The effects of big data on process innovation
An exploratory study
Master Thesis Joris Jongerius 10324038
MSc Business Administration: International Management University of Amsterdam, Amsterdam Business School Supervisor: Dr. M.P. Paukku
Statement of originality
This document is written by Student Joris Jongerius 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.
Abstract
Big data is a relatively new phenomenon that is becoming increasingly important (McAfee and Brynjolfsson, 2012). The impact of big data is already identified within several business areas, for example, marketing and supply chain management. However, the effects of big data on innovation have not been investigated yet, although being a significant driver of competitive advantage. According to many researchers, big data is the next big thing within innovation (Strawn, 2012; Gobble, 2013). This research investigates the opportunities of big data within process innovation. This exploratory research has acquired qualitative data through semi-structured interviews with four data managers and four data analysts. These data employees are working at four major insurance companies in the Netherlands. Overall, the interviewees generally agreed with the three propositions. Therefore, they agree that big data influences process innovation in such a way that it becomes more focused, faster and more varied. These results indicate that firms can benefit from big data. This research can be used as a first step for further research about the relationship between big data and process innovation.
Table of contents
1. Introduction and research question 6
1.1 Goal 8 1.2 Relevance 8 1.3 Outline 9 2. Literature review 10 2.1 Definitions 10 2.2 Big data 11
2.3 Big data in practice 13
2.3.1 Conclusion 15
2.4 Innovation 16
2.5 Process innovation 19
3. Big data and process innovation 25
3.1 Motivation and propositions 25
4. Research design 28
4.1 Research structure 28
4.2 Research approach 29
4.3 Research setting 30
4.4 Data collection 31
4.5 Validity and reliability of the study 32
4.6 Case description 33
5. Analysis and results 36
5.1 Analysis 36
5.2.1 Conclusion 42
5.3 Data analysts 43
5.3.1 Conclusion 48
5.4 Comparison of data manager and data analyst within the company 49
5.4.1 Delta Lloyd 49 5.4.2 AEGON 50 5.4.3 Zilveren Kruis 52 5.4.4 Nationale Nederlanden 53 5.5 Overall result 54 6. Discussion 57 6.1 Theoretical implications 57 6.2 Practical implications 59 7. Conclusion 60
7.1 Limitations and suggested research 61
1. Introduction and research question
Every interaction on our mobile phone, the Internet or social media is creating data. These are only a few examples of actions that create data. According to Verhoeye (2015), 2.5 billion
gigabytes of data were produced every day in 2012. As more and more people will have access to the Internet and will use a smartphone, one can expect that this amount will increase
significantly. These enormous amounts of data can be referred to as ‘big data’ (McAfee and Brynjolfsson, 2012). Given this massive scale, it could be logic to understand big data solely in terms of volume. But next to volume, there are two other requirements when one can speak about big data. These are variety and velocity (McAfee and Brynjolfsson, 2012). As said, the volume is about the enormous amount of data. The variety is about the fact that the data that is collected is highly unstructured, and the velocity is added because it is fast moving data (McAfee and Brynjolfsson, 2012).
According to the McKinsey Global Institute (2011), big data is becoming a key source of a firm’s competitiveness and its advantages. The Institute estimated that the overall annual gains from big data would be US$ 610 billion due to increased productivity and cost savings. EY published a report in 2014 stating that firms that derive value from their data will have an advantage over their competitors. This advantage, which is called a performance gap, will continue to grow, as more relevant data will be generated. However, no numbers are given to verify the advantage. McAfee and Brynjolfsson did a similar study in 2012. They identified that companies characterised themselves as users of big data have better financial and operational results. To be more specific, companies that are using big data are about six percent more profitable and five percent more productive than their competitors (McAfee and Brynjolfsson, 2012). Also, the downside of big data has to be considered. Abuse of large databases and privacy violations are potential threats that big data possesses. However, big data will bring
improvements in everyday life (Bollier and Firestone, 2010). Further, it is expected that firms will be competing on analysing big data in the next few years. Firms that will not use big data will lose the competition (Bollier and Firestone, 2010).
As can be seen from above-mentioned examples, big data can have a positive effect on a firm’s performance. At this moment, the management literature of big data can be divided into two main categories. The first category is about the results and implications of business analytics and is concerned about how big data can be used for optimisation of business (Chen, Chiang and Storey, 2012). The second category is about marketing analytics and is focused on understanding how customer data can reveal future customer behaviour (Sagiroglu and Sinanc, 2013). The reason why these two categories are getting so much attention is that their results can be easily measured within firms, which means the importance of big data can easily be validated
(Davenport, 2006).
However, according to Koellinger (2008), the potential value of big data can go further: The insights gained from previously mentioned analytics might provide a firm with valuable knowledge about the business and market. These could then be used for the innovation of
products and services. Kemp, Folkeringa, Jong and Wubben (2003) identified that innovation is a huge determinant of a firm’s performance. When using big data for innovation, it could even increase the success rate of the innovation activities. Therefore, big data-driven innovation is gaining attention by scholars, but it is still not fully understood and thus studies need to be conducted for its validity. Of course, innovation can be used for several purposes. One can think about product innovation, process innovation, brand innovation and so forth. However, according to Davenport (2014), big data analytics is mostly used for process innovation. At first, it was used within supply chains and practical processes, but is nowadays also used for process
never can stop with innovating their processes. Otherwise, competitors will adopt the same process innovations and can quickly catch up.
As can be concluded from above, the relationship between big data and innovation is not yet fully understood. Although innovation can be done in several ways, process innovation can be seen as the most important form of innovation. Therefore, this thesis will focus on the effects of big data on process innovation and thus the following research question can be formulated:
‘’What are the effects of big data on process innovation?’’
1.1 Goal
The objective of this study is to give insights into the problem that has been identified in the previous section. The study will focus on the effects that big data has on innovation and, in particular, process innovation. As this study aims to explore the effect of big data on process innovation, it is important to review the existing literature about big data, innovation and process innovation, before exploring the usage of big data and see what its effects are for process
innovation. As there is not much known about this relationship yet, this study will be an exploratory study. Therefore, semi-structured interviews will be held.
1.2 Relevance
This study should contribute to the relationship between big data and process innovation, one that is currently not well investigated and understood. As already mentioned before, firms that use big data will have a competitive advantage, therefore making big data important. As process
between these two. The findings of this research give insights about what the effects of big data are on process innovation and may help to convince firms to use big data in process innovation.
1.3 Outline
This thesis is organised as follows: in the second chapter, the current definition of big data will be given next to a general overview of what big data is. The literature review will continue with current literature about innovation and process innovation. Continuously, the working
propositions for the research will be derived from the literature. These will be used to investigate the effects of big data on process innovation. Thereafter, the data collection and research method that is used will be given. Following this, the results will be presented and be discussed and an answer is given to the research. Lastly, a conclusion will be drawn from the findings, limitations of the research will be discussed and suggestions for future research are given.
2. Literature review
In this chapter, relevant literature about big data, innovation and process innovation for this thesis will be discussed. It starts with an overview of what big data is and how it should be used. The next part will be about how big data is implemented in business nowadays. Certain businesses are selected where big data already has positive results, such as marketing and supply chain
management. Finally, innovation will be discussed and in particular process innovation and some working propositions will be developed. Because big data, innovation and process innovation are the key terms within this study, a short definition of them will be given below in the definition section. However, within the specific sections, there will be elaborated further on the key terms.
2.1 Definitions
Although there is not one overall accepted definition, big data can be best described as data that has a large volume, comes at a high velocity, and has a wide variety. Next to these properties, big data also has to create value for a company and a company needs to ensure that the data is
trustworthy.
According to Damanpour (1996), innovation is a change within an organisation, either as a response to changes in the external environment or as a pre-emptive action to influence the environment. Innovation encompasses new process technologies, new products or services, new organisation structures, new administrative systems, and new plans or programs for members within the organisation.
Process innovation is, according to the OECD (2005), the implementation of a new or significantly improved delivery or production method. This includes changes in equipment, software or techniques.
2.2 Big data
In the past few years, many scholars and researchers have created a number of definitions and explanations about big data. In the beginning, researchers referred to big data when they were talking about the size of it. Cuzzocrea, Song and Davis (2011) refer to big data as enormous amounts of unstructured data that are produced by high-performance applications. According to Gobble (2013), big data can be seen as a dataset that is too big to capture, store, manage and analyse. It is a large dataset, which requires advanced storage and management to use it (McKinsey Global Institute, 2011). Brynjolfsson and McAfee (2012) add numbers to quantify how enormous this big data is. The volume of data that was created in 2012 was 2.5 Exabyte each day. The data was then expected to double every forty days. Therefore, it can be defined as an enormous amount of data, which consists of Terabytes, Exabytes or even Zettabytes (Aiden and Michel, 2013).
Whereas in the beginning, the definition of big data was about how big the data was (volume), the definition switched to another point of view. Researchers and scholars added two dimensions to the definition of big data, namely velocity and variety. According to Zikopoulous, Parasuraman, Deutsch, Giles, and Corrigan (2013) and McAfee and Brynjolfsson (2012), the definition of big data is about the three ‘V’s’ – Volume, Variety and Velocity.
Volume is the first property of big data because of the enormous quantity of data. This quantity is big enough to make specific decisions. Every second nowadays there is more data created than there was in total twenty years ago (McAfee and Brynjolfsson, 2012). Not only more different kinds of data are collected, but also more data of a specific phenomenon.
Variety is the second property of big data. This is because there are many different kinds of data available nowadays. One can think about GPS locations, updates and pictures posted on
social media, customers that click on certain things at websites. Most of this data is highly unstructured (Zikopoulous et al., 2013).
Velocity is the third and last property that Zikopoulous et al. (2013) describe. There is constantly new data available. New technologies have made it possible to increase the speed at which this data is collected. It is about the frequency of data delivery and frequency.
As big data is under continuous development, so is the definition of big data. After the introduction of the three V’s by McAfee and Brynjolfsson (2012) and Zikopoulous et al. (2013), two additional V’s were introduced. The fourth V that is part of the definition is veracity.
Veracity is added because not all the data is trustworthy. The inherent unpredictability of some data requires analysis of big data to gain reliable prediction. For example, one out of three managers does not trust the data that they use for their decision-making processes.
The fifth and final V is value. Value is added to see what big data is actually worth. Does big data generate economically worthy insights and benefits through extraction and
transformation? According to Verhoeye (2015), a company is becoming more profitable when its access to and collection of data is better and higher. It can, therefore, create a competitive
advantage. As said before, the definition of big data is still under development. However, in this paper, when referring to big data, it is about big data as described in the 5 V’s – volume, variety, velocity, veracity, and value.
According to Gandomi and Haider (2015), big data on its own is worthless. This means that big data alone cannot lead to a competitive advantage. Only through analysis, big data can become valuable. McAfee and Brynjolfsson (2012) state that big data can be seen as a tool to increase a firm’s performance. For example, the decision-making process is enhanced through a better visualisation of a firm’s operations and also improved performance measurement
that it opens up new opportunities. Therefore, using big data to create or support innovative ideas seems to be a good strategy for firms to engage in.
2.3 Big data in practice
Because of big data, managers can know more about their business. They can directly translate this knowledge into improved decision-making and performance in almost every part of the firm (McAfee and Brynjolfsson, 2012). They give as an example that big data has led to
improvements in supply chain management. This is one of the business areas where big data has an effect on. In this section, some examples will be given about big data and its positive effect on particular business areas, like marketing and supply chain management. These business areas are chosen, because it is clear that big data has actual positive effects on these areas. For example, as big data contains so much information, it can lead to increased understanding of consumer needs (EY, 2014). Several other mechanisms, influenced by big data, are supporting marketing
initiatives, enhancing fraud monitoring, and improving decision-making. All of these can be increased, which leads to increased value, because big data contains rich information (EY, 2014). However, it should be clear that not every business area is included. George and Lin (2017) support the findings of EY. They argue that big data transformed the organisations, because big data enables firms to better manage their manufacturing processes and create efficiencies in managing customers. However, according to them, it is not yet investigated how big data can influence innovation (George and Lin, 2017). Although they created a framework how to approach innovation with big data, no results or implications for innovation have been given. Innovation can create a competitive advantage, however, the effects of big data on this business area have not been discovered yet (George and Lin, 2017).
Supply chain management
Waller and Fawcett (2013) did a study about the effect of big data on supply chain design and management. Big data has deep implications for supply chain management, as it will transform the way supply chains are designed and managed. This will lead to new opportunities. Some examples of potential applications of big data are human resources, customer and supplier relationship management, and forecasting. Within these categories one can think of forecasting time of delivery, improvement in inventory system accuracy, taking into account traffic
congestion and more effective monitoring of productivity (Waller and Fawcett, 2013). It is only suggested in which parts of supply chain management big data can be used, meaning that there are not yet actual results. However, it is expected that using big data will lead to a reduction in costs, especially in inventory, labour, and transportation costs (Waller and Fawcett, 2013).
Service-oriented strategies
In 2015, Opresnik and Taisch studied the value of big data in service-oriented strategies among manufacturers. According to them, services in manufacturing are a necessary condition for creating competitive advantage. Big data has effect in multiple ways on how these services should be operationalized. First of all, new revenue streams can be created, as big data offer valuable insight in which services are missing in the current market. It also involves entering new markets (Opresnik and Taich, 2015). Second, there is a possibility to decrease prices for product-services, because new knowledge can be created through big data. As big data is known for its variety, knowledge can come from many different types of data. For example, through big data, the company can see which services are valued most and which services are redundant (Opresnik and Taich, 2015). Third and last, manufacturers can differentiate themselves from others that are not using big data. On the one hand, this is because companies that are using big data have more
information and insights about their services. On the other hand, gathered information can be reused and even, in the future, can be sold to other service-oriented manufacturers (Opresnik and Taich, 2015).
Marketing
Normal data, which is not as big, fast and varied as big data, provides behavioural insights about consumers and marketers try to translate these insights into market advantage. The rise of big data, meaning more volume, velocity and variety, will lead to new ways of understanding consumer behaviour (Erevelles, Fukawa, and Swayne, 2016). Marketers with access to these big data will understand its consumer needs better. This means in practice that marketers are able to make better decisions based on evidence at a given time, rather than on intuition or laboratory-based consumer research. Therefore, also personalised marketing will have a better effect, as the company now has more information about someone. Finally, big data offers firms new ways how to differentiate its products from competitors. According to Erevelles et al. (2016), big data has the potential to impact marketing in almost every area. Firms that will not use big data will need to create another competitive advantage; otherwise, they will not survive the big data revolution.
2.3.1 Conclusion
In several business areas, big data is already playing an important role. For example, within supply chain management, big data leads to improved inventory system accuracy and more effective monitoring of productivity (Waller and Fawcett, 2013). Within marketing, big data leads to a better understanding of consumer behaviour, which enables marketers to make better decisions (Erevelles et al., 2016). However, the effect of big data is not investigated and
innovation. Innovation is an important business area, because it is a key aspect of creating competitive advantage. However, this will be discussed more in detail in the next section.
2.4 Innovation
Within the literature, several authors (McAfee and Brynjolfsson, 2012; Weber, 2016) have stated that big data is a very helpful tool to drive the innovation process. The variety and velocity of big data are useful for new ways to serve markets with innovative products and services. Big data has not only the potential to have an effect on product and service innovation, but also can be used for optimisation of business processes and consumer insights (LaValle, Lesser, Shockley, Hopkins, and Kruschwitz, 2011). Gobble (2013) states that big data is ‘the next big thing in innovation’, whereas Brown, Chui, and Manyika (2011) state that ‘big data can be the next frontier for innovation, competition and productivity’ (p. 24). Big data can help organisations to create business value by discovering new organisational capabilities and value (Davenport, Barth, and Bean, 2012). Most researchers agree that big data can lead to positive effects for innovation. As big data is a relatively new phenomenon, it could be seen as an innovation on its own.
However, this study focuses on what the effects of big data are on innovation. This means that big data is considered to support innovation, by, for example, making it faster or more focused rather than being an innovation itself. Now, innovation will be explained and how it is
operationalized.
Big data has a few different definitions and the same can be said about innovation. At this moment, the simplest definition to understand, however not used the most, comes from
Thompson. He states that ‘innovation is the generation, acceptance and implementation of new ideas, processes, products or services’ (Thompson, 1965, p. 2). Wong, Tjosvold, and Liu proposed a more recent but relatively same definition in 2008. They define innovation as ‘the
effective application of processes and products new to the organisation and designed to benefit it and its stakeholders’ (Wong et al., 2008, p. 3). However, the most quoted definition of innovation till today is the definition by Damanpour (1996, p. 694): ‘Innovation is conceived as a means of changing an organisation, either as a response to changes in the external environment or as a pre-emptive action to influence the environment. Hence, innovation is broadly defined here to encompass a range of types, including new product or service, new process technology, new organisation structure or administrative systems, or new plans or program pertaining to
organisation members’. Research done by Baregheh, Rowley, and Sambrook (2009) showed that there are more than sixty definitions of innovation at that time. Cooper already suggested in 1998 that one of the challenges of innovation is the lack of a common definition. However, when speaking about innovation in this thesis, it is in line with the most used and quoted definition given by Damanpour in 1996.
Furthermore, next to the lack of a common definition, there is sometimes confusion about the difference between invention and innovation. Invention can be seen as the first occurrence of an idea for a new product or process, whereas innovation can be seen as the first
commercialisation of the idea. Next to this difference, inventions may be carried out anywhere, for example in universities. Innovations, on the other hand, mostly occur in firms in the
commercial sphere (Rogers, 1995).
Leaving definitions aside, the important questions about innovation are of course how it is operationalized and what the effect of innovation is. According to many different researchers, the innovation process consists of four steps. These are ideation, selection, development, and
implementation (Tidd, Bessant, and Pavitt, 2005; Chandra and Neelankvil, 2008). As the word already says, ideation is the phase in which new ideas and plans are formed. These new ideas can come from different sources, for example, customers, advertisements, and data. The second
phase, selection, is the phase where every idea that has been developed in the first phase is critically analysed and eventually chosen if it has potential value. The third phase is developing these new ideas or plans. One can think of testing or validating a new service or build a prototype of a new product. The fourth and final stage is implementation. A new service, product, or
process is added to the value chain of a firm (Chandra and Neelankvil, 2008). Belkenstein already developed a relatively similar model about the innovation process in 1998. His model is called the ‘innovation funnel’. This funnel consists of the same four stages as the above-described model. Only the naming of the four phases is different. One important aspect he mentioned is that the opening of the funnel has to be as big as possible. This way, every idea that comes from the environment of a company will be taken into consideration. One can think about ideas that come from suppliers, customers, partners, and competitors (Belkenstein, 1998).
Understanding how innovation is operationalized is important to firms because successful innovation is a necessity in order to survive competition (Crossan and Apaydin, 2010).
Innovation can create resources that are valuable, rare, inimitable and costly to copy by
competitors, therefore creating a competitive advantage (Barney, 1991). Cho and Pucik (2005) argue that innovation leads to an increase of the firm’s strategic resources and to sustainable competitive advantage. Continuously, a positive relationship between innovation and firm performance is expected because firms can profit from developing new products, new business models and more efficient processes. This is in line with findings by Porter in 1990. He stated that firms want to stay competitive in their market. In order to do so, firms have to renew their way of production, renew the products and services they are offering, or change the market in which they want to compete. Furthermore, other researchers have found that the relationship between innovation and firm performance is positive. For example, Bowen, Rostami and Steel (2010) conducted a study to test this relationship based on 55 empirical studies conducted
between 1975 and 2005. They came to the conclusion that the relationship is indeed positive and is an important condition for creating a competitive advantage. Within innovation, there are of course many forms. One can think of product innovation, brand innovation, and process innovation. However, covering all these forms of innovations is too broad for this thesis.
According to Salge and Piening (2014), the capability to introduce and use process innovation is one of the most important competitive resources. When prices and products are relatively the same, firms should compete on processes. The firm with the most efficient processes can lower its prices and therefore create competitive advantages. Furthermore, George and Lin (2017) suggest that big data can be used for process innovation. For example, analysing big data to reduce cycle times between concept and testing of new processes. According to Conway and Klabjan (2013), there are many best practices for implementing big data, but the most important two are data quality optimisation and process innovation. Taking these arguments into
consideration, and to create a more focused research, this thesis will focus on process innovation.
2.5 Process innovation
The OECD (2005) states that process innovation can be seen as the adoption of technologically new or notably improved production methods. Keupp, Palmié and Gassmann (2012) added two more dimensions where process innovation can have an effect. According to them, process innovation can be best described as the introduction of new or notably improved production methods, administrative processes, or supply chains. Process innovation is different from product innovation as it leaves the product functionality unchanged, but is lowering the cost of production or processes (Adner and Levinthal, 2001). First, one must understand the existing process in order to create an idea about how innovations can be implemented. Thereafter, one has to design the new process in detail before it can actually be implemented (Davenport, 1992). Information
about the processes is, of course, the most important resource for enabling and implementing process information. It should be clear that accurate and real-time information on process performance is of great significance for the effectiveness of process innovation (Davenport, 1992).
According to Fuglsang (2008), process innovations consist of three phases. These are the idea phase, the development phase, and the implementation phase. Whereas innovation consists of four phases, process innovation consists of three phases, according to Fuglsang (2008). This is because the ideation and selection phase are taken together by Fuglsang (2008) in the idea phase. During the first phase, ideas are generated through inspiration, talking with other people or from information sources. This phase consists of searching for innovate solutions, recognising a need for an innovation, studying existing innovations and selecting potential process innovations (Fuglsang, 2008). In the development phase, proposed ideas are evaluated. After acceptance, a development process starts to develop the idea. One can think of implementing and testing the new process for a few weeks. The final phase is the implementation phase. During this phase, the organisation is prepared for the use of the innovation, the acceptance of the innovation, and making the innovation a routine. During all these phases, information is crucial (Fuglsang, 2008). According to George and Lin (2017), information collection always has a significant role in innovations. Through analysing information, more ideas could be generated and improved decisions could be made. However, because analysing big data reveals significantly more information, big data is highly suitable for process innovations.
Within process innovation, there are two forms distinguished. These are technological process innovation and organisational process innovation (Reichstein and Salter, 2006). Technological process innovation encompasses new IT-infrastructure, new processing
lower production costs. New software technology can be used to monitor operations more precisely, improve manufacturing efficiency, or increase customer satisfaction (Shallock, 2010). This form of process innovation can be seen as the most difficult form of the two. This is because new technologies can build on existing knowledge to some extent, but often requires big changes to existing structures and skills. This can lead to confusion, uncertainty, and a lack of consensus and mutual understanding. However, when more information is available and communicated regarding the innovations, these problems can be overcome (Frishammar, Floren and Wincent, 2011).
Organisational process innovation refers to new ways of organising structures and work activities within a firm. It includes the development and implementation of change to
organisational structures, administrative systems, existing processes, and management procedures (Damanpour and Aravind, 2012). Birkinshaw, Hamel and Mol (2008) identify three reasons why organisational innovation can be a challenging task. First, organisational process innovation needs serious amounts of knowledge and is difficult to identify. Second, many companies lack expertise in developing and implementing this form of innovation. Third, changing the current structure of the organisation frequently leads to uncertainty amongst employees (Birkinshaw et al., 2008).
Although there are two different forms of process innovation distinguished within the literature, most of the time process innovation encompasses both technological and organisational innovation (Reichstein and Salter, 2006). For example, Hervas-Oliver and Sempere-Ripoll (2015) provided evidence from a large survey to argue that the effectiveness of both forms of process innovation are mutually reinforcing. By combining efforts of technological and organisational innovation, firms can achieve significant performance improvements. Likewise, Hollen, Bosch, and Volberda (2013) suggest that technological process innovation requires organisational
process innovation to face the managerial challenges that arise from the technological process innovation.
Competitive advantages and positions arise quickly from process innovation, because of changing customer needs, legal frameworks, and market situations (Damanpour, Walker and Avellaneda, 2009). According to Van der Panne, Beers and Kleinknecht (2003), process innovation is also playing a key role in the financial performance of an organisation. This is because it makes organisations more profitable and grow faster, as process innovation contributes to cost reductions, quality improvements, and productivity gains. Implementing these cost
reductions or supply chain technologies help a firm either to retain a higher profit margin or reduce the price for consumers due to the cost savings. This might lead to higher sales and market shares. Robertson, Casali, and Jacobson (2012) agree with the previous argument. According to them, other things being equal, the firm with the most efficient process will be the most profitable firm. When the products are almost equal, price becomes the central strategic variable. By
lowering cost structures and due to more efficient processes and as a consequence the price, firms can gain market share (Robertson et al., 2012). Continuously, process innovation contributes to turnover growth. Likewise, an increase in the intensity of using new or improved processes was found to boost the firm’s sales growth (Klomp and Van Leeuwen, 2001). According to Davenport (1992), almost all process innovation initiatives arise from the need of improving financial
performance. For example, improved customer service and quality are assumed to translate into higher sales. Another aspect why firms focus on process innovation is the objective of process time reduction. By innovating the process, redundant variables can be excluded from the process, which, in turn, makes the whole process faster (Davenport, 1992).
As can be concluded from the previous section, process innovations can create a competitive advantage. Enhancing these process innovations can lead to an even better
competitive advantage (Robertson et al., 2012). According to Purcell and McGrath (2013), external knowledge is important for all sorts of innovation processes. Firstly, the use of external knowledge avoids the overreliance on internal knowledge. Secondly, external knowledge allows firms to gather new knowledge. Combining internal and external knowledge can also create new knowledge. This external knowledge has an increased level of variance, which in turn can increase the variance of search outcomes for innovations, meaning that there are more ideas formed during the ideation phase (Purcell and McGrath, 2013).
Jurado, Gracia, and Lucio (2009) conducted a study about the effects of external knowledge on process innovation. The study was based on external knowledge gathered from industrial and scientific agents. They found that process innovation is largely driven by external knowledge. The external knowledge can be a good source of information to give insights how existing processes work and whether new processes are needed or existing processes need to be innovated (Jurado et al., 2009). With these insights, firms are able to better structure their processes and know better where processes need improvements or innovations. This means that firms can better target process innovations (Jurado et al., 2009). Although Jurado et al. (2009) do not give an indication which stage of the process innovation is mostly influenced, one can argue that it is the idea (and selection) phase. This is because, through the external knowledge, more ideas are generated and firms know better where to innovate which can be used to select the best ideas.
According to Davenport (1992), information is an important tool for developing and implementing process innovations. To increase the performance of these innovations, accurate and real-time information is a prerequisite. When more information is collected, it is likely that new insights are gathered. As a consequence, decisions regarding the innovations tend to go easier and faster. This leads to a decrease in the total time a process innovation takes (Davenport,
1992). However, Davenport (1992) only mentioned the developing and implementing stage of process innovation and not the ideation stage. George and Lin (2017) suggested, as also
mentioned before, a relatively same option. By using big data, the cycle times between concept and testing and between scaling and prototyping can be reduced. Therefore reducing the total time within the development phase of the innovation.
The examples above all indicate that external knowledge and information have a positive effect on process innovation. As big data is part of these external knowledge flows, but was not part of the above research, it is likely that big data also has an impact on process innovation.
As can be concluded from this part, process innovation is an important source for firm performance and creating competitive advantage. As already stated by Gobble (2013), big data is the next big thing for innovation. Therefore, it is important to investigate the relationship between big data and process innovation. In the next section, propositions will be composed in order to test whether big data has a positive effect on process innovation.
3. Big data and process innovation
Within different segments of the business, it is becoming clear that big data has a positive effect. For example, on marketing and supply chain management. However, at this moment, it is not clear what kind of impact big data has on innovation and, in particular, on process innovation. There are researchers stating that big data is the next thing that will have great influence on innovation (McAfee and Brynjolfsson, 2012; Gobble, 2013; Strawn, 2012). At the moment, only a few studies have identified what the real potential of big data is and how it can support several business areas. However, this is not done for the business area of innovation yet. Therefore, there are still many questions left. Although researchers suggest it, does big data really have a positive influence on innovation? Will big data be the best source to make decisions regarding innovation or will big data only be a helpful additional source? Does big data increase the speed of process innovation? An answer, to at least some of these questions, will identify how and what kind of effect big data can have on process innovation. To give insights into the effect of big data on process innovation, some propositions have to be developed. This thesis will now continue with the motivations for and defining of these propositions.
3.1 Motivation and propositions
Section 2.5 already identified some ways – focus, speed, and variety – in which process innovations can be influenced and this section will elaborate further on these characteristics. According to Davenport, a process is ‘simply a structured, measured set of activities designed to produce a specified output for a particular customer or market. It implies a strong emphasis on how work is done within an organisation’ (1992, p. 5). Without some focus on critical processes, the resources and time of an organisation for innovation will not be used in full potential
Therefore, big data contains a lot of potential information. This information can be used to discover new or deeper insights about this set of activities. As a consequence, big data can be used to improve the identification of which activities within a process can be innovated best. This leads to the first proposition:
P1: Usage of big data leads to more focused process innovations
Furthermore, as process innovation is an important source of creating competitive advantage, it is of great value if these innovations are done as fast as possible. If the innovations are quicker, this can, in turn, save time and capital (Damanpour et al., 2009). Because big data has a high velocity, a constant stream of information is coming to the firm. The continuous stream of data may lead to, for example, faster decision-making, which leads to a faster innovation process. This means that the total time a process innovation takes, from the ideation phase to the implementation phase, decreases. George and Lin (2017) support this argument by arguing that big data can potentially decrease the cycle times between concept and testing as well as between prototyping and scaling. Therefore, the total time of a process innovation should decrease. This leads to the second proposition:
P2: Usage of big data leads to an increased speed of process innovations
According to Davenport and Dyché (2013), the volume of big data is not that important. The most important property of big data is its variety. They did a survey at more than fifty large companies. The main finding was that the biggest reward of big data comes from the ability to analyse several data sources and new forms of data. Companies can use these new and additional
data sources also for innovation. However, big data has not been considered a source of variation for process innovation. Therefore, this new information may lead to process innovations in new areas, which increases the diversity of innovations. This leads to the third and last proposition:
P3: Usage of big data leads to a wider variety of process innovations
Of course, there are more possibilities how big data can have an effect on process innovation. However, according to the available literature, these three propositions cover the most important factors. Figure 3.1 illustrates once again how big data has an impact on process innovation.
Figure 3.1: The effect of big data on process innovation
At this moment, the effects of big data on process innovations are not clear. This research will try to identify the effect through the three proposed propositions. The three propositions test whether big data increases the focus, the speed, and the variety of process innovations.
In the next section, the research design will be given.
Big data
P1: focus
Process
innovation
P2: speed
innovation
Process
P3: variety
Process
innovation
4. Research design
This section will give an overview of the research design. The goal of this study is to gain a deeper understanding of how big data can have an effect on process innovation. To gain insights, data managers and data analysts of four different companies were interviewed. This study will now continue with the used research structure, before giving the method how the data was collected and the strengths and limitations of this research design.
4.1 Research structure
According to Yin (2003), there are three different forms of research one can choose from. These are descriptive, exploratory, or explanatory. The first form, descriptive studies, is used to describe processes that are happening. The second form, exploratory studies, is conducted to analyse problems that are not clearly defined yet. The third and final form, explanatory studies, is
conducted to elaborate on causal relationships (Yin, 2003). At this moment, little is known about the relatively new phenomenon big data and even less about how big data can have an effect on process innovation. Therefore, this research has an explorative nature.
Secondly, it will be a qualitative research design. This is because qualitative research is especially suitable for explaining phenomena that are not covered in a large amount of existing literature (Given, 2008). There are several qualitative research methods to gather information. One can think about in-depth interviews, archival data, surveys or observations (Given, 2008). The first option is chosen because it enables the researcher to gather information and knowledge from people that have experience with the aforementioned topics.
Thirdly, one has to decide whether the research will be a single case study or a multiple case study. As this study tries to investigate the impact of big data on process innovation across
multiple firms, it can be stated that this study adopts a multiple case design. A case study is excellent for conducting exploratory research (Yin, 2013).
Finally, as big data is used within firms, the unit of analysis is the firm. Depending on the organisation, big data can be used at different levels within the organisation. Therefore, this study focuses on multiple function levels within a firm.
4.2 Research approach
To get the information that is necessary to answer the research question, in-depth interviews were held. The participants were selected based on a few criteria. Firstly, the interviewees, of course, have to work with big data and use this data, at least partially, for innovation of the processes. This is to make sure that the research targeted the right employees.
The second criterion is that the companies work within the same sector. This was done in order to increase the comparability of the results. According to the Central Bureau of Statistics (CBS) in the Netherlands, the sector that uses big data most is the tourism sector. Of every company within the tourism sector in the Netherlands that has more than ten employees, 44 percent uses big data (CBS, 2016). To get access to these companies, employees of the
companies were identified by using LinkedIn. However, after two days of calling and emailing, only one out of the fifteen contacted companies was willing to cooperate. As a consequence, the study shifted to the sector that is the second largest in using big data. This is the insurance sector, where 39 percent of the companies are using big data (CBS, 2016). These companies were better willing to cooperate and interviews were held at four of them. Later on in this research design, a description of the insurance companies will be given.
The third criterion is that only companies based in the Netherlands are selected for practical purposes. There are no further criteria considered in this study. Therefore no distinction
is made with respect to size or a certain type of insurance companies, such as house-, life-, or health insurance companies.
As already said before, this study focuses on different levels of employees. This is done because, for example, to see whether managers and analysts have the same ideas about big data or have different thoughts about big data and its effect on process innovation. However, when arranging the interviews, it became clear that there are, at this moment, only two function levels. These are the data manager and the data analyst. The data manager is responsible for where big data is implemented and why it is implemented. The data analysts are actually gathering and analysing the big data.
4.3 Research setting
The participants were invited through two different methods. As the author did not have any connection with any employee working at the insurance companies, every contact had to be built from scratch. The first method to establish contact was by identifying people that are working as a data analyst or data manager through LinkedIn. These employees were telephoned or emailed. The second method was, when the first contact was established, to ask participants whether they could provide names of familiar employees that want to cooperate. This resulted in ten
appointments with five data managers and five data analysts in five different insurance
companies. Thus, within every company one data manager and one data analyst. However, one company cancelled the meetings last minute. Therefore, there are actually eight interviews done in total within four companies. All these interviews were held face-to-face.
By using semi-structured interviews, rich and empirical data can be collected in a highly efficient way. Another advantage of this form of interviews is that the interview is clearly guided, but deviation from the original topic is possible if needed (Eisenhardt and Graebner, 2007). In
order to test whether everything is covered and to check whether the questions are well structured or have to be adjusted, a pilot interview has been done. This pilot interview was done with a friend of the author’s brother. This friend is working with big data and innovation. However, he wanted to stay anonymous. The setting was informal and this interview was not recorded because the purpose was to increase the quality and efficiency of the questions and not gathering actual results.
Every interview was recorded with a recording program on a telephone and transcribed afterwards, after approval from the interviewees. A full transcription of the data is important for analysing the data. To analyse the collected data, the software program called NVivo has been used. However, the first fully transcribed interview was also checked on useful information in person by highlighting sentences with a marker. This was done in order to see whether the generated codes for NVivo were correct or adjustments in the codes were needed.
4.4 Data collection
The qualitative data was collected through semi-structured interviews with respondents from four different companies. As the respondents all work with big data, these respondents obviously know a lot about big data. To give an overview of the interviewees, table 4.1 shows the characteristics.
Table 4.1: Overview of respondents
Name Position Company Employees
Frits Goeijenbier Data analyst Delta Lloyd 5.208
Ralph van Gelderen Data analyst AEGON 31.530
Mirjam Wolters Data manager AEGON 31.530
Respondent 5 Data analyst Zilveren Kruis 3.698
Robin Gerichhausen Data process manager
Zilveren Kruis 3.698
Respondent 7 Data analyst Nationale Nederlanden 11.461 Respondent 8 Strategic data
manager
Nationale Nederlanden 11.461
Note: Employees based on 2015, worldwide. Some respondents wanted to stay anonymous
4.5 Validity and reliability of the study
According to Yin (2014), there are four commonly used tests in order to determine the quality of a research design. These are internal and external validity, the reliability of the case study
research, and construct validity.
First, internal validity is created by ensuring that the relationships found by the
researchers are correct (Yin, 2014). Internal validity is most important when conducting a causal or explanatory study. However, it also has to be taken into consideration when doing an
exploratory study. To create internal validity, one must derive the research framework from the literature and through the development of propositions (Yin, 2014). This is done within this research and therefore internal validity is guaranteed.
Second, external validity describes the generalizability of particular findings (Yin, 2014). This study is a multiple case study, instead of a single case study and therefore the results are relatively more generalizable. The fact that the findings are from companies operating in the same sector enhances the external validity further.
Third, there is the reliability of the case study research. According to Yin (2014),
reliability is created by repeating the same steps within a study and thereby generating the same results. To ensure reliability, all the data gathered from the interviews is recorded and thereafter fully transcribed. Furthermore, to increase the chance of getting the same results, the same questions were asked during the interviews. However, there was, of course, room to deviate from the original questions.
The fourth and final test proposed by Yin (2014) is the construct validity. Construct validity defines how well a study measures the construct it claims to be measuring. One way to increase construct validity is to take different perspectives into account. This ensures data triangulation (Yin, 2014). This research focuses on different function levels, namely data managers and data analysts. To ensure a good overview of the different viewpoints, first, the results of the data managers will be presented and thereafter the results of the data analysts. Finally, also a comparison between the data manager and data analyst from the same company will be made.
4.6 Case description
In this part, an overview will be given of the companies that were willing to cooperate.
According to the CBS (2016), 39 percent of the insurance companies are working with big data. Only within the tourism sector there are more companies working with big data, namely 44 percent. However, as discussed in section 4.2, this sector was not willing to cooperate.
Recently, the Covenant of Insurers of the Netherlands, in cooperation with PwC (2017), published a report about innovation within the insurance sector. 54 percent of the insurers indicated that process and product innovations were their top priority to increase growth. Examples of process innovations were trying to make processes more standardised or simpler.
Another example was advanced analytics with big data (Covenant of Insurers, 2017).
As said before, all these companies are active in the insurance sector. However, this is the only similarity. They differ in, for example, products they offer, size, and countries they are active in. The four companies are Delta Lloyd, AEGON, Zilveren Kruis and Nationale
Nederlanden. These four companies are all in the top ten of biggest insurance companies in the Netherlands, with Zilveren Kruis at the first place. Together, these four companies represent 43 present of the total insurance market. In 2015, they had a total turnover together of almost 17 billion euro (Covenant of Insurers, 2015).
The abbreviated name of the company has been given next to the full name between brackets, because these will be used in the results part.
Delta Lloyd (DL)
Delta Lloyd is a company with 5.208 employees in 2015. It offers many different forms of insurances, ranging from health insurances to boat insurances. The company has about eight percent market share, making it the fifth largest insurance company in the Netherlands. Delta Lloyd also has a department where one can get a mortgage, meaning that Delta Lloyd is not only focusing on insurances. The company is also active in Belgium. Later this year, Delta Lloyd will become part of Nationale Nederlanden.
AEGON (AE)
In the Netherlands, there were 4.322 people working at AEGON in 2015. Worldwide, AEGON had 31.530 employees in 2015. The company offers insurances and pension products to more than 47 million people each year. AEGON is a company that is active in almost every part of the world. Examples are the United States of America, Brazil, China, and India.
Zilveren Kruis (ZK)
In 2015, 3.698 people were working at Zilveren Kruis. The company is only offering healthcare insurances, but has more than 3.3 million customers. This makes it the biggest insurance
company in the Netherlands, with respect to healthcare. At the moment, Zilveren Kruis is only active in the Netherlands.
Nationale Nederlanden (NN)
In 2015, Nationale Nederlanden had 6.493 employees in the Netherlands. Worldwide there were 11.461 employees working at Nationale Nederlanden in 2015. The company serves more than four million customers each year, making it one of the biggest companies in the Netherlands. In line with Delta Lloyd, it also offers a wide range of different insurances. Next to the insurance products, the company also offers products and services for banking, mortgages, pensions, and investments and deposits. Nationale Nederlanden is active in seventeen European countries and in Japan.
5. Analysis and results
This chapter will present the analysis and results of the gathered data. By giving these, it will become clear whether the propositions are supported or rejected. This will be done as follows. First, the three propositions will be analysed through the answers of the four data managers. Second, the three propositions will be analysed through the answers of the four data analysts. Third, an analysis will be made within the firm, meaning that the answers of the data manager and data analyst of the same firm will be compared. Fourth, an overall overview of the results will be given. But first, an overview will be given about how the results are analysed.
5.1 Analysis
The results will be analysed in the following way. Every subchapter will contain a table with the data managers, the data analysts, or both. Within this table, the initials of the interviewees will be given in the first column (unless they wanted to stay anonymous). The second column contains the abbreviated name of the firm. The third column contains whether he or she is in favour of the proposition or not. The fourth and last column contains the supporting argument why or why not.
To determine whether an interviewee agrees or disagrees with a proposition, the
interviews were coded with NVivo, after they were fully transcribed. These codes are created to find relevant sections of information for the propositions. As already mentioned before, one interview was also checked manually to check whether NVivo covers every relevant section and that the coding was justified. Found was that NVivo identified more sections that might contain relevant information. Therefore, the other interviews are only analysed with this software program. The first aggregate codes that were created to identify the main sections were data, big data, innovation, and process innovation.
means that the coding was also done in Dutch. However, supporting statements or comments will be translated literally into English and presented in the tables.
According to Shenton (2004), in order to allow others to follow and validate the process of analysing the results, an overview of the codes should be given. Therefore, the overview will now be given. With the use of NVivo, sub-codes were created to identify whether an interviewee agrees, disagrees, or did not give a relevant answer regarding the propositions.
For the first proposition, codes such as guidance, direction, stimulate, deviation, and distraction were created. The sentences that contain these words were checked to see if an interviewee agreed or disagreed with the proposition. It is also possible that none of the codes were found within the transcription of the interview, indicating that no relevant answer was given. Continuously, it is also possible that the interviewee was not clear about whether he agrees or disagrees. For the second and third proposition, the same procedure has been applied. To check the opinion of the interviewee about propositions two, codes such as quicker, increase, impact, slower, and decrease have been created. For proposition three, codes such as wider, more variety, range, smaller, and less variety have been created.
This paper will now continue with presenting the results of the data managers.
5.2 Data managers
In this part, the results of the interviews with the data managers are presented. The first proposition is that usage of big data leads to more focused process innovation. The results are presented in table 5.1 below.
Table 5.1: Data managers about proposition one
Manager Firm Agree with proposition I
Supporting argument
A [E.S.] DL Yes ‘We have an 80-20% rule here, meaning that 80% of the problems [within processes] can be solved by focusing on the 20% biggest problems … with using big data, we come much closer to this 80-20% rule’ and ‘you might have ten elements within a process that do not function optimal, but if you use big data, you can indicate the top three and that already solves a lot of the problems.’
B [M.W.] AE Yes ‘The more data you have the better and if you apply some machine learning techniques with it, you can learn things, thus also where you can innovate your process the best.’
C [R.G.] ZK Yes ‘We have several systems that give us relevant insights of our processes from big data. These insights can help us with improving the services for our customers because we now know better where [the process] can be innovated or optimised.’ D NN Yes ‘Big data-analysis is in our company mostly used to
see which process innovation should get priority, by checking how often the process is used and guessing
how much profit is gained with the proposed innovation’ and ‘you can exactly address that what is the best use of your time.’
Note: As manager D wanted to stay anonymous, no initials are given.
The results for the first proposition show that all four interviewed data managers agree with the first proposition. This means that they were confident that big data leads to more focused process innovation. Manager A states that through the use of big data more problems are solved within the processes than with normal data. He pointed out that big data could help to innovate the processes and therefore solving the problem forever, instead of creating a one-time solution. Manager B was also very clear that big data has a positive effect on knowing where processes can be innovated best. However, this cannot be achieved without using machine learning, meaning that big data on its own is considered to be insufficient. According to manager C, big data leads to better and new insights, at least in services for customers. He could not tell whether big data has the same effects within other parts of the company, because this was out of reach of his function. Manager D also agreed with the first proposition, however from a different perspective than, for example, manager A. Whereas manager A focuses more on using big data for solving problems, manager D focuses more on creating value for the company. Although the perspectives differ how big data can lead to more focused process innovation, the managers concluded overall that big data has a positive effect.
The second proposition states that big data leads to an increased speed of process innovation. In table 5.2 below the results are presented.
Table 5.2: Data managers about proposition two
Manager Firm Agree with proposition II
Supporting argument
A [E.S.] DL Yes ‘I think you become twenty to twenty-five percent quicker [with innovation] when using big data.’ B [M.W.] AE Yes ‘I think the best way that big data can help with
process innovations is by making decisions automatically instead of humans making these choices.’
C [R.G.] ZK Not clear ‘Big data has enormous potential [for innovation], but it can also use unnecessarily much time and money if a company does not have the right focus or the right capabilities. Realising impact is not always as simple as it seems.’
D NN No ‘The effect of using big data is that it does not make it faster or more frequent, I would rather say slower as those analyses are not easy…’
Note: As manager D wanted to stay anonymous, no initials are given.
The results for the second proposition are mixed. Only two out of the four data managers agree that big data makes process innovation quicker. However, only one out of the four managers disagrees with the proposition. Manager A indicated relatively precise what the effect of big data is on the speed of process innovation. He was sure that process innovation becomes twenty to twenty-five percent faster when using big data. Manager B is also supporting the second
statement. She highlighted that big data can make decisions within the process faster and
therefore, the total time it takes for a process innovation decreases. Manager C did not provide a clear answer. On the one hand, he indicated that the potential of big data is enormous. Although enormous potential can lead to faster process innovations, he did not explicitly state this. On the other hand, he also mentioned that a company needs the right focus and capabilities in order to fully capture the potential benefits of big data. He was not sure whether his company is already doing so. Manager D highlighted the complexity of big data. Because of this, the process innovation takes more time, rather than less time. However, manager D mentioned that big that also has positive effects, such as decreasing costs of innovation.
The third proposition states that big data leads to a wider variety of process innovations. The results are structured in the same way as the results from proposition one and two and are presented below in table 5.3.
Table 5.3: Data managers about proposition three
Manager Firm Agree with proposition III
Supporting argument
A [E.S.] DL No ‘We always focus on current problems within the processes… within these problems we try to innovate or enhance the process to solve it, but we never go [innovate] outside these problems. And I do not see more solution variety within the problems.’
B [M.W.] AE Yes ‘I think that we will constantly find new applications of processes [through big data].’
C [R.G.] ZK Yes ‘Because you are getting more and more big data, you can innovate on more positions. Also in parts of your products or processes where you have never thought about yet.’
D NN No relevant
answer
-
Note: As manager D wanted to stay anonymous, no initials are given.
From the table can be concluded that only two of the four data managers agree with the third proposition. Only manager A explicitly disagrees with the proposition. However, he disagrees because their work method does not allow finding new options of process innovation, rather than stating that big data has no effect on finding these new options. Manager B and C both agree with the third proposition. Manager B argues that big data has the potential of constantly finding new ways to innovate the processes. Manager C says that big data leads to more process innovations. Moreover, big data leads to unexpected innovation in new areas. Unfortunately, manager D did not say anything relevant related to the third proposition.
5.2.1 Conclusion
All data managers agreed on the first proposition. The arguments given to support this proposition were quite the same. Two managers stated that through big data they know best where to innovate the process. Another manager agrees on that, however, he is clearer by telling that he focuses more on the problems that arise from the processes. The last manager ranks the possibilities of process innovations through big data and determines which is best and focuses thereon.
Unlike the first proposition, only two managers agreed with the second proposition, stating that big data increases the speed of process innovation. The first manager was really clear by telling that through big data the process innovations become twenty to twenty-five percent faster. The second manager supports the proposition by telling that big data makes decisions faster than humans, therefore decreasing the total time of process innovations. Due to the
complexity of big data, according to another respondent, big data slows process innovation down. Two managers also support the third proposition. Both indicate that big data leads to newly discovered process innovations. Another manager disagrees with the proposition. However, he only uses big data for analysing and solving problems that occur. Therefore, it might be the case that this manager would agree with the other managers if the company uses big data not only for analysing problems but also in a more open way.
5.3 Data analysts
In this part, the results of the interviews with the data analysts are presented. This will be done in the same way as has been done with the data managers. The first proposition is that usage of big data leads to more focused process innovation. The results are presented in table 5.4 below.
Table 5.4: Data analysts about proposition one
Analyst Firm Agree with proposition I
Supporting argument
A [F.G.] DL Yes ‘I think there are chances, especially that you can apply it [big data] more effectively. I mean, that you know better where you can innovate your process
and therefore you can use [the process] it better.’ B
[R.v.G.]
AE Yes ‘There are processes that operate differently than they should and with big data you can identify these. Therefore, you can just target on these because of that big data.’
C ZK Not clear ‘I do not know for certain if the focus would
increase. I think that it does not make any difference with normal data. However, maybe I am wrong, but at this moment we are not that far with big data.’ D NN Yes ‘There are so many processes within this company;
the employees can never check them all. You really need data for it. Therefore, I think that data helps you to find where you can innovate your processes at best.’
Note: As both analysts C and D wanted to stay anonymous, no initials are given.
From table 5.4 can be seen that data analysts A, B, and D support the first proposition. Only analyst C was not clear about supporting the proposition. Both analysts A and D state that through big data, one can know better where to innovate the process, which will result in better processes. Analyst D adds that big data will replace, or at least will complement, the employees in identifying where can be innovated best. Analyst B argues that one can identify the flaws within a process due to big data. Only analyst C did not give a clear answer. He argues that at this moment normal data is enough. However, there might be potential for big data, but this has not been discovered yet. Therefore, as there might be potential, the proposition was not rejected.
