What Motivates the Commercialization of Innovations?

7 downloads 62604 Views 573KB Size Report
If, on the other hand, a developer's software products did not meet Apple's ... The major change with the introduction of the Apple App Store was structural, ..... the quasi-legal nature of the jailbreak community, it is not possible to account for.
Filthy Lucre: What Motivates the Commercialization of Innovations?

Ethan Mollick Wharton School of Management University of Pennsylvania 2000 Steinberg-Dietrich Hall 3620 Locust Walk Philadelphia, PA 19023 215-898-6361 [email protected]

Acknowledgements: I would like to acknowledge the support of the Wharton Entrepreneurship and Family Business Research Centre at CERT and The Centre of Excellence for Applied Research and Training Term Fund. Significant help in gathering data was provided by Isaac VanDuyn, Jason Johnson, and anonymous members of the iPhone hacking community, including Big Boss. I would also like to thank Emilie Feldman, attendees at the Bowman Seminar at Wharton, the University of Maryland Entrepreneurship Conference, and the Wharton Junior Faculty Group, and the MIT User Innovation Conference for their input and insight.

1

Electronic copy available at: http://ssrn.com/abstract=1742380

Filthy Lucre: What Motivates the Commercialization of Innovations?

Abstract: It is generally assumed that individuals choose whether to commercialize novel products or innovations based on their expected economic return. However, many innovators with valuable products may be motivated by non-economic incentives that can often be antithetical to profit-seeking. Despite the importance of understanding the motivations of individuals facing commercialization decisions, academic studies of the topic are lacking, in part due to the difficulty of identifying nascent entrepreneurs and early-stage innovators. Using a unique natural experiment involving the launch of the Apple iPhone to address these empirical difficulties, I find that expected returns play little role in commercialization decisions for many highly innovative individuals, while non-economic motivations are significant, and may suppress commercialization. This suggests that which innovations become commercialized is more dependent on individual variation than the expected value of the innovation.

2

Electronic copy available at: http://ssrn.com/abstract=1742380

Introduction Standard economic logic predicts that individuals who spot profitable opportunities or develop economically valuable inventions will seek to commercialize them. This rational economic assessment is what is assumed to drive entrepreneurs to launch companies and inventors to license technologies, and thus serves as the engine of the cycle of creative destruction and innovation (Kirzner, 1997; Schumpeter, 1934). Despite the importance of the decision to commercialize in understanding innovation and entrepreneurship, there has been little research on the topic, and what work there has been has emphasized the demand side of the phenomenon (Thornton, 1999). The expectation has been that perceived market demand is the major driver of whether individuals choose to commercialize their products (Choi & Shepherd, 2004). The nature of demand may be filtered through the cognitive lenses and psychology of the entrepreneur (Haynie, Shepherd, & McMullen, 2009; Hmieleski & Baron, 2009), and weighed against opportunity cost (Shane, 2003), but, ultimately perceived demand, and the revenue associated with it, is assumed to drive the calculus of commercialization. Neglected in this analysis is the supply side, the potential commercializers themselves, who may have motivations that differ from achieving pure economic profit (Scott Morton & Podolny, 2002). Understanding the influence of non-economic motivations upon commercialization is especially critical because individuals can have goals that are often antithetical to profit-seeking, such as among innovators or scientists whose goal is to increase their standing within a peer community (Jeppesen & Frederiksen, 2006; Stern, 2004) or among individuals who seek to maximize control over their own work environment (Hurst & Pugsley, 2010; Wasserman, 2006).

3

Electronic copy available at: http://ssrn.com/abstract=1742380

Complicating our understanding of the motivations to commercialize is the difficulty of studying individuals as they are choosing to commercialize innovations. It is hard to identify nascent entrepreneurs, inventors before they patent, and other individuals facing commercialization decisions. Using a natural experiment to overcome the difficulty of studying individual innovators before they commercialize their products, this paper examines the role of human agency in commercialization decisions, and the extent to which economic profit and noneconomic rationales affects the motivations of innovators with potentially valuable products. I find that, contrary to the untested expectation that demand is the critical factor in commercialization, expected returns play little role in commercialization decisions for many highly innovative individuals, while non-economic motivations are significant, and may suppress commercialization, even when commercialization seems economically rational.

The Motivation of Commercialization Decisions Commercialization is a phenomenon of great importance, as it encompasses a number of approaches by which the ideas and innovations of individuals become part of the wider economy by delivering some of the value of the innovation to the individual who created it. Individuals may engage in entrepreneurship, creating new ventures or spinning out from existing organizations (Shane & Venkataraman, 2000). Alternately, an innovator may decide to license an idea commercially, or patent a concept so that it may be licensed (Stuart & Ding, 2006). Or, they may engage in small-scale dabbling with commercialization to determine if there is product need before engaging in full production (Shah & Tripsas, 2007). Individual innovators may even explore multiple paths, attempting entrepreneurship only to later shift to a different strategy. All of these methods result in an idea being commercially exploited, with value being captured by

4

the entrepreneur and created for society at large (Lepak, Smith, & Taylor, 2007; Schumpeter, 1942). Most studies of commercialization occur within the literature on entrepreneurship and examine when individuals, based on expected demand, choose to start their own ventures to take advantage of an opportunity they discover or create. Becoming an entrepreneur is a complicated process, with many possible definitions (Hannan & Freeman, 1993; Reynolds, 1997). However, commercialization through entrepreneurship is generally assumed to require a minimum of two steps. The potential entrepreneur must perceive and evaluate an opportunity for entrepreneurship, and then decide to exploit the opportunity by launching a new venture (Alvarez, 2007; Shane, 2003). Drawing on the perspectives of Austrian economics (Hayek, 1945; Kirzner, 1997), a wide range of researchers have examined the nature of these entrepreneurial opportunities and how individuals identify them (for an overview of 64 such papers, see Short, Ketchen, Shook, & Ireland, 2009). Less is known about the decision to exploit an opportunity (also called “venture initiation,” see Ruef, 2005), perhaps because of the difficulty in identifying potential entrepreneurs before they make a commercialization decision. As a result, our current view of the commercialization decision is guided more by assumption than by evidence. The assumption is that exploitation is based on information, as nascent entrepreneurs gain more positive information about their potential venture, they are more likely to exploit it (Shane, 2003). Individuals chose to commercialize because, conditional on their psychology and opportunity set, they have uncovered or created an opportunity that they expect to be worth the opportunity cost of exploiting it as an entrepreneur. While this approach allows for differences in cognition or information (Hayward, Shepherd, & Griffin, 2006; J.S. 5

McMullen, Shepherd, & Shepherd, 2006; Shane, Locke, & Collins, 2003), it assumes a common profit-driven motivation for potential entrepreneurs in which there is little room for individual preferences, as economic rationality is the primary determinant in decision-making. A similar assumption underlies other studies of commercialization in other contexts, such as the literature on user innovators, those individuals who develop important innovations to solve their own problems (Von Hippel, 1988). Among user innovators, commercialization, when it occurs, is again driven by expected profit. In the most comprehensive study of the subject, Shah and Tripsas (2007) studied when these user innovators chose to exploit their ideas commercially. They observed that 82 percent of the founders of juvenile product firms were non-professional user innovators, usually parents, who created products based on their own needs and interests. Rather than planning on becoming entrepreneurs, amateur user innovators only gradually realize the value of their solution to a wider market, taking input from the community of users as they continued to refine their innovation. As opposed to the traditional view of entrepreneurship, in the conception of Shah and Tripsas, user innovators first exploit their own opportunity, and, only after sharing it with the community, do they formally identify the opportunity to become entrepreneurs. Though the process is reversed from traditional entrepreneurship, ultimately the decision to commercialize is driven by the knowledge of potential demand. However, there is reason to believe that a pure economic assessment does not always explain commercialization decisions, as the literature currently assumes, and that the nature and social context of the individuals themselves have a profound influence on the process. One example of this can be found in the work of Scott Morton and Podolny (2002) in the wine industry. They found that hobbyists vintners have different goals than producers who seek to 6

maximize financial returns. Hobbyists emphasize the utility that comes from being a quality producer over the profits of being an efficient producer, even though that may result in lower financial returns. This effect is not limited to wineries: in an examination of the data from the Panel Study on Entrepreneur Dynamics, Hurst and Pugsley (2010) found that many small business owners were motivated to start businesses for reasons other than pure financial gain. If this is the case for entrepreneurs, who are under market pressure to generate an economic profit to keep their firms operating, then non-economic motivation is likely even more common among new innovators and nascent entrepreneurs who have not yet commercialized their work. The effect of non-economic motivations on commercialization decisions may go even further, supplanting economic returns as the key decision-making criteria for innovative individuals. While hobbyist wineries are willing to give up some economic profit in return for higher quality products, among many populations of potential entrepreneurs, non-economic motivations may actively discourage profit.

For example, among communities of innovators,

such as open source software developers, the value of sharing and free-exchange among community members makes commercialization unlikely (Harhoff, Henkel, & Von Hippel, 2003; Von Hippel & Von Krogh, 2003). This is, in part, due to the values of the community, but also because many potentially commercializable products are group efforts, making partitioning of economic profits difficult. Many communities of innovators emphasize making innovations free, including communities around sports equipment (Franke & Shah, 2003; Hienerth, 2006), video games (Edery & Mollick, 2009), and those supporting the early development of the airplane and personal computers (Meyer, 2007). Non-economic, and often anti-commercial, motivation extends to scientists, who give up potential profits in order to be allowed to participate in the scientific community (Stern, 2004), and are less inclined to engage in commercial activity if the 7

norms of their institutions do not support economic motivations (Gans, Hsu, & Stern, 2002; Stuart & Ding, 2006). Thus, non-economic incentives can be in direct opposition to economic incentives among many groups of innovative individuals. Given the preponderance of research showing non-economic incentives relating to commercialization in fields ranging from research science to software development, it is somewhat surprising that the concept of profit maximizing entrepreneurs remains prevalent.

The Study Setting: Commercializing the iPhone Part of the reason that the commercialization decision has been understudied is due to the difficulty of finding nascent entrepreneurs and potential innovators before they decide to commercialize. To avoid this problem, I identified a natural experiment associated with the launch of the iPhone smart phone that allowed me to sidestep this problem. This is an especially compelling sample because of the importance of the innovations developed by the individuals in the sample. In addition to eventually starting a number of successful ventures, these innovators developed key new software categories for the iOS platform: the first photography applications, the first games, the first electronic reading applications, and many others. To understand the choices of these individuals in commercializing their work, it is necessary to understand the launch of the iPhone. For the first year after its release, from June of 2007 to July of 2008, the iPhone was technically a closed system, only the original Applesupplied software programs were officially permitted. However, as would be expected from the literature on user communities (Mollick, 2005), it was not long before dedicated individuals began to “hack,” or modify, the iPhone operating system, writing their own software for the iPhone platform (Al-Zarouni & Al-Hajri, 2007). Moving beyond creating original software, 8

these enterprising individuals also created their own shadow distribution system for this software, so that anyone who opened their iPhone to modifications was also able to download a wide range of software. Using these underground software packages required a process called “jailbreaking.” Jailbreaking modified the iPhone in a manner harmless to the phone, but which was strongly discouraged by Apple (installing outside applications would void the warranty of the phone, among other issues). Despite this, millions of iPhones were jailbroken, a minimum of 25% of all iPhones sold (Krazit, 2007). Since the iPhone operating system was closely related to the Macintosh OS, which in turn was derived from UNIX, it proved a relatively easy programming environment for computer enthusiasts. The result was, in the words of Wired magazine, “an entire underground ecosystem: the Jailbreak community.”(Chen, 2009) Hundreds of jailbroken applications were developed, the vast majority of them as free software. To support these individuals, large number of websites and discussion forums served as foci for community members developing applications, and as areas where new applications were discussed. This period of furious underground innovation ended on July 12, 2008, when Apple created the iPhone Application Store while simultaneously updating the software of the iPhone. With the new software, users of the iPhone could purchase (or download for free) software by outside developers that was legally developed for the iPhone and distributed through Apple’s own App Store. Such software was approved by Apple, so it had to meet certain relatively welldefined criteria. However, Apple allowed developers to price software as they saw fit, and gave developers 70% of the revenue from any sales. Within two months 3,465 applications were

9

listed on iTunes. Apple created the possibility of commercialization, where previously there had been none. In this new environment, individuals who had written pre-existing software faced a choice as to whether they would exploit the opportunity afforded by the Application Store. All software was written for earlier versions of the iPhone operating system, and had to be updated to the new system, which was a trivial process. After that was done, individuals could either continue to make the software available for free, or else begin to charge for it using Apple’s App Store. If, on the other hand, a developer’s software products did not meet Apple’s standards, or if the developer simply refused to work through the App Store, they could still make their software available for free through the underground distribution methods that the community maintained. Adding to the options for commercialization, the underground distribution system, now called Cydia, was modified to allow individuals to charge for software, creating a commercialization channel outside of the official App Store. The launch of the App Store and its effect on the vibrant jailbreak community provides an excellent natural experiment to test hypotheses involving commercialization and individual choice. The major change with the introduction of the Apple App Store was structural, creating a market where there had previously not been one. Prior to the App Store, individuals had already created valuable applications for the iPhone, and given them away for free, since there were no other options. Now, at the same moment, and in the same strategic context, all of these amateurs now faced a clear choice among three options: commercialize their work, continue their commitment to free software, or else walk away and abandon their project.

10

To commercialize existing work, or to continue to give it away, individuals needed to update their software to the new iPhone software, or it would not work on new phones. Doing so was trivial. In the words of one developer, “It was an incredibly easy process; you are writing the same code [as you did before the App Store launched]. The difference is that Apple is providing the documentation in one case, and in the other you have to reverse engineer it. It is the same application.” Thus, opportunity costs, a key consideration in economic decisionmaking, were minimal for the developers of jailbroken apps. The result is an experiment with a three outcomes (individuals could choose commercialization, free distribution, or abandon their creations), but a variety of paths to reach them, as can be seen in Figure 1. Originally, developers made their jailbreak applications available for free. A few individuals in these early days had attempted to sell their jailbroken application, but this was rare (less than 10% of developers) and difficult because there were no trusted mechanisms available by which individuals could charge customers. For the vast majority of developers, their first opportunity to commercialize their work occurred when the new App Store was released. They could choose to update their software for the new operating system version, or else abandon their projects. If they did choose to update the software, they were faced with a second choice, they could also choose continue to offer it for free or commercialize it.

[FIGURE 1 ABOUT HERE]

11

The potential economic profit associated with commercialization proved to be vast. For example, one firm that was started by innovators in the jailbreak community was later sold to the Walt Disney Company for over $40 million. Thus, it would be entirely rational to expect individuals, with almost no opportunity cost and essentially complete products, to pursue commercialization. However, as suggested by Scott Morton and Podolny (2002), pure assessment of profit is not the only lens through which individuals view commercialization, even with such vast potential profits available.

Hypotheses The iPhone natural experiment gives us a case where individuals are suddenly faced with the opportunity to commercialize their work. Many take advantage of the opportunity to generate economic profit, but, surprisingly, not all do. This offers us a strong analytical purchase on the question of when commercialization occurs, as the opportunity cost is trivial and significant information is available to innovative individuals about the potential reception of their products. As discussed, we would expect two different factors to influence whether individuals commercialize their products. First, there is the general assumption in the commercialization literature that commercialization is driven by economic profit , filtered through the psychology and individual tastes of the nascent entrepreneur (Shane & Venkataraman, 2000). The higher value an opportunity is, the more likely an individual is to exploit it through entrepreneurship (Kirzner, 1997; Schumpeter, 1934). Specifically, as Choi (2004) argues, the more information a potential entrepreneur has about their customer needs, the more likely they are to exploit an opportunity. Similarly, in Shah and Tripsas’s (2007) model it is implied that, as individuals learn more about 12

the value of the opportunity they have identified, they are more likely to become entrepreneurs. This suggests: H1. As the expected economic value of a product increases, the likelihood of commercialization increases. The standard view of entrepreneurial motivations would argue that, given a similar opportunity cost (Amit, Muller, & Cockburn, 1995) among individuals, H1 would be the key predictor of whether individuals to commercialize their products. However, some individuals may seek to maximize other forms of utility over profitability, which may result in an alternative set of criteria for the commercialization decision. In some cases, individuals chose to emphasize quality over profit (Scott Morton & Podolny, 2002), in other contexts, issues of lifestyle or control take precedence (Hurst & Pugsley, 2010). In the case of the early jailbreak iPhone community, where financial rewards were impossible, this alternate form of motivation was evident in the heavy participation of individuals associated with the open source community and “hacker” communities. For open source and hacker communities, the rewards for developing products are not financial, but come from a mix of intrinsic motivation, skill development, and community interaction (Lakhani & Wolf, 2005; Raymond, 2001; Shah, 2006). For example, the widelypublished Hacker Ethic states that individuals participating in the open source software space have “the belief that information-sharing is a powerful positive good, and that it is an ethical duty of hackers to share their expertise by writing open-source code and facilitating access to information and to computing resources wherever possible.” (Himanen, Torvalds, & Castells, 2002). For individuals working on open source, free software is more than an economic choice; 13

it is also a philosophic approach. Free flow of information, and its formal expression in the open source movement, represents an idealistic statement about the nature of property and ownership, which, in turn, motivates the communities of open source developers (Bagozzi & Dholakia, 2006). For individuals who demonstrate ties to open source, we would therefore expect that the rewards of community affiliation often contrast directly with economic motivation, as they do in other innovative communities, such as academic scientists (Stuart & Ding, 2006; Von Hippel, 2005). H2. As individual commitment to non-economic systems of reward increase, the likelihood of commercialization decreases. These are two deeply contrasting views of the motivation to commercialize, though they are likely to operate independently. That is, if H1 is true, then we would expect demand to matter to individuals whether or not they are also interested in alternative reward systems, though perhaps at a lesser rate. It is the existence of these effects, and their magnitude as well, that is of particular interest. In the following empirical analysis, I hope to tease apart the motivations for innovative individuals facing the choice to commercialize. On one hand, the literature assumes that higher expected profits would increase the chances of commercialization, while, on the other, non-economic motivations may suppress profit-seeking.

Methodology Data All of the variables, except for outcome data, came from the period before the launch of the App Store, when only the unofficial jailbreak community existed. Data was collected on 14

individual applications from the largest distribution site for software before the App Store, and was supplemented by email correspondence with approximately 20% of the developers of the software. The initial data from the software distribution site consisted of the name of the application developed, the number of downloads, and the number of versions of the software that were released. Applications that consisted solely of text, images, or which were add-ons to existing applications (such as holiday themes for a photography application) were dropped from the sample. Also eliminated from the sample were the developers who had attempted to charge for jailbroken applications. For each piece of software, two research assistants collected additional data using a directory of all iPhone applications (ModMyi) to identify whether a the package was updated for the new iPhone firmware, as well as the package’s creators and whether the project was open source. Since ModMyi may not always be complete, it was supplemented with another directory, AppSafari, as well as with direct searches for product home pages and for product information in the largest Apple iPhone discussion forum, Hackint0sh. The author and both research assistants agreed on the final coding of each application. Each application was then cross-referenced with the iTunes Application Store, both under its original name, the author’s name, and by category. The price of any software identified in the iTunes Application Store was noted, as was any additional software by that developer. It is possible that through a combination of name changes and title changes, that some applications may have been listed in the iTunes Application Store but not identified, but interviews with developers did not locate any errors. A total of 158 jailbroken applications that fit the criteria were identified from the list, of which 88 were abandoned by the inventor, 35 were kept free, and 35 were commercialized. 15

These jailbroken applications were created by 88 developers, of which 32 abandoned all their development efforts, 19 continued to develop only free software, and 37 commercialized by selling at least one software product (8 of these commercialized by creating new for-charge software products for the App Store after the launch of the App Store, some of which were variations on jailbroken software). Figure 2 shows these outcomes. [INSERT FIGURE 2 HERE]

Variables The variables are summarized in Table 1. The dependent variable is whether the innovative individual abandoned development on all of their applications, updated at least one application and kept all of their updated applications available for free, or else updated their applications and placed at least one for sale. For a developer to be considered to be noncommercial, all of their software needed to be available for free after the launch of the App Store. For a developer to be considered to have commercialized, at least one application needed to be listed for sale on either the App Store or through Cydia. Additionally, if a developer created a new paid application within two years after the launch of the App Store, they were considered to have commercialized, even if they kept their jailbroken applications updated and free. Expected market demand was measured as the log of the mean number of times each piece of software created by a developer was downloaded during the period before the Application Store was opened by Apple, though the number of downloads for the most popular app for each developer was also tested, with the same result. Downloads are a strong indicator of demand for a product, and act as a very direct signal of market interest. The level of data 16

provided to the individual about demand for their product from potential customers was therefore quite precise – highly downloaded products were likely to be in high demand after the legal release of the App Store. This logic is supported by the first publicly available survey of installed applications, conducted seven months after the launch of the iPhone. Of the top ten most installed iPhone applications (out of over 32,000 available applications at that time), two had originated within the jailbroken community, and they represented the first and second most downloaded original games before the launch of the Application Store (Jurutka, 2009), demonstrating that pre-App Store downloads acted as a strong signal of actual demand. As was verified in discussions with members of the sample, the relevant non-commercial affiliation among early Apple iPhone hackers was the open source community. Thus, community affiliation was measured by the proportion of a developer’s applications where the source code was ever revealed as an open source project before the launch of the App Store. In many cases, open source software does not require that the product be free, and, even in cases where free software was required by an open source license, developers could create other products that were not free. Thus, individuals could still commercialize even if they maintained some set of open source software. The open source licenses used included 31 variants on the GNU license, 4 using the BSD license, and 14 others (including licenses not clearly identified, MIT licenses, and other variants). A number of control variables were also used. An individual’s personal commitment to the software package was measured by the number of times a developer updated their free software before the release of the Application Store. Individual psychology and risk tolerances may affect how an opportunity is perceived (Brockhaus & Horwitz, 2002). Thus, I would expect 17

the effort that an individual expends exploring an opportunity to have an impact on how an opportunity is perceived, and therefore whether it is exploited. This is because sunk costs and escalation of commitment are likely to bias an individual (Brockner, 1992; Staw, 1981), but also because effort is an indicator of serious interest on behalf of a developer in continuing to create a program, and therefore a positive bias. An additional control variable was whether or not the developer had software that was banned from the App Store by Apple. Apple bans software for a variety of reasons, most notably because it duplicates other Apple functions, is offensive, or allows users to modify the underlying features of how an iPhone works. A ban on one piece of software does not ban individuals from releasing other software for the iPhone. However, banned software is presumably harder to sell (though better methods for selling jailbroken software on the iPhone were developed after the launch of the App Store), and presents more of a structural barrier to commercialization. This variable was determined by statements by the developers themselves, and also an analysis by the research assistants of whether a software package would be obviously disallowed by Apple. If the software was not obviously disallowed, and there were no statements to state that the software was banned, it was not coded as banned. A dummy was also used to indicate whether a software package was a game, since it might be expected that entertainment applications had different characteristics than utilities or other applications. A final variable was used to indicate whether the developer had previously asked for donations to support their free software effort prior to the launch of the App Store. This usually consisted of a request in the application description for financial support via PayPal. Table 1 shows the summary of variables by developer, Table 2 shows the correlation matrix. 18

Given the quasi-legal nature of the jailbreak community, it is not possible to account for individual characteristics of the developers themselves, who tend to be elusive. I will draw on the rich qualitative data available from the sample of individuals that were willing to answer questions later in this paper, but it is important to note that opportunity costs were extremely low to the point of being trivial. Actually updating the application involved almost no work, and Apple handled all of the commercial aspects of the sale of the software, simply sending checks to the developers. Additionally, the iPhone application market represents a substantial opportunity: one jailbroken application sold over $1m a month in revenue after it was released on the Application Store.

[INSERT TABLE 1 HERE] [INSERT TABLE 2 HERE]

Results The outcome of each software product is summarized in Figure 2. The effect of the Apple App Store launch was measured with McNemar’s Chi Squared Test, using the full sample (including applications that were not free prior to the launch of the App Store). As might be expected from the reported data in Figure 2, the launch of the App Store led to commercialization at better than p=.0001. [INSERT FIGURE 2 HERE] To examine the two hypotheses, I next used a multinomial logistic regression. There are three outcomes: commercialization, non-commercialization, or abandonment of all applications. 19

Table 3 reports the results, and Table 4 the marginal effects, for ease of interpretation. The result indicates that affiliation with open source communities reduces the chances of commercialization, and increases the chance of abandonment, supporting H2. Contrary to H1, however, expected demand is only marginally related to whether or not individuals choose to commercialize, with higher demand being weakly associated with non-commercialization. [INSERT TABLE 3 HERE] [INSERT TABLE 4 HERE]

Though the lack of strong support for H1 may seem surprising, the results proved robust across a variety of alternative specifications at the developer level. There was no indication of an interaction effect between the expected demand and open source affiliation. Demand did not predict commercialization even if all individuals associated with open source were removed, if all other variables beside demand were removed, and under various non-linear specifications. These results were also robust to measuring demand by the developers’ most successful applications, rather than the demand of the average application. However, expected demand is predictive of which specific applications individuals chose to commercialize, if they decide to commercialize at all. As can be seen in Table 5, within the population of individuals who chose to commercialize their applications, H1 is supported – the applications with the highest demand are most likely to be commercialized. Individuals who respond to economic profit are therefore consistent in their application of economic logic, commercializing their most valuable innovations. 20

[INSERT TABLE 5 HERE] Many individuals quickly commercialize when presented with an attractive opportunity. However, commercialization is not universal, nor is it based simply on a present-value assessment of the opportunity presented by charging for the software. First, the expected economic value of commercialization (as determined by number of downloads prior to the launch of the App Store) is not generally predictive of whether or not individuals commercialize. Secondly, as hypothesized, individuals who have previously responded to non-economic incentives by engaging in open source activities are less likely to commercialize their products.

Discussion Given the robustness of the results, perhaps the most intriguing result is the weak support for the first hypothesis. Strong indicators for demand did not predict which individuals would commercialize their innovations. However, for those individuals who did commercialize their work, demand was important in deciding which potential applications to commercialize. These two findings suggest something important about the population of innovative individuals – in aggregate, these innovators does not respond as expected to economic profit alone, but those individuals who do chose to commercialize seem to generally behave in a rational manner when evaluating potential economic returns. The fact that sample of innovators as a whole did not respond to obvious economic incentives may seem surprising. Individually, of course, many individuals were motivated by economic profit, and, indeed, had started successful entrepreneurial ventures based on their innovations. Those individuals who commercialized products responded as expected to 21

economic incentives, as is shown in Table 5. However, many of the innovators in the sample did not react as predicted to economic incentives to commercialize. These individuals had alternate views of the norms and values of economic exchange that mitigated the importance of economic profit as a goal (Scott Morton & Podolny, 2002; Sine & Lee, 2009). As a result, these individuals tended not to conceive of themselves as potential profit-maximizing entrepreneurs, and did not take an entrepreneurial approach to commercialization (Cardon, Wincent, Singh, & Drnovsek, 2009; Fauchart & Gruber, 2011). This made the process of commercialization itself seem extremely challenging, even if it was objectively straightforward. Discussions with App Store developers led some context supporting this view. One popular application developer, when asked why he did not sell his apps on the App Store, suggested that actually commercializing his application required an underlying set of attributes that he did not possess: “It takes a lot more than being a good programmer to make an app developer. It takes good business skills, and I don’t feel I have that.” Similar thoughts were expressed by other developers. Another open source developer who had created a number of innovative apps with millions of free downloads reported that “It is not that I have some allergy to making money. I know how stupid I have been to not make money. But I don’t have the resources to put together the business stuff.” From the perspective of a rational individual assessing an opportunity, statements like these make no sense. Apple handles all of the business operations for developers, developers merely have to submit an application and accept a check for 70% of revenues. Business skills are not required to commercialize an application, and, in any event, it would have been possible to hire or partner with any number of “business people” to handle any issues that might arise.

22

Instead, these sentiments make sense in the context of a group of individuals who do not conceive of themselves as potential commercializers of products. Growing evidence suggests that entrepreneurial self-identity can play an important role in how individuals address the problems associated with commercializing innovations (Cardon et al., 2009; Farmer, Yao, & Kung-Mcintyre, 2011; Fauchart & Gruber, 2011). The iPhone developers who did not commercialize despite high demand clearly did not identify as profit-seeking entrepreneurs, making the missing “business skills” or “business stuff” mysterious. Thus, even though these developers had produced well-regarded applications, they decided not to commercialize. Pure market demand and near-zero economic opportunity cost were outweighed by a perspective that made commercialization seem daunting and distant. So what does motivate these individuals to continue to develop and refine their innovations? The strong support for H2 suggests that the non-economic incentives associated with the open source and hacker communities were often of greater influence than economic profit. Individuals with strong ties to open source felt bound to adhere to the free revealing norms of their community, as suggested in earlier work on open source groups (Bagozzi & Dholakia, 2006). One developer of a highly downloaded open source application wrote that, “I released my app as open source, and I intended not to break it, so I would've had to do this all for free.” In this case, the individual’s commitment to open source meant that they would have refused to charge for the application, but would have still felt obligated to support the application for free. To an outside observer, the obvious solutions would have either been to sell the application, or else release it for free without any support, so that individuals could gain some benefit from the free work. However, the non-economic incentive of the community’s ideals

23

precludes either of these options, and several developers mentioned the fear of disappointing the community by failing to support commercial products. The incentives that come from working with a community also surfaced in other ways. The joy of pure discovery and innovation unencumbered by commerce that is common among innovative groups was frequently invoked (Lakhani & Wolf, 2005; Mollick, 2005). As one developer told me, “there is a certain degree of satisfaction to pushing the boundaries of what is possible” by developing novel applications for the “beautiful” iPhone operating systems. This sense of purpose over profit could be seen in the developer of a popular text adventure game who chose to release his app for free in order “to promote the enjoyment and development of interactive fiction.” Despite the fact that the application was free, the developer devoted significant time to the project, spending “3-5 weeks, evenings and weekends developing new features or fixing bugs” once every few months. Economic returns clearly were not the driving motivation. That so many innovative individuals are motivated by the benefits of being part of a community, rather than substantial economic rewards, should not be surprising. Scientists, after all, are willing to pay for the privilege of being part of the scientific community (Stern, 2004). And many studies of user communities have shown the importance individuals attach to being part of an innovative collective (Bagozzi & Dholakia, 2006; Franke & Shah, 2003; Jeppesen & Frederiksen, 2006). These motivations do not apply to all individuals, many were happy to commercialize their work, but the overall role of demand in determining commercialization decisions was weak, at best.

24

Generally, expected economic profit did not predict who would commercialize their innovations, while non-economic utility from community membership was significant. For individuals associated with communities, commercialization decisions are driven by very different motives than are usually assumed. It also suggests that, if, economic profit is not the only driver of commercialization for individuals, many potentially valuable innovations and ideas may never be released into the wider economy, since their creators are not driven purely by economic incentives. To the extent that commercialization decisions do vary from those predicted by pure economic incentives, it would suggest that a key part of the process by which innovations are commercialized and new companies are born does not operate as expected. There are some limitations to this research. First, the community around the iPhone may not be representative of innovators generally. However, while it is certain that there are idiosyncrasies around the iPhone development community, there is little reason to suspect it differs from similar groups of challenge-driven tinkerers and innovators, including scientists and software developers. Another limitation is that the data is somewhat constrained by the fact that much of the jailbreak community operates with pseudonyms and some degree of secrecy. However, where the data was backed up by interviews, there have been no substantial errors found. Finally, it is possible that individuals may have anticipated the launch of the App Store and created free products to create demand prior to charging for them. This view is not supported by the interviews that I conducted, which indicated that the nature of the App Store was a surprise, and that individuals did not anticipate the market, but this cannot be ruled out. Further, there was no indication of any two-sided markets for free applications in the analysis – free applications were not launched as vehicles for advertising or demonstration versions.

25

Conclusion Using the launch of the Apple App Store as a natural experiment allows us to examine the motivations behind commercialization. Out of the developers who continued to update software after the launch of the App Store, the majority chose to make their work commercially available. Many individuals, however, chose not to commercialize. Contrary to what a rational economic approach to opportunity exploitation might suggest, the potential demand for a product did not predict commercialization. Individuals decisions are driven by utility beyond profit maximization, which can lead to seemingly irrational choices about whether to commercialize (Scott Morton & Podolny, 2002). Further research is needed to understand when rational economic assessment is more influential, but these finding suggest that motivations for commercialization are more complicated than often assumed. It takes more than economic attractiveness to drive commercialization, and understanding what influences the will be an important factor in the study of entrepreneurship and the spread of new innovations.

26

References Al-Zarouni, M., & Al-Hajri, H. (2007). Taxonomy of iPhone Activation and SIM Unlocking Methods. Information Security Management Conference (p. 157). Alvarez, S. (2007). Discovery and creation: alternative theories of entrepreneurial action. Strategic Entrepreneurship Journal, 1-46. Amit, R., Muller, E., & Cockburn, I. (1995). Opportunity costs and entrepreneurial activity. Journal of Business Venturing, 10, 95-106. Bagozzi, R., & Dholakia, U. (2006). Open source software user communities: A study of participation in Linux user groups. Management Science, 52(7), 1099. Brockhaus, R., & Horwitz, P. (2002). The psychology of the entrepreneur. In N. Krueger (Ed.), Entrepreneurship: Critical Perspectives on Business and Management (Vol. 2). Routledge. Brockner, J. (1992). The escalation of commitment to a failing course of action: Toward theoretical progress. Academy of Management Review, 17(1), 39–61. JSTOR. Cardon, M., Wincent, J., Singh, J., & Drnovsek, M. (2009). The nature and experience of entrepreneurial passion. The Academy of Management Journal, 34(3), 511-532. Chen, B. (2009, November). Jailbreakers Battle Apple for Control of iPhone. Wired. Choi, Y., & Shepherd, D. (2004). Entrepreneurs’ Decisions to Exploit Opportunities. Journal of Management, 30(3), 377-395. doi:10.1016/j.jm.2003.04.002 Edery, D., & Mollick, E. (2009). Changing the Game: How Video Games are Transforming the Future of Business. Financial Times Press. Farmer, S. M., Yao, X., & Kung-Mcintyre, K. (2011). The behavioral impact of entrepreneur identity aspiration and prior entrepreneurial experience. Entrepreneurship Theory and Practice, 35(2), 245– 273. Wiley Online Library. Fauchart, E., & Gruber, M. (2011). Darwinians, Communitarians, and Missionaries: The Role of Founder Identity in Entrepreneurship. The Academy of Management Review, 54(3). Franke, N., & Shah, S. (2003). How communities support innovative activities: an exploration of assistance and sharing among end-users. Research Policy, 32(1), 157–178. Elsevier. Gans, J., Hsu, D., & Stern, S. (2002). When does start-up innovation spur the gale of creative destruction? RAND Journal of Economics, 33(4), 571-586. Hannan, M., & Freeman, J. (1993). Organizational Ecology. Cambridge, MA: Harvard University Press. Harhoff, D., Henkel, J., & Von Hippel, E. (2003). Profiting from voluntary information spillovers: how users benefit by freely revealing their innovations. Research Policy. Hayek, F. (1945). The use of knowledge in society. The American Economic Review. Haynie, J. M., Shepherd, D., & McMullen, J. S. (2009). An Opportunity for Me? The Role of Resources in Opportunity Evaluation Decisions. Journal of Management Studies, 46(3), 337-361. doi:10.1111/j.1467-6486.2009.00824.x Hayward, M. L. a., Shepherd, D., & Griffin, D. (2006). A Hubris Theory of Entrepreneurship. Management Science, 52(2), 160-172. doi:10.1287/mnsc.1050.0483 Hienerth, C. (2006). The commercialization of user innovations: the development of the rodeo kayak industry. R&D Management, 36(3), 273–294. Wiley Online Library. Himanen, P., Torvalds, L., & Castells, M. (2002). The Hacker Ethic (p. 256). Random House. Hmieleski, K. M., & Baron, R. A. (2009). Entrepreneurs’ optimism and new venture performance: A social cognitive perspective. Academy of Management Journal, 52(3), 473–488. Academy of Management. Hurst, E., & Pugsley, B. (2010). The Non Pecuniary Benefits of Small Business Ownership. University of Chicago, working paper. Jeppesen, L. B., & Frederiksen, L. (2006). Why do users contribute to firm-hosted user communities? The case of computer-controlled music instruments. Organization Science, 17(1), 45. INFORMS.

27

Jurutka, B. (2009). Tapulous’s Tap Tap Revenge Has Been Downloaded by 1 out of 3 iTunes Application Users. Comscore. Reston, VA. Kirzner, I. (1997). Entrepreneurial discovery and the competitive market process: An Austrian approach. Journal of economic Literature, 35(1), 60-85. Krazit, T. (2007). Apple: 250,000 iPhones bought to Unlock. CNet News. Lakhani, K., & Wolf, R. G. (2005). Why hackers do what they do: Understanding motivation and effort in free/open source software projects. In J. Feller (Ed.), Perspectives on Free and Open Source Software. Cambridge, MA. Lepak, D. P., Smith, K. G., & Taylor, M. S. (2007). Introduction to special topic forum: Value creation and value capture: a multilevel perspective. The Academy of Management Review, 32(1), 180–194. Academy of Management. McMullen, J.S., Shepherd, D., & Shepherd, A. (2006). Entrepreneurial action and the role of uncertainty in the theory of the entrepreneur. Academy of Management Review, 31(1), 132. Academy of Management. Meyer, P. (2007). Network of Tinkerers: A Model of Open-Source Technology Innovation. BLS Working Paper No. 413. Mollick, E. (2005). Tapping into the underground. MIT Sloan management review, 46(4), 21. Raymond, E. S. (2001). The cathedral and the bazaar: musings on Linux and open source by an accidental revolutionary. O’Reilly & Associates, Inc. Sebastopol, CA, USA. Reynolds, P. D. (1997). Who starts new firms?–Preliminary explorations of firms-in-gestation. Small Business Economics, 9(5), 449–462. Springer. Ruef, M. (2005). Origins of organizations: the entrepreneurial process. Research in the Sociology of Work. Schumpeter, J. (1934). Theory of Economic Development. Harvard University Press. Schumpeter, J. (1942). Capitalism, Socialism and Democracy. Economica (Vol. 11, p. 40). Harper. doi:10.2307/2549943 Scott Morton, F. M. S., & Podolny, J. M. (2002). Love or Money? The Effects of Owner Motivation in the California Wine Industry. Journal of Industrial Economics, 50(4), 431-456. Blackwell Publishing Limited. doi:10.1111/1467-6451.00185 Shah, S. (2006). Motivation, governance, and the viability of hybrid forms in open source software development. Management Science. Shah, S., & Tripsas, M. (2007). The accidental entrepreneur: The emergent and collective process of user entrepreneurship. Strategic Entrepreneurship Journal, 1(1), 123–140. Wiley Online Library. Shane, S. (2003). A general theory of entrepreneurship: the individual-opportunity nexus (p. 327; 9781843769965). Edward Elgar Publishing. Shane, S., & Venkataraman, S. (2000). The Promise of Entrepreneurship as a Field of Research. Academy of Management Review. Shane, S., Locke, E., & Collins, C. (2003). Entrepreneurial motivation. Human resource management review, 13, 257-279. Short, J. C., Ketchen, D. J., Shook, C. L., & Ireland, R. D. (2009). The Concept of “Opportunity” in Entrepreneurship Research: Past Accomplishments and Future Challenges. Journal of Management, 36(1), 40-65. doi:10.1177/0149206309342746 Sine, W. D., & Lee, B. H. (2009). Tilting at windmills? The environmental movement and the emergence of the US wind energy sector. Administrative Science Quarterly, 54(1), 123–155. Johnson Graduate School. Staw, B. M. (1981). The escalation of commitment to a course of action. Academy of management Review, 6(4), 577–587. JSTOR. Stern, S. (2004). Do scientists pay to be scientists? Management Science. National Bureau of Economic Research Cambridge, Mass., USA.

28

Stuart, T., & Ding, W. (2006). When Do Scientists Become Entrepreneurs? The Social Structural Antecedents of Commercial Activity in Academic Life Sciences. American Journal of Sociology. Thornton, P. H. (1999). The sociology of entrepreneurship. Annual Review of Sociology, 25(1), 19–46. Von Hippel, E. (1988). The sources of innovation (Vol. 132). Oxford University Press New York. Von Hippel, E. (2005). Democratizing innovation. the MIT Press. Von Hippel, E., & Von Krogh, G. (2003). Open source software and the“ private-collective” innovation model: Issues for organization science. Organization Science, 14(2), 209–223. JSTOR. Wasserman, N. (2006). Rich Versus King: The Entrepreneur’s Dilemma. Academy of Management.

29

Figure 1: Outcome tree for jailbroken software Charge for app

Jailbroken app developed

Commercial

Noncommercial

Release for charge on App Store

Commercial

Release for free on Cydia

Noncommercial

Release for charge on Cydia

Commercial

Update app

Free app

Do not update app

Release of App Store

30

Release for free on App Store

Abandon

Figure 2: Results for all application developers

Charge for app

Commercial

N=7

N=7

Jailbroken app developed

Update app

N=98

N=58

Release for free on App Store

Noncommercial

Release for charge on App Store

Commercial

Release for free only on Cydia

Noncommercial

Release for charge only on Cydia

Commercial

Free app N=91*

N=9

N=38

N=10

N=1

Do not update app

Abandon

N=33

N=33

Variables recorded before the App Store launch:

tab Variables recorded after App Store launch:

Name of developer, number of apps, demand (downloads), open source commitment, number of app versions, app category, and whether donations were requested.

Outcome (commercialization, noncommercialization, abandonment) and whether software was banned by Apple

Release of App Store * Includes 3 developers not included in the analysis because of unrecorded download numbers.

31

Table 1. Summary Statistics by Developer

VARIABLES Observations N Apps Versions logDemand Opensource Games Banned Donationware

32

Abandoned

Commercial

mean (sd)

mean (sd)

Non Commercial mean (sd)

32 1.188 (0.471) 3.609 (4.130) 5.464 (0.436) 0.563 (0.504) 0.0469 (0.195) 0.0938 (0.296) 0.203 (0.399)

38 2.500 (4.329) 5.509 (5.067) 5.573 (0.401) 0.259 (0.415) 0.184 (0.393) 0.158 (0.370) 0.213 (0.405)

18 1.500 (1.425) 3.008 (2.995) 5.341 (0.604) 0.528 (0.499) 0 (0) 0.389 (0.502) 0.0833 (0.257)

Table 2. Correlation Table by Developer

Commercial Commercial Apps Versions lDemand Opensource Game Ban Donation

33

Apps

1.0000 0.0729 1.0000 0.0539 -0.0187

Versions

lDemand

Opensource

Game

Ban

Donation

1.0000

0.0219 0.0691 0.2410 -0.0971 -0.1405 0.0695 -0.0215 -0.0615 0.2333 0.2531 0.2510 -0.0223 -0.0820 -0.0454 0.1693

1.0000 0.1056 0.1345 0.1206 0.0780

1.0000 -0.1180 1.000 0.1084 -0.135 1.0000 -0.0783 0.217 -0.115

1.0000

Table 3: Results of Multiple Logistic Regression by Developer

N Apps Versions logDemand Opensource Games Banned Donationware Constant

Commercia l 2.138* (0.863) 1.144* (0.0825) 1.110 (0.694) 0.222** (0.136) 5.382 (5.936) 1.444 (1.295) 0.322 (0.288) 0.224 (0.761)

88 37.35 0.000 0.201

N chi2 p Psuedo R2 Relative risk ratios shown z-statistics in parentheses *** p