and their solutions. These study results were found, using a deductive and inductive content analysis approach of which the outcomes are outlined in table 3 and 4 below. First, table 3 describes the sub codes that were deducted from the 31 articles (out of the total amount of 61 articles) that related to the question of what challenges companies faced when outsourcing their big data analytics capabilities to another company. Table 4, then, depicts the themes that arose when exploring the 20 articles (out of the total amount of 61 articles) for the challenges that related to the pursuit of big data-driven innovation.
Table 3 - Organizational culture challenges of big data analytics implementation through partnerships
Sub codes (Themes) Description Number of articles in
which sub code was mentioned out of 31 Adaptation Adapting business processes, systems, and retrieval
and storage to the data analytics provider and the big data analytics system.
20
Communication with external (data analytic) partner
Sharing insights with data analytics partner and other players in the market to exchange ideas.
15
Table 4 - Organizational culture challenges of pursuing big data-driven innovation
Sub codes (Themes) Description Number of articles in
which sub code was mentioned out of 20 Top management
support
Commitment from top management to promote the potential of big data analytics, to encourage
employees to use it and to facilitate big data analytics insight exploitation.
10
Adopting big data analytics insights
Adoption of big data insights in communication with internal peers and in practice. Big data-driven innovation thrives when insights are shared between different employees within a company.
10
The following sections will elaborate on the themes displayed in tables 3 and 4 and will connect them to the extant literature that is available about these issues. Then, through the conclusions from the interviews and again the existing theoretical knowledge, solutions for these challenges are proposed in the next subchapter.
4.1.1. Organizational culture challenges of big data analytics implementation through
partnerships
Table 3 presents the results of categorizing the articles and interviews related to the organizational culture challenges a company faces when it attempts to implement big data analytics by using a partnership with a data analytics provider. ‘Adaptation’ and ‘Communication with external (data
analytics) partners’ were found to be the most prominent challenges in the articles and interviews. In the coming sections these issues will be discussed by first defining the challenge, then the impact these challenges have on the business will be discussed using examples from the excerpts that were found in the analysis of the articles and interviews and then it will also be explained what a firm will benefit if these issues are solved. In the next subchapter, the solutions to these challenges will be discussed.
Adaptation
The decision to outsource big data analytics is often not only made because a company lacks the appropriate technological resources, but often also because of a lack of the right in-house data analytics labor. Outsourcing big data analytics often means for companies that their already existing data analytics teams need to collaborate with the employees of the data analytics startup with which they are to partner. If a firm has selected the right data analytics partner, their team would be able to augment current data analytics capabilities with its experience and expertise (Hoffman, 2007). Working together with another partner who plays a major role in the daily practices of a company presents additional issues to an organization’s culture, especially when company-specific data is involved. Therefore, adapting business processes, methods and the retrieval and storage of data to the big data analytics partner, was frequently found in the articles and interviews to be 20articles it was mentioned that adaptation was hindering the implementation phase for big data analytics companies that had used a data analytics partner. The issue of cultural adaptation, thus, means that the organizational culture of a firm must be aimed towards the generation of new insights. In order for big data analytics practices to become successful, the data that is being analyzed needs to have the potential to deliver value. In an interview with company C, a manager stated that big data analytics success is much tied to the quality and availability of data that an organization can contribute. In a similar fashion, one article reports that “when bad data goes in, bad insights will come out”. When information is not placed into the right context of the business needs, big data analytics initiatives would not yield the desired results. Companies that rely mostly on their data
analytics partners to obtain big data insights will need to deliver quality data from which their partners can retrieve valuable insights.
Adapting the culture to insight generation is critical to the success of the partnership between a data analytics firm and its client and also for the success of the implementation of big data analytics capabilities. The company that receives the data analytics capabilities must be able to adapt its own processes and systems in order to provide the data analytics firm with the right raw materials to build big data insights from. This does not only hold true for the analytics department of the receiving firm, but any other department that may have to do with the provision of data or knowledge, such as marketing teams, sales teams, HR teams, production facility teams, accounting teams and administration teams. All the employees need to adapt to new rules and regulations that will make the data analytics provider work more easily. They must obtain an understanding of the value of big data analytics and have to report new big data analytics possibilities when they arise. This is not only a task for (data) analytics teams, but for the whole organization, because big data analytics possibilities for the generation of data or the usage of new insights can come from a variety of sources. Thus, outsourcing big data analytics requires from an organization that employees at all levels and departments adopt a ‘big data analytics mentality’: meaning that they perform daily activities with the intention to improve the organization’s big data analytics capabilities, either through the usage of big data analytics insights or by exploring new big data analytics possibilities. Moreover, in order for an organization to efficiently work together with another organization it may take a lot of time, effort and energy before all systems are set in place. When a firm builds its big data analytics capabilities in-house, rather than through a partnership, it acquires and develops its own combination of tools, methods and trained professionals. Outsourcing big data analytics capabilities and adapting to one data analytics firm makes it unnecessary for a firm to develop their own big data analytics capabilities. It is therefore easy for competitors to imitate their way of competing on data. Next to increased profitability and a better market position, one reason why companies choose to implement or acquire big data analytics capabilities is for the improvement of
