&
Organization Size
Is the level of digital innovation in an organization related to the size of the organization?
Summary: This thesis empirically investigates the effect of the size of the organization on the level
of digital innovation within organizations. We argue that small organizations have a higher level of digital innovation than large organizations. Large organizations may have a hard time in keeping up with the speed and complexity of digital innovation due to lack of visionary leadership, decision making speed and skills. We investigate our hypothesis through a questionnaire that is disseminated among organizations in the Netherlands via one Big Four firm. We offer support for our hypothesis with three observations. First, small organizations report a higher score on the diagnostic tool included in our questionnaire compared to large organizations. Second, this score is mainly caused by the higher score on digital evolution scanning and skills and in a weaker sense improvisation for the small organizations. Third, the previous results are robust to including industry differences.
MSc Thesis track: Managerial Economics & Strategy
ECT: 15
Written by: Nicole Rhebergen
Student number: 11122803
2 Statement of Originality
This document is written by Student Nicole Rhebergen 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
3 Table of Contents
1. Introduction……… 4
2. Theoretical Background………. 6
2.1. Innovation and Organization Size………... 6
2.2. Digitization and Organization Size……….10
2.3. Digital Innovation and Organization Size………... 12
3. Methodology……….. 16 3.1. Research Setting………..16 3.2. Population………... 18 3.3. Data Collection………... 19 3.4. Outcome Variables………..20 3.5. Independent Variables……….22 3.6. Statistical Analysis……….. 24 4. Results……… 25 4.1. Sample……….25
4.2. Background Characteristics Small and Large Organizations………..26
4.3. Differences in the Level of Digital Innovation………... 29
4.4. Determinants of Digital Innovation……… 31
4.5. Robustness Checks………. 34 4.6. Limitations……….. 35 5. Conclusions……… 36 6. Related Literature………...39 7. Appendix A……… 43 8. Appendix B……… 48 9. Appendix C……… 57 10. Appendix D……… 63
4 1. Introduction
Extensive research has been done in the field of innovation and the size of the organization. The traditional Schumpeterian theory argued that small organizations are the driving force of innovation (Schumpeter, 1934). Later, the Schumpeterian theory moved away from the argument of creative destruction by small organizations and moved towards creative accumulation of large organizations (Schumpeter, 1942). This was summarized by Schumpeter (1942) as the following:
As soon as we go into the details and inquire into the individual items in which progress was most conspicuous, the trail leads not to the doors of those firms that work under conditions of comparatively free competition but precisely to the doors of large concerns… and a shocking suspicion draws upon us that big business may have had more to do with creating that standard of living than with keeping it down. (p. 82).
Many empirical studies supported the later Schumpeterian thinking that small organizations are less likely to innovate than large organizations.
One would expect that the positive relationship between innovation and the size of the organization also holds for digital innovation and the size of the organization. However, we have seen how organizations struggle with implementing digital innovation processes in a successful way which is leading to severe consequences sometimes (Lucas & Goh, 2009). Many large organizations have vanished like Kodak or Blockbusters, the organization lifespan is decreasing and small and/or relatively young organizations, like Uber or Amazon, seem to have an advantage when it comes to digital innovations.
Nowadays, organizations are operating in a world that has to cope with digital technologies. Digital technologies are increasingly embedded in products and services. For example, there are cars, watches and houses that include software based digital capabilities now. This changes the nature of product and service innovations since digital technology generates complex and new innovation
5 challenges. One important part of the complexity is the speed with which digital innovations follow up each other. Recent research has also shown that the unique aspects of digital technology result in new types of innovation that are distinctly different from non-digital innovation (Henfridsson
et al., 2014; Yoo et al., 2012). To summarize, organizations increasingly have to deal with digital
innovation.
This raises our research question: “Is the level of digital innovation of an organization related to
the size of the organization?”. The literature has studied the relationship between innovation and
the size of the organization extensively but these articles are dated from many years back and fail to give an explanation for the observation that large organizations increasingly cannot cope with this new type of innovation, namely digital innovation. To test our research question, we will use a diagnostic tool, first presented by Nylén and Holmström (2015). This diagnostic tool is developed to give insight into the level of digital innovation of organizations.
This thesis contributes to the existing literature in several ways. First, we test the diagnostic tool of Nylén and Holmström (2015) in practice by disseminating a questionnaire via one Big Four firm and show the predictive strength of the tool. We refer to a Big Four organization as one out of the four largest accounting organizations worldwide that offer accounting services to all kinds of organizations. The organizations that form the Big Four are Deloitte, Ernst & Young (EY), KPMG and Pricewaterhousecoopers (PwC). We refer to a Big Four organization and not to the exact name of the organization in order to maintain the privacy of their clients and their employees. Second, we examine the correlation between digital innovation and the size of the organization and we suggest why the relationship between digital innovation and the size of the organization could be negative, even though the relationship between non-digital innovation and the size of the organization is positive. Third, we give insight in the differences between large and small organizations when it comes to the resources and capabilities that are needed to digitally innovate successfully. Fourth, we report on the difference in the level of digital innovation for small and large organizations by collecting information directly from the people that are involved in the digital agenda. As such, this supports the trustworthiness of our results.
6 We find support for a negative relationship between the level of digital innovation and the size of the organization. That is, large organizations have a lower level of digital innovation than small organizations. When looking at the resources and capabilities of the organizations, the observed result is mainly explained by the high level of digital evolution scanning and skills and in a weaker sense improvisation in the small organizations.
The thesis continues as follows. Section 2 presents the theoretical background that is used for this thesis. Section 3 describes the methodology that has been used, whereas the results of the thesis can be found in Section 4. Section 5 concludes and presents some limitations and directions for further research.
2. Theoretical Background
Research on innovation dates back to the 1900s, in which it was already argued that product and process renewal was essential for economic and technological change (Baregheh et al., 2009). However, there was no clear term for this product and process renewal and there was much often talked about invention instead of innovation. The differentiation of invention from innovation occurred when Schumpeter (1939) stated that the term invention is referred to as an act of intellectual creativity while the term innovation is referred to as organizations figuring out how to transform inventions into changes in the business model of the organization (Godin, 2008). Based on this, in this thesis we will refer to innovation as a change in the business model of the
organization by the development of new, unique concepts such as 1) improving the product or service value and 2) how products and services are delivered to customers.
2.1. Innovation and organization size
Extensive research has been done in the field of innovation activity and organization size. The relationship between innovation activity and organization size was first investigated by
7 Schumpeter (1934). In his early work, Schumpeter argued that small organizations are the driving force behind innovation indicating a negative relationship between innovation activity and organization size (Mark I). However, in his later work, Schumpeter moved away from his argument that small organizations were the ones causing creative destruction. Schumpeter (1942) argued that large organizations are the ones causing creative accumulation indicating that large organizations are the driving force behind innovation (Mark II). Thus, the final conclusion by Schumpeter (1942) was that large organizations have a higher innovation activity than small organizations. Nowadays, Mark I and Mark II are seen as complementary, resulting in a non-linear inverted U-shaped relation between innovation activity and organization size (Aghion et al., 2005). However, are small and large organizations also equally good in exploiting innovations? In the famous article written by Teece (1986), he elaborates on the question why small entrepreneurial organizations, that often have developed interesting ideas, may fail in exploiting these ideas and why large organizations may succeed in exploiting ideas even though they might have less innovative ideas than small organizations. The reason for this, according to Teece (1986), is that large organizations possess the right resources and capabilities for exploiting a new idea which can turn an idea into a successful introduction of the product or service into an existing market or a new market. In the literature, we could identify four reasons why large organizations may be better in exploiting innovations by focusing on the resources and capabilities of organizations.
2.1.1. Fixed costs of R&D
When innovating a product, service or process, the organization has to invest a substantial amount of money in research and development (R&D). There are many empirical studies that support the hypothesis that the amount of money invested in R&D is a necessary condition for the innovation activity of the organization. Among others, Frenkel et al. (2001) hypothesizes that the amount of money invested in R&D can be taken as a proxy for the innovative activity of an organization and this hypothesis has been proven to be true. Next to R&D expenditures being used as a proxy for the innovation activity of an organization, Acs and Audretsch (1988) found that the number of innovations is positively related to the amount of money invested in R&D. Finally, when organization size is taken into account, it can be concluded that organization size has a positive
8 effect on the amount of money invested in R&D (Acs & Audretsch, 1988; Cohen et al., 1987). Therefore, organization size could have a positive effect on the innovation activity. The reason for this is that investing in R&D requires high fixed costs. Large organizations are better able to cover these costs because large organizations have a higher expected output, and thus income, to cover these costs compared to small organizations (Cohen, 1995). To conclude, R&D is needed in order to innovate and large organizations have a higher amount of money at their disposal to invest in R&D, which suggests that large organizations are in a better position to innovate than small organizations.
2.1.2. Innovative input versus innovative output
Next to the fact that organizations have to invest a substantial amount of money in R&D when innovating a product, service or process, the innovative output of the invested money is at least as important. There are several studies that have focused on the relationship between the amount of money invested in R&D, called the innovative input, the innovative output and organization size. These studies concluded that large organizations have an advantage over small organizations when it comes to the ratio of innovative input relative to the innovative output. Among others, Acs and Audretsch (1988) found that the number of innovations is positively related to the expenditure on R&D and that the number of innovations increase when the money invested in R&D by the industry increases. However, this is at a decreasing rate. In the first section, we concluded that large organizations are, in general, better able to cover the large fixed costs of R&D. Therefore, large organizations have, in general, a higher innovative input compared to small organizations. Since there is a strong relationship between innovative input and innovative output, the higher the innovative input, the higher the innovative output of an organization. It follows that large organizations will have a higher innovative output than small organizations because these organizations, in general, have a higher innovative input. Thus, there exists a positive relationship between innovative input and innovative output. Furthermore, large organizations have a higher level of money at their disposal that they can invest in R&D (innovative input) than small organizations which, all taken together, shows that large organizations have a higher innovative output and are thus better able at exploiting innovations compared to small organizations.
9
2.1.3. Diversification
So far, we have looked at what amount goes in an organization and what comes out of the organization with regard to innovation, but we have not taken a look at the characteristics of the organization itself and that is where diversification plays a role. In this case we refer to diversification as an organization entering multiple industries (industrial diversification) or markets (global diversification) besides the core business model (Denis et al., 2000). Following Aron (1988) and Lang and Stulz (1997), there exists a positive relationship between diversification and organization size. Thus, the larger the organization, the more industries in which the organization operates and the more product lines the organization has. The reason for this is that large organizations have the resources available to diversify into different markets. Furthermore, diversification is positively related to innovation which means that the more industries and markets the organization is active in, the more innovations will come from the hand of this organization. Cardinal and Opler (1994) found that industrially diversified organizations are more efficient innovators than other organizations. This is mainly because these organizations minimize incentive problems in R&D. This might suggest that large organizations have a substantial amount of valuable information at their disposal. It can place them in a position that they are better able to exploit the new knowledge leading to surprising and new innovations. Next to that, Hitt et al. (1997) focus on the relationship between global diversification and innovation and explain three reasons for a positive relationship between them. First, globally diversified organizations have an incentive to invest in developing capabilities needed for a higher innovative input. This incentive exists because it leads to a higher innovative output and, ultimately, a better competitive advantage for the organization. Second, global diversification can generate resources that are necessary for R&D intense organizations. Third, global diversification can generate resources that are necessary for building innovation capabilities. To conclude, large organizations are more diversified than small organizations and more diversified organizations are better able to generate innovations which, all taken together, suggests that large organizations are better able to generate innovations than small organizations.
10
2.1.4. Market concentration
After we have taken a look at the organization itself, we will move to the environment in which the organization operates. Schumpeter (1942) hypothesizes that there exists a positive relationship between market concentration and the amount of money invested in R&D. He argued that organizations operating in a concentrated market have more money at their disposal to invest in R&D because of the higher profits of these organizations. Among others Doukas and Switzer (1992) find support for the Schumpeterian hypothesis and state that investors see an organization’s R&D announcement as a positive sign, when the organization is operating in a concentrated market. On the contrary, investors see an organization’s R&D announcement as a negative sign when the organization is not operating in a concentrated market. Furthermore, Acs and Audretsch (1987) found that large organizations tend to operate in markets that are more concentrated compared to small organizations which might suggest that large organizations have a higher amount of money available to invest in R&D and hence, have a higher innovative output. To support this, Cohen and Klepper (1996) found that large organizations have a higher portion of their own profits they can invest in R&D and Becchetti and Trovato (2002) show that large organizations are in a better position to make use of external finance. Rogers (2004) supported the relationship between R&D intensity and organization size by using a regression analysis and found that the R&D intensity increases as the organization size increases. To summarize, there exists a positive relationship between market concentration and the amount of money invested in R&D and there also exists a positive relationship between market concentration and organization size. Therefore, organizations operating in a concentrated market, which are large organization, have an advantage over organizations operating in a less concentrated market when it comes to innovative input and hence, innovative output.
2.2. Digitization and organization size
Much research on innovation in a broad sense and on the relationship between innovation activity and organization size in particular, dates back to before the 2000s. From the literature it becomes clear that the term innovation, as we described it in this thesis, evolves over time. After the 2000s the focus of the literature shifts toward the use of Internet to innovate an organization, called
11 digitization or e-commerce. Slywotzky and Morrison (2000) refer to digitization as “a disruptive, creative force that is revolutionizing how people work, play, communicate, buy, sell, and live” (p. 7). Building on this, in this thesis we refer to digitization as a change in the business model of the
organization by the development of new, unique concepts such as 1) improving the product or service value and 2) how products and services are delivered to customers by the use of Internet.
During the mid-1990s, the focus was on the effect of the Internet on communication, marketing and information technology and later on, the focus shifted towards the effect of the Internet on the industry structure, profit distribution and transaction costs (BarNir et al., 2003).
However from 2000 onwards, empirical research around the topic of digitization was conducted so that the implications of the Internet for the competitive positioning of organizations could become clear. In this thesis, we focus especially on the consequences of the introduction of the Internet on the competitive positioning of small and large organizations. Kowtha and Choon (2001) hypothesize that a positive relationship exists between organization size and the organization’s strategic commitment to e-commerce because of the lack of resources and capabilities in the first stages. This hypothesis has proven to be true. Furthermore, Zhu et al. (2003) investigated facilitators of e-business adoption and concluded that large organizations have an advantage when it comes to the adoption of e-business compared to small organizations. However, small organizations seem to be better off in countries with a high e-business intensity compared to countries with a low e-business intensity because of network effects. BarNir et al. (2003) investigated the topic further and argued that the relationship between digitization and organization size should be negative because organizational change would be harder for large organizations due to routines and controls. However, BarNir et al. (2003) found the opposite results and concluded that large organizations seem to have a higher level of digitization because they have the money available to invest in digitization and they have a higher pool of knowledgeable employees that have the skills to incorporate digitization into the organization.
12 Altogether, this strengthens the evidence that large organizations seem to have more resources and capabilities available to exploit ideas and transform it into successful product or service introductions for existing markets or new markets.
2.3. Digital innovation and organization size
Around 2010, the focus of the literature switched from the use of the Internet to innovate an organization’s business model to the incorporation of digital tools and/or technologies into products and services to innovate an organization’s business model, which is called digital innovation. Based on this, in this thesis we will refer to digital innovation as a change in the
business model of the organization that is accomplished by incorporating digital components (digital tools or digital technologies) into products and services that are offered to its customers.
It is important to stress that in this thesis, we focus on digital innovation of the organization from inside out, namely to the customers. We do not focus on the digital innovation within the organization, namely digital innovation with regard to the organization’s processes.
One might expect that the positive relationship between innovation and organization size and between digitization and organization size also holds for digital innovation and organization size. The reason for this is that large organizations are still expected to be better able to cover the fixed costs of R&D, to have a higher amount of money available to invest in R&D and to be better able to exploit innovations. However, this raises the question why large organizations seem to have a hard time when it comes to digital innovation, like Blockbusters or Kodak, and why small organizations seem to succeed at the cost of large organizations while the resources and capabilities of organizations have not changed. Or have they? Put differently, are large organization still in a better position to exploit (digital) innovations?
The relationship between innovation and organization size has been studied extensively and has led to important insights with regard to the strengths and weaknesses and the resources and capabilities of different organizations. Furthermore, in the literature light was shed on the relationship between digitization and organization size. However, the world in which organizations
13 are operating has changed significantly over the last couple of years and the insights from previous literature may not be fully relevant anymore. Digital technologies are increasingly integrated into products and services leading to new innovation challenges, referred to as digital innovation. The resources and capabilities of established, and often large, organizations can stand in the way of innovating (Christensen, 1997) and we have seen how organizations struggle with exploiting digital innovations in a successful way which is leading to severe consequences sometimes (Lucas & Goh, 2009). A shortcoming of the previous literature is that it fails to address the reason for the fact that large organizations increasingly have trouble in exploiting digital innovation challenges. This calls for a new analysis of the resources and capabilities of different organizations (with respect to their size) necessary to exploit digital innovation.
Digital innovation goes further than innovating the products or services of the organization by the use of the Internet, as was the case with digitization. Digital innovation and the accompanied digital technologies go beyond the internal boundaries of the organization and will be integrated into the products and services that the organization offers to its customers (Lucas & Goh, 2009). This leads to new innovation challenges because digital technologies are characterized by speed, complexity and they are often hard to predict (Henfridsson et al., 2014; Yoo et al., 2012; Yoo et
al., 2010). Furthermore, it becomes clear that digital technologies, and the unique aspects, result
in new types of innovation compared to innovation as we have discussed in earlier chapters of this thesis (Henfridsson et al., 2014; Yoo et al., 2012). So, there seems to be an empirical and a theoretical trend. On the empirical side, large organizations are increasingly struggling with new types of innovation. On the theoretical side, the literature notes the fact that digital innovation is different from non-digital innovation. However, there is no literature present that combines these trends and analyses the relationship between digital innovation and organization size. This suggests that there is a need to fill this gap in the literature that could possibly explain the observations regarding digital innovation and organization size.
Why might large organizations fail to exploit digital innovations and why might small organizations have an advantage because of this? The literature offers three explanations.
14 First of all, the lack of visionary leadership might play a role in explaining why large organizations fail in exploiting digital innovations. Tellis (2006) wonders why some large organizations succeed in lasting a long period of time and why other large organizations self-destruct themselves. He argues that the latter is due to the lack of vision regarding change. Tellis (2006) argues that established organizations (who are most of the time large in their size) who lack vision are characterized by being happy with past successes and current customers, products, and services are the focus point of op this type of organization. Because these organizations focus strongly on their existing business model, they do not want to cannibalize this by shifting the focus to future products, services, customers or markets. What happens with an organization that does not focus on the future, is that they can easily miss an important innovation. Especially when this innovation is characterized by speed and complexity like digital innovation. On the other hand, small organizations do not have a solid business model yet and are very flexible in adapting to changes due to, among other things, their size. Another point is that small organizations have to focus on surviving as a business so they have to develop themselves quickly. Therefore, small organizations are more future oriented. This results in the fact that small organizations are better able to scan digital innovations and exploit them in the end.
Second, the speed of decision making in large organizations can take longer compared to small organizations. Indeed, Baum and Wally (2003) found a negative and significant relationship between organization size and the speed of decision making. In their research, the CEO’s who reported that decision making in their organization was slow, were characterized by slow or stable growth and profitability. As stated in their research, these are the characteristics of large and established organizations. The reason for this is that large organizations consist of many layers and most of the time have a formal way of taking decisions. Decisions have to be approved by a number of parties. To the contrary, small organizations are less relying on formal rules and can communicate with their employees directly because of their size. Fast decision making is essential in adopting to changes that are characterized by speed and complexity in order to stay competitive and active in the market.
Third, large organizations may have a hard time in making sure that their employees are prepared for digital innovation and have the appropriate skillset to deal with this type of innovation. Every
15 wave of innovation brought new jobs as well as it eliminated some existing jobs. Organizations therefore had to make sure that they attracted new skilled employees and teach the rest of the employees the right skills. However, digital innovation leads to continuous new improvements that are following up each other faster than the organizations ever dealt with. This leads to the fact organizations have to try even harder to make sure that the skills of their employees are up to date (Brynjolsson & McAfee, 2012). Large organizations may have a harder time when it comes to this since they are bound to tight rules and recruitment programs and their workforce has clear task descriptions whereas small organizations are flexible and less bound to task descriptions.
To test whether the relationship between digital innovation and organization size is indeed negative, we make use of the diagnostic tool that is presented by Nylén and Holmström (2015). In their article, they presented a diagnostic tool that is supposed to test the level of digital innovation within the organization. Furthermore, the diagnostic tool consists of five components that each measure the level of digital innovation in a specific area of business, namely: user experience, value proposition, digital evolution scanning, skills and improvisation. However, the diagnostic tool has not been used in practice before in the Netherlands and to our knowledge, we will be the first ones to test this tool in practice and use the tool to test whether there exists a negative relationship between the level of digital innovation and the size of the organization.
Insight into this relationship can potentially explain what we observe in practice, namely that large organizations are increasingly struggling with digital innovation compared to small organizations. Moreover, by making use of the diagnostic tool we can observe what might be the underlying reason for the potential difference in the level of digital innovation between small and large organizations because the tool is broken down into five important areas of business. We can thus observe in which areas of business the small and large organizations differ.
16 3. Methodology & Data
3.1. Research setting
This thesis will follow the positivist approach since data to answer hypotheses will be gathered using quantitative sources and is not focused on the feelings of the participants. Furthermore, this thesis will follow a deductive approach which means that we will base our hypotheses on the existing theory and develop a method to test the hypotheses in order to investigate whether our findings support the existing literature or speak against it.
The possibility to collect the primary data for this thesis depended on access to the relevant parties within different organizations. Via the Customer Relationship Management (CRM) system of a Big Four organization and via the personal contacts of employees within this Big Four organization, we gained access to the relevant positions and accompanied people within the different organizations.
The strategy to collect the data was through a quantitative questionnaire which was disseminated among multiple organizations, and the corresponding responsible employee, where the Big Four office is doing business with. The participants filled out the questionnaire in their work environment and could decide themselves whether they wanted to take part in it or not. The main part of the questionnaire consisted of ordinal questions based on a Likert-scale. Next to that, nominal questions were included to get more information on the background of the respondent and of the organization the respondent is currently employed.
The questions regarding the digital innovation activity of the organization are based on Nylén and Holmström (2015). They presented a diagnostic tool that should be able to measure the digital innovation activity, or the level of digital innovation, of an organization. This thesis has two objectives. The first objective is to perform a pilot study to test whether our questionnaire, including the diagnostic tool, works in practice. The other objective is to test the questionnaire, including the diagnostic tool, in practice to measure the digital innovation activity of organizations
17 that are different in size in order to test our research question. Next to that, the data collection regarding the respondent’s and organization’s characteristics is based on several classifications that have previously been used in practice to make sure that the independent variables are not suffering from lack of precision.
Through the use of the questionnaire, the following conceptual framework and accompanied hypotheses will be tested:
In the literature, a positive relationship was found between innovation and organization size and between digitization and organization size. However, the relationship between digital innovation and organization size has not been tested before. On the contrary, we have seen how large organizations increasingly fail to keep up with the speed and complexity of digital technologies and that small organizations increasingly have an advantage when it comes to exploiting digital innovations (Lucas & Goh, 2009). Therefore, one might expect that the opposite relationship will be found for digital innovation and organization size, which leads to the following hypothesis.
Figure I
18
Hypothesis 1
— Small organizations have a higher digital innovation activity than large organizations.
Next to that, it is important to get insight in the individual parts of which the diagnostic tool that is testing the digital innovation activity is build up. The diagnostic tool takes five important resources and capabilities needed for digital innovation into account. When a breakdown into the five resources and capabilities is made, we can make statements about whether large or small organizations excel at certain points. Therefore, the following hypotheses are based on the five resources and capabilities that will together measure the level of digital innovation of the organization.
Hypothesis 2
— Small organizations have a higher score on user experience than large organizations.
Hypothesis 3
— Small organizations have a higher score on value proposition than large organizations.
Hypothesis 4
— Small organizations have a higher score on digital evolution scanning than large organizations.
Hypothesis 5
— Small organizations have a higher score on skills than large organizations.
Hypothesis 6
— Small organizations have a higher score on improvisation than large organizations.
3.2. Population
A questionnaire will analyze a sample of individual responses from an entire population. The questionnaire will be spread among clients with which the Big Four organization is doing business.
19 Moreover, the questionnaire will be spread among the Dutch clients with which the Big Four organization is doing business. We refer to a Big Four organization and not to the exact name of the organization in order to maintain the privacy of their clients and their employees. The reason for this is that the Big Four organization serves all kinds of clients varying in industry, size and age. Next to that, access was given to their clients in the Netherlands solely for the purpose of this thesis. Therefore, the population where we will focus on in this thesis are the clients of the Big Four organization that have an office in the Netherlands with their own board of directors.
3.3. Data collection
A desirable result would be obtained by spreading the questionnaire among all the clients of the Big Four organization and obtain results from all of them. However, due to time constraints and the lack of other necessary resources, a random sample will be drawn from the entire organization. The random sample drew random organizations varying in industry, size and age from the entire population to make sure that every type of organization was equally represented.
The data will be collected by emailing the questionnaire to the random sample of the clients of the Big Four organization. There has been chosen for an online questionnaire because time constraints will be eliminated. By the use of an online questionnaire, we have access to numerous organizations in a short amount of time without the constraint of being geographically far apart from each other (Wright, 2005). The questionnaire was spread among 800 organizations out of which 72 organizations responded, resulting in a response rate of 9.00%.
3.3.1. Pilot to test the diagnostic tool
A pilot of the questionnaire will be the first step because we first have to test whether the diagnostic tool will prove to work in practice. The data that serves as input for performing a pilot questionnaire is obtained from two contradictory industries with regard to their innovation activity. The innovation activity of an organization will be used as a proxy for the digital innovation activity because no prior research is available that tested the differences in digital innovation activity among industries.
20 Two contradictory industries that will be used to test the survey are the agricultural industry, which is often characterized by low innovation activity, and the technology industry, which is often characterized by high innovation activity. Likewise, we expect that the digital innovation activity in the technology industry is relatively high compared to other industries. In short, we expect that the digital innovation activity in the agricultural industry is lower than the digital innovation activity in the technology industry.
The pilot has shown that the same results were obtained as the results that were predicted according to theory. Thus, the organization operating in the technology industry reports a higher digital innovation activity than the organization operating in the agricultural industry. Next to that, no other remarks came forward about the questionnaire.
3.4. Outcome variables
The dependent variable or outcome variable will be the digital innovation activity of the organization, referred to as the 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦. This variable is a sum of five other variables that, when added together, represent the total amount of digital innovation activity in an organization. There are five composite variables that will together measure 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦, namely user experience (𝑈𝑠𝑒𝑟 𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒), value proposition (𝑉𝑎𝑙𝑢𝑒 𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛), digital evolution scanning (𝐷𝑖𝑔 𝐸𝑣 𝑆𝑐𝑎𝑛), skills (𝑆𝑘𝑖𝑙𝑙𝑠) and improvisation (𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛). The composite variables are created by combining multiple individual variables, the indicators, into one variable. The previous is consistent with Nylén and Holmström (2015). However, in this thesis, all the indicators are measured using a 4-point Likert-scale which ranges from ‘totally don’t agree’ (1 point) to ‘totally agree’ (4 points). The more the organization agrees on the question, the higher the organization scores on the indicator and the higher the digital innovation activity. The fifth category that is added to the Likert-scale questions is called ‘does not apply’ since we have to account for the fact that some questions do not apply to certain organizations. When the question is answered with ‘does not apply’, the organization scores zero on digital innovation activity for that indicator. Furthermore, we have split a few indicators into two questions because in the original diagnostic tool, the two questions were asked
21 in one question which can lead to confusion among respondents and ultimately less trustworthy answers. Below, the dependent variable, the composite variables, the indicators and the links between them are stated. Furthermore, an explanation for the dependent variable, the composite variables and the indicators can be found in Appendix A.
Figure II
DEPENDENT VARIABLE, COMPOSITE VARIABLE, INDICATORS AND LINKS
Question Indicator Organization's response Organization's score Composite variable Dependent variable score
Agree Average of 3 and 2
Don't agree 2.5
Does not apply 0 5.5
Agree 3
Totally agree Average of 4 and 4
Totally agree 4
Agree Average of 3 and 4
Totally agree 3.5 9.5
Don't agree 2
Totally don't agree 1
Agree 3 Agree 3 7 Totally agree 4 Don't agree 2 Agree 3 Totally agree 4 Agree 3 Agree 3 𝑈𝑠𝑎 𝑖𝑙𝑖𝑡𝑦 𝑈𝑠𝑎 𝑖𝑙𝑖𝑡𝑦 𝐴𝑒𝑠𝑡 𝑒𝑡𝑖𝑐𝑠 𝐸𝑛𝑔𝑎𝑔𝑒𝑚𝑒𝑛𝑡 𝑆𝑒𝑔𝑚𝑒𝑛𝑡𝑎𝑡𝑖𝑜𝑛 𝑆𝑒𝑔𝑚𝑒𝑛𝑡𝑎𝑡𝑖𝑜𝑛 𝑢𝑛 𝑙𝑖𝑛𝑔 𝑢𝑛 𝑙𝑖𝑛𝑔 𝑜𝑚𝑚𝑖𝑠𝑠𝑖𝑜𝑛𝑠 𝐷𝑒𝑣𝑖𝑐𝑒𝑠 𝑎𝑛𝑛𝑒𝑙𝑠 𝑒 𝑎𝑣𝑖𝑜𝑟𝑠 𝑒𝑎𝑟𝑛𝑖𝑛𝑔 𝑜𝑙𝑒𝑠 𝑒𝑎𝑚𝑠 𝑆𝑝𝑎𝑐𝑒 𝑖𝑚𝑒 𝑜𝑜𝑟 𝑖𝑛𝑎𝑡𝑖𝑜𝑛 𝑈𝑠𝑎 𝑖𝑙𝑖𝑡𝑦 𝑆𝑒𝑔𝑚𝑒𝑛𝑡𝑎𝑡𝑖𝑜𝑛 𝑢𝑛 𝑙𝑖𝑛𝑔 𝐴𝑒𝑠𝑡 𝑒𝑡𝑖𝑐𝑠 𝐸𝑛𝑔𝑎𝑔𝑒𝑚𝑒𝑛𝑡 𝑜𝑚𝑚𝑖𝑠𝑠𝑖𝑜𝑛𝑠 𝐷𝑒𝑣𝑖𝑐𝑒𝑠 𝑎𝑛𝑛𝑒𝑙𝑠 𝑒 𝑎𝑣𝑖𝑜𝑟𝑠 𝑒𝑎𝑟𝑛𝑖𝑛𝑔 𝑜𝑙𝑒𝑠 𝑒𝑎𝑚𝑠 𝑆𝑝𝑎𝑐𝑒 𝑖𝑚𝑒 𝑜𝑜𝑟 𝑖𝑛𝑎𝑡𝑖𝑜𝑛 𝑈𝑠𝑒𝑟 𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒 𝑉𝑎𝑙𝑢𝑒 𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐸𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑆𝑐𝑎𝑛𝑛𝑖𝑛𝑔 𝑆𝑘𝑖𝑙𝑙𝑠 9 𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛 10 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦 41 Table I
22 To illustrate how the score of an organization on digital innovation activity is established, an example is presented above (Table I). The score on the dependent variable can take a value between zero and 60. The score on each composite variables can take a value between 0 and twelve.
The composite variables are being measured by asking three to five questions per composite variable in the questionnaire. An overview of the questions belonging to the composite variables is presented in the questionnaire which can be found in Appendix B.
3.5. Independent variables
The variable of interest in our analysis will be the size of the organization. During our analysis, this variable will be of main influence when looking at the effect on the digital innovation activity of an organization. The number of employees will be used as a proxy to for the size of the organization (𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑆𝑖𝑧𝑒). To measure the organization size, we implemented a question in the questionnaire with regard to how many employees are working at the organization on a permanent basis in the Netherlands. This variable consists of seven size categories. However, for the Independent T-test in our analysis the categories will be reduced to two categories, namely small and large organizations so that we can make a reliable comparison between sizes when taking our sample size into account. The classification that is used for reducing the size categories to smaller categories is the Dutch Chamber of Commerce size classification. An overview of this classification and how we reduced the size categories following this classification can be found in Appendix A.
A number of other independent variables will be included in the questionnaire as well. . First, we paid attention to two background characteristics of the organization that will be used as control variables. Second, a number of variables are included that will measure to what extent the organization finds digital innovation important. Finally, a number of variables will be included with regard to the background characteristics of the respondent.
23
3.5.1. Control variables
It is important that the following two characteristics of the organization will also be taken into account, namely the years of existence of the organization (𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝐴𝑔𝑒) and the type of industry in which the organization is operating (𝐼𝑛 𝑢𝑠𝑡𝑟𝑦). The organization age and the industry are two very important control variables that have to be taken into account when testing the relationship between digital innovation activity and organization size.
3.5.1.1. Organization age effects
The relationship between digital innovation activity and organization size might be affected by the organization age. Most of the time, large organizations tend to be older in age. Similarly, small organizations tend to be younger in age. This means that organization size and organization age are closely related. However, there might be small organizations that are old in age and large organizations that are young in age. When there is not controlled for the organization age, the result might be that the relationship between digital innovation activity and organization size will be disturbed by age effects resulting in a bad representation of the true relationship. Therefore, by not taking the organization age as a control variable, the effect of it on the relationship between digital innovation activity and organization size will be eliminated. However, there exists a strong correlation between organization size and organization age in our analysis. Furthermore, in our sample, there are only small organizations that are relatively young and large organizations tend to vary in the age of the organization but most of them are relatively old. Therefore, for the rest of our analysis we will eliminate the organization age.
3.5.1.2. Industry effects
In different industries, there exists a different level of innovation and a different potential to innovate. Similarly, there might exist a different level of digital innovation in different industries. For example, within a technology industry there might be more digital innovation compared to an agricultural industry mainly because the technology company is built around creating advanced, most of the time digital, systems. An agricultural industry is more focused on creating offline systems naturally. However, we should keep in mind that the agricultural industry is also evolving
24 quickly. When all industries are taken together and one industry is represented by more organizations than the other, the effect of industry might disturb the true relationship between digital innovation activity and organization size. Therefore, it is very important to take the industry into account and use it as a control variable in our analysis. In our questionnaire, 17 industry categories are depicted. However, these are a lot of categories which reduces the probability of a good analysis. The 17 industries are first reduced to seven, and finally to four industries by the use of the Standard Industry Classification (SIC). An overview of this classification and how we reduced the categories following this classification can be found in Appendix A.
3.5.2. Background characteristics of the respondent
A number of variables will be included with regard to the background characteristics of the respondent. The variables measure the percentage of females in the organization (%𝐹𝑒𝑚𝑎𝑙𝑒), the job-level (𝐽𝑜 𝑒𝑣𝑒𝑙) and the job title (𝐽𝑜 𝑖𝑡𝑙𝑒) of the respondent and to whom the respondent is reporting to ( 𝑒𝑝𝑜𝑟𝑡 𝑜). The variables will be used purely to identify the background characteristics of the respondent.
3.5.3. Satisfaction level
The variables that test whether the organization has a clear digital business strategy (𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝑢𝑠𝑖𝑛𝑒𝑠𝑠 𝑆𝑡𝑟𝑎𝑡𝑒𝑔𝑦) and whether the organization is satisfied with their level of digital innovation ( 𝑒𝑣𝑒𝑙 𝑜𝑓 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛) will be used to test to what extend the organization finds digital innovation an important aspect within their organization.
An overview of all classifications used for the independent variables and an explanation of the independent variables can be found in Appendix A. Furthermore, the full questionnaire can be found in Appendix B.
3.6. Statistical analysis
25 𝑌𝑖 = 𝛼 + 𝜌𝑉𝑜𝐼 + 𝛽𝑋 + 𝜀𝑖
Where 𝑌𝑖 is the respective outcome variable (𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦) or one out of five
composite variable (𝑈𝑠𝑒𝑟 𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒, 𝑉𝑎𝑙𝑢𝑒 𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛, 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐸𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑆𝑐𝑎𝑛𝑛𝑖𝑛𝑔, 𝑆𝑘𝑖𝑙𝑙𝑠 and 𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛), 𝑉𝑜𝐼 is the variable of interest (𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑆𝑖𝑧𝑒), 𝑋 is a measure for the control variable 𝐼𝑛 𝑢𝑠𝑡𝑟𝑦 and 𝜀𝑖 is the error term. 𝑉𝑜𝐼 and 𝑋 will be the same in every model. However, 𝑌𝑖 will change in every model. In the table below (Table II), there is specified which hypothesis is represented by which model and which outcome variable is measured by the model.
4. Results
4.1. Sample
We recall from the previous section that our sample consists of 72 organizations. We divided the organizations into small and large organizations according to the Dutch Chamber of Commerce classification, resulting in 31 organizations that are classified as small organizations and 41 organizations that are classified as large organizations. When looking at the sample, it became clear that some organizations responded ‘does not apply’ on a substantial amount of questions, resulting in a very low score on digital innovation. We identified seven organizations that responded ‘does not apply’ on more than nine out of 18 questions regarding digital innovation. Furthermore, we have identified three organizations that responded ‘does not apply’ on more than seven out of 18 questions. This results in 10 organizations that have a substantial lower score on
Tested hypothesis Tested model Outcome variable, Variable of Interest,
Hypothesis 1 Model 1 Hypothesis 2 Model 2 Hypothesis 3 Model 3 Hypothesis 4 Model 4 Hypothesis 5 Model 5 Hypothesis 6 Model 6 𝑌𝑖 𝑉𝑜𝐼𝑖 𝑌 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦) 𝑌 𝑈𝑠𝑒𝑟 𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒) 𝑌 𝑉𝑎𝑙𝑢𝑒 𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛) 𝑌 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐸𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑆𝑐𝑎𝑛𝑛𝑖𝑛𝑔) 𝑌 𝑆𝑘𝑖𝑙𝑙𝑠 𝑌 𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛) 𝑉𝑜𝐼 𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑆𝑖𝑧𝑒) 𝑉𝑜𝐼 𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑆𝑖𝑧𝑒) 𝑉𝑜𝐼 𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑆𝑖𝑧𝑒) 𝑉𝑜𝐼 𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑆𝑖𝑧𝑒) 𝑉𝑜𝐼 𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑆𝑖𝑧𝑒) 𝑉𝑜𝐼 𝑂𝑟𝑔𝑎𝑛𝑖𝑧𝑎𝑡𝑖𝑜𝑛 𝑆𝑖𝑧𝑒) Table II
26 digital innovation than the other organizations in the sample. Therefore, we will classify these 10 organizations as missing values. We have to point out that all missing values belong to the group of small organizations. A possible explanation for this observation is that some small organizations are in the very first stage of developing their business and simply do not cover all the aspects of the business yet and those are sometimes businesses that are included in the diagnostic tool within the questionnaire. The sample we use in our analysis thus contains 62 organizations out of which 21 are classified as small organizations and 41 are classified as large organizations.
When looking at the reduced sample of 62 organizations, it becomes clear that the respondents are mainly man (51 out of 62). Furthermore, 32 of the respondents are C-level managers, 17 are Middle-level managers, eight are Lower-level managers and five respondents report to have another description for their position within the organization.
4.2. Background characteristics small and large organizations
Now that we have identified the sample and divided the sample into groups of small and large organizations, we can take a look at some important characteristics of both groups (Table III).
Small organizations (group = 1) Large organizations (group = 2)
N % N % p-value
2 9.52 25 60.98 0.001***
21 100.00 41 100.00
Yes, enterprise wide 17 80.95 22 53.66
Yes, within business units 0 0.00 10 24.39
No, but working on one 2 9.52 7 17.07
No 2 9.52 2 4.88 0.051*
21 100.00 41 100.00
Manufacturing 2 9.52 7 17.07
Transportation, Communications,
Electric, Gas, and Sanitary Service 13 61.90 7 17.07
Services 4 19.05 20 48.78 Other 2 9.52 7 17.07 0.005*** N 21 21 41 41 62 𝑒𝑣𝑒𝑙 𝑜𝑓 𝑖𝑔𝑖𝑡𝑎𝑙 𝑖𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 % 𝑛𝑜𝑡 𝑠𝑎𝑡𝑖𝑠𝑓𝑖𝑒 ) 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝑢𝑠𝑖𝑛𝑒𝑠𝑠 𝑠𝑡𝑟𝑎𝑡𝑒𝑔𝑦 𝐼𝑛 𝑢𝑠𝑡𝑟𝑦 𝑆𝐼 ) Table III
BACKGROUND CHARACTERISTICS OF GROUPS
Note 1: The p-value of the independent variables is based on a Chi-Square test.
27 The table depicts the number of observations, the percentage and the p-value following from Chi-Square tests. First, we take a look at possible identifiers of the level of digital innovation in the organizations, namely the satisfaction level with regard to digital innovation and whether the organizations had a digital business strategy in place. It becomes clear that many large organizations are unsatisfied with the level of digital innovation in their organization (60.98%) while only a few small organizations are unsatisfied (9.52%). A Chi-Square test shows that the difference in groups is significant (p≤0.010). Next to the satisfaction level, we identified whether the organizations had a digital business strategy. The largest part of the small organizations (80.95%) has an enterprise wide business strategy, while only two out of 21 reported not to have a digital business strategy but that they are working on one (9.52%) or did not have a digital business strategy at all (9.52%). None of the small organizations reported to have a digital business strategy within individual business units which may be explained by the fact that small organizations are simply too small to have multiple business units. The largest part of the large organizations also has a digital business strategy (78.05%). However, 53.66% implements the strategy enterprise wide while 24.39% implements the strategy within individual business units. However, there are still seven out of 41 large organizations (17.07%) that do not have a digital business strategy and is working on one. The smallest part of the large organizations reports not to have a digital business strategy at all (4.88%). A Chi-Square test shows that the difference between the groups with regard to having a digital business strategy is significant at the 10% level (p=0.051). Thus, both groups differ significantly when it comes to implementing a digital business strategy. Second, we determined the industries in which the organizations are operating to check for differences between the groups. We reduced Two of the small organizations (9.52%) and seven of the large organizations (17.07%) are operating in the Manufacturing industry. Furthermore, 13 (61.90%) of the small organizations and seven of the large organizations (17.07%) are operating in the Transportation, Communication, Electric, Gas and Sanitary Service industry. Four of the small organizations (19.05%) and 20 (48.78%) of the large organizations are operating in the Services industry. Finally, two small organizations (9.52%) and seven large organizations (17.07%) are operating in the industry category “Other”. A Chi-Square test shows that the differences in industries are significant (p≤0.010) between the small and the large organizations.
28 Later on, we will analyze whether the industry differences are having an effect on the level of digital innovation when this variable is included in the regression.
Furthermore, it is interesting to look at the digital innovation satisfaction level and the level of digital innovation obtained from the diagnostic toolin more detail (Figure III). However, we did not include the satisfaction level of the respondent into the OLS regression because the variable is endogenously determined. When looking at the figure, it can be concluded that the organizations who are satisfied with the level of digital innovation in their organization, also report a higher score on the level of digital innovation compared to the organizations who are not satisfied with the level of digital innovation in their organization. Furthermore, the spread in the level of digital innovation of the large organizations who are not satisfied with the level of digital innovation is very high, ranging from a score of 24 to a score of 53. Finally, the small organizations report a higher score on digital innovation when they are satisfied compared to the large organizations when they are satisfied.
Second, in the questionnaire we also asked the respondents whether they had a digital business vision/strategy. When having a digital business vision/strategy or not is offset against the level of digital innovation (Figure IV), it shows that the small organizations who have an enterprise wide business strategy have a higher level of digital innovation than large organizations. The scores of
Figure IV
DIGITAL BUS INES S S TRATEGY AND LEVEL OF DIGITAL INNOVATION
Figure III
S ATIS FACTION LEVEL OF DIGITAL INNOVATION AND LEVEL OF DIGITAL INNOVATION
29 these organizations are also more centered when they are in the small organization category compared to the large organization category. Furthermore, the level of digital innovation is declining faster as the organization moves from having an enterprise wide digital business strategy to not having a digital business strategy at all when in the large organization category compared to the small organization category.
4.3. Differences in the level of digital innovation between small and large organizations
We will now move on to the differences between small organizations and large organizations with regard to the dependent variable 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦 and the composite variables 𝑈𝑠𝑒𝑟 𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒, 𝑉𝑎𝑙𝑢𝑒 𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛, 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐸𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑆𝑐𝑎𝑛𝑛𝑖𝑛𝑔, 𝑆𝑘𝑖𝑙𝑙𝑠 and 𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛 (Table IV).
The table depicts the mean, standard deviation and p-value following from Independent T-tests. Out of 60 points in total, the small organizations score 47.762 and the large organizations score 41.366 on 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦 which is significant (p≤0.010) and is suggesting that small organizations have a higher level of digital innovation compared to large organizations. Furthermore, we will look at the composite variables that might cause this significant difference in the level of digital innovation between groups. The small organizations have a higher score on 𝑈𝑠𝑒𝑟 𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒 compared to large organizations, however this difference is not significant (p=0.153). The same holds for 𝑉𝑎𝑙𝑢𝑒 𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 (p=0.533). However, when looking at 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐸𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑆𝑐𝑎𝑛𝑛𝑖𝑛𝑔, the score of the small organizations is significantly higher
Small organizations (group = 1) Large organizations (group = 2)
Mean Std. Dev. Mean Std. Dev p-value
47.762 5.517 41.366 6.765 0.001*** 9.500 2.975 8.598 1.918 0.153 8.691 2.358 8.354 1.800 0.533 10.333 1.426 8.781 1.904 0.002*** 10.238 1.700 7.561 1.500 0.001*** 9.000 2.429 8.073 1.752 0.090* N 21 21 41 41 62 Table IV
DIFFERENCES IN DIGITAL INNOVATION BETWEEN GROUPS
𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦 𝑈𝑠𝑒𝑟 𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒 𝑉𝑎𝑙𝑢𝑒 𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐸𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑆𝑐𝑎𝑛𝑛𝑖𝑛𝑔 𝑆𝑘𝑖𝑙𝑙𝑠 𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛
Note 1: The dependent variable 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦 can take a value between 0 and 60. The composite variables 𝑈𝑠𝑒𝑟 𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒, 𝑉𝑎𝑙𝑢𝑒 𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛, 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐸𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑆𝑐𝑎𝑛𝑛𝑖𝑛𝑔, 𝑆𝑘𝑖𝑙𝑙𝑠 and 𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛 can take a value between 0 and 12.
Note 2: The p-value of the dependent variables is based on an Independent T-test.
30 (p≤0.010) than the score of the large organizations. Also for 𝑆𝑘𝑖𝑙𝑙𝑠, the score of the small organizations is significantly higher (p≤0.010) than the score of the large organizations. Finally, the score on 𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛 is significantly higher but weaker (p≤0.100) for small organizations compared to large organizations. This suggests that mainly the composite variables 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐸𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑆𝑐𝑎𝑛𝑛𝑖𝑛𝑔 and 𝑆𝑘𝑖𝑙𝑙𝑠 and in a weaker sense 𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛 are responsible for the difference in the level of digital innovation between groups. Based on the previous analysis, we have obtained the following results:
Result 1
— A negative correlation is observed between the size of the organization and the level of digital innovation.
Result 2
— The observed negative correlation between the size of the organization and the level of digital innovation is mainly caused by the negative correlation between the size of the organization and 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐸𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑆𝑐𝑎𝑛𝑛𝑖𝑛𝑔 and 𝑆𝑘𝑖𝑙𝑙𝑠, and 𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛 in a weaker sense.
When we offset the level of digital innovation against the seven different size categories of the organizations (Figure V), it shows that the level of digital innovation decreases when the size of the organization increases. Furthermore, it shows that the spread of the level of digital innovation increases as the size of the organization increases suggesting a growing difference in the level of digital innovation when the size of the innovation increases.
Figure V
ORGANIZATION S IZE AND LEVEL OF DIGITAL INNOVATION
31
4.4. Determinants of digital innovation
We ran OLS regressions to control for background variables when we compare the different scores on the dependent variable (and the composite variables) for the group of small organizations and the group of large organizations. In this thesis, we would like to include the industry in which the organizations are operating. Including the industry enables us to control for the effects of the industry on the level of digital innovation so that we can check whether the results from the Independent T-tests are consistent when including the industry. Because the variables organization size and industry are both categorical variables, we have to create dummy variables in order to include these variables into the regression. Because we are working with a categorical size variable, we take the average number of employees from every organization size category and include these values into the regression. For the organization size, there are thus six dummy variables and one reference size included in the regression. For the industry, three dummy variables and one reference industry was included in the regression. It is important to note that OLS regression is more precise and efficient in large samples than in small samples. However, the OLS regression may lead to important insights that could be a starting point for further research even though our sample size is relatively small.
First, we ran an OLS regression with the dependent variable 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦 and independent variables organization size and industry. After that we ran separate OLS regressions for every composite variable, namely 𝑈𝑠𝑒𝑟 𝐸𝑥𝑝𝑒𝑟𝑖𝑒𝑛𝑐𝑒, 𝑉𝑎𝑙𝑢𝑒 𝑃𝑟𝑜𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛, 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐸𝑣𝑜𝑙𝑢𝑡𝑖𝑜𝑛 𝑆𝑐𝑎𝑛𝑛𝑖𝑛𝑔, 𝑆𝑘𝑖𝑙𝑙𝑠 and 𝐼𝑚𝑝𝑟𝑜𝑣𝑖𝑠𝑎𝑡𝑖𝑜𝑛 (Table V).
The table depicts the coefficients and the standard errors following from the OLS regression. Furthermore, the significance levels are stated behind the coefficients. First of all, we will look at the OLS regression with the dependent variable 𝐷𝑖𝑔𝑖𝑡𝑎𝑙 𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑜𝑛 𝐴𝑐𝑡𝑖𝑣𝑖𝑡𝑦 and it becomes clear that that for all organization sizes, the level of digital innovation will decrease compared to the reference size of five and that this is also significant in all the cases.
