Paper to be presented at the 25th Celebration Conference 2008 on
ENTREPRENEURSHIP AND INNOVATION - ORGANIZATIONS, INSTITUTIONS, SYSTEMS AND REGIONS Copenhagen, CBS, Denmark, June 17 - 20, 2008
USER ENTREPRENEURSHIP IN ONLINE COMMUNITIES: LEAD USER CHARACTERISTICS, AGENDA SHAPING AND SOCIAL STANDING Lars Frederiksen I&E Group, Imperial College London
[email protected] Linus Dahlander I&E Group, Imperial College London
[email protected] Erkko Autio I&E Group, Imperial College London
[email protected]
Abstract: In this paper we develop and validate a model that explains entrepreneurial intent and activity of individual product users. These users entrepreneurial intent and activity are influenced by their personal characteristics (i.e. lead user qualities) and their position in a social network (i.e. agenda-shaping activities and social standing). Beyond employing survey data the conceptual model is validated by using unique data from the social network of an online user community covering all communication within the community over its eight-year life span. We find that an individual s lead user characteristics and agenda-shaping activities affect her entrepreneurial intent. Entrepreneurial intent in turn mediates the effect of an individual s lead user characteristics and agendashaping activities on entrepreneurial activity. Also, we demonstrate that an individual s social standing in the community impacts her entrepreneurial activity. Our model suggests that online user communities not only construct the knowledge concerning new entrepreneurial opportunities, they also support users in opportunity creation and shape the entrepreneurial roles required for their pursuit. The paper contributes new insights to: 1) the entrepreneurship literature emphasising that a different organisational setting from firms and universities, supports individuals to become entrepreneurs and 2) the growing literature on user innovation by emphasising how users not only innovate and how that is important for many established firms but occasionally also users develop their own firms to appropriate value from their innovations.
JEL - codes: M13, O32, O31
User Entrepreneurship in Online Communities: Lead User Characteristics, Agenda Shaping and Social Standing
Word count: ABSTRACT:
1
In this paper we develop and validate a model that explains entrepreneurial intent and activity of individual product users. These users’ entrepreneurial intent and activity are influenced by their personal characteristics (i.e. lead user qualities) and their position in a social network (i.e. agenda-shaping activities and social standing). Beyond employing survey data the conceptual model is validated by using archive data from the social network of an online user community covering all communication within the community over its eight-year life span. We find that an individual’s lead user characteristics and agenda-shaping activities affect her entrepreneurial intent. Entrepreneurial intent in turn mediates the effect of an individual’s lead user characteristics and agenda-shaping activities on entrepreneurial activity. Also, we demonstrate that an individual’s social standing in the community impacts her entrepreneurial activity. Our model suggests that online user communities not only construct the knowledge concerning new entrepreneurial opportunities, they also support users in opportunity creation and shape the entrepreneurial roles required for their pursuit. The paper contributes fresh insights to: 1) the entrepreneurship literature emphasising that a different organisational setting from firms and universities, supports individuals to become entrepreneurs and 2) the growing literature on user innovation by emphasising how users not only innovate and how that is increasingly important for many established firms but occasionally also users become entrepreneurs and create their own firms to appropriate value from their innovations.
2
INTRODUCTION
“I gave away sound fills and other products for free, and so it was a matter of having fun and sharing knowledge with others in the community…everyone was trying to find out new ways to create new features and sounds…I noticed a lot of users in the community knew my name, I got kicked out from some servers because of overload on my user account. So, I felt, yeah, that there could be a business in this” (User entrepreneur and lead user in the Propellerhead online community, Dec. 2007)
The role of individual users in innovation has long since been recognised (von Hippel, 1988). Often, product users are the first to experience new needs. If the needs are pressing enough, users skilled enough, and the market is unable to supply a product or service capable of satisfying the need, ‘lead users’ (i.e., those individuals who are sensing the need most acutely and standing to benefit the most if a solution is developed) may take it upon themselves to create solutions to solve their problems (Morrison, Roberts, & Midgley, 2004; von Hippel, 1988). Lead users tend to dominate innovative activity particularly when their potential gains from innovation exceed those of manufacturers and suppliers, as is often the case in very small market niches or in emergent markets (Baldwin, Hienerth & von Hippel, 2006; Shah & Tripsas, 2007; von Hippel, 1988). As the market gains momentum and economies of scale emerges, established manufactures tend to step in, taking over from lead-user innovators. Innovative activity by lead users is widely documented in a variety of sectors, e.g., scientific instruments and electronics (von Hippel, 1988); sports equipment (Franke & Shah, 2003; Hienerth, 2006; Luthje, Herstatt, & von Hippel, 2005); specialised software (Jeppesen & Frederiksen, 2006); and juvenile products (Shah et al., 2007). While the importance of users in innovation has been well established in the literature, research and theorising on the role of users as entrepreneurs is of more recent origin (Baldwin et al., 2006; Hienerth, 2006; Shah et al., 2007). Some individuals not only 3
innovate – they sometimes, for various reasons, act as the pioneering manufacturers of their own innovations, thereby paving the way for new market creation and firm formation. However, while anecdotal evidence suggests that the phenomenon is important (Shah et al., 2007), little is known about the dynamics of this phenomenon. Although researchers have studied lead user innovation since the 1980s, research on user entrepreneurship is the product of the current decade and remains largely descriptive and theoretical, as researchers attempt to gauge the scale and dynamics of the phenomenon at the level of enterprising individual. To date, there exists little theory-grounded research that would seek to empirically validate the micro-process mechanisms that cause some individuals to become user entrepreneurs. This is the problem that this study addresses. An important motivation for the recent interest on user entrepreneurship has been provided by the emergence of physical and online user communities, often facilitated by the Internet (Dahlander & Wallin, 2006; Shah, 2003). User communities provide an important facilitating mechanism for social interaction and learning, thereby enabling collective experimentation, information sharing and feedback (Jeppesen & Frederiksen, 2006). These have all been identified as important influences on user innovation and entrepreneurship (Shah et al., 2007). As an important, hitherto less recognised mechanism also online user communities facilitate the construction of online identities and online social status and prestige (Himanen, 2001). As individuals interact in a non-restricted public forum, they build reputations and network relationships that enable them both to discover and assess opportunities, as well as to access and mobilise resources for their pursuit. Active and high-profile members of the user community may due to higher levels of interaction get more alert and thus see more opportunities than others, and their visibility and insider status enjoyed within the community may furthermore enable them to actively create and shape opportunities and thereby facilitate the adoption of their
4
inventions (Brown & Duguid, 1991; Lave & Wenger, 1991). The legitimacy conferred upon some members of the online community may endow these with agenda-shaping abilities, such as the ability to influence opinions and the ability to help legitimise novel products and designs. Such agenda-shaping activity is underlined in the community of practice literature as inherited power structures of controlling dialogue among community members (Brown & Duguid, 2001). Such agenda-shaping individuals have also been identified taking roles as external gatekeepers and internal stars by Tushman & Scalan (1981). Community embedded opinion leaders would not only enjoy privileged access to resources and information, but also, possess power to influence development, experimentation and social learning agendas toward desired directions. As a consequence, a high-visibility status would endow high-visibility individuals with stronger perceptions of entrepreneurial self-efficacy, thereby facilitating entrepreneurial intentions and actions (Eckhardt & Shane, 2003). In short, we suggest that online communities not only assist experimentation and social learning concerning user-initiated products, they may also help construct the social identities required for resource mobilisation, agenda shaping and for the commercialisation of opportunities thus constructed and discovered. In this study, we use online community data to study the effects of an individual’s lead user characteristics, agenda-shaping and social standing on her perception of opportunity, entrepreneurial intentions and entrepreneurial behaviour. Rather than explaining user entrepreneurship as rational behaviour based on economic calculations, we examine the influence of an individual’s social environment as an external contingency that constructs role definitions and related expectations and regulates individuals’ access to information and resources. We draw on social network, opportunity recognition and resource dependence theories to build and test a model which suggests that not only is an individual’s innovator role an important determinant of her recognition
5
of entrepreneurial opportunities, but also, her social standing and influence within her community exercise an important influence on her perceptions of entrepreneurial selfefficacy, and hence, on her entrepreneurial behaviours. Because of her network position, the individual gets variably exposed to opportunities. Because of her social standing within the network, she is variably likely to pursue opportunities perceived. Using unique data that covers the life-span of the Swedish-hosted digital music software community Propellerhead, we build and test a model that explicates the effect of an individual’s social standing within a user-developer community on her entrepreneurial intent and entrepreneurial actions. By so doing we seek to advance the understanding of userentrepreneurship processes beyond today’s state-of-the-art. Even though there is increasing data documenting the extent of the user entrepreneurship phenomenon in a number of sectors (e.g., Franke et al., 2003; e.g., Hienerth, 2006; Luthje et al., 2005), as well as increasing theorising concerning the external conditions within which user entrepreneurial activity is more likely to occur (Baldwin et al., 2006; Shah et al., 2007), there have been no studies to examine how individual perceptions are balanced with social processes within a given user community to influence individual entrepreneurial propensities. In summary, we seek four distinctive contributions in our study. First, we add to an emerging research stream on user entrepreneurship by developing and validating a model that articulates the effect of user’s social position and status on her entrepreneurial intents and behaviours. We hereby contribute by suggesting a more fine-grained conceptualisation and operationalisation of the user entrepreneurship notion as we analyse the affect of various explanatory variables on both entrepreneurial intention and entrepreneurial activity. Second, by so doing, we extend the user entrepreneurship research agenda beyond descriptive and theoretical studies toward theory testing and
6
validation. Third, our data provides a rare opportunity to examine how social identities constructed in online communities accord some individuals greater freedom and motivation to pursue entrepreneurial opportunities, thereby lending support to the notion that individual entrepreneurial actions are not only determined by innate and given individual characteristics, but also, by her evolving social standing. Finally, we demonstrate how user communities not only provide spaces for experimentation and social learning, but also, for the social construction of roles required to commercially exploit opportunities discovered.
THEORY Emerging Research on User Entrepreneurship Even though entrepreneurship research has advanced well beyond the simplistic notions of entrepreneurship as an innate behavioural orientation that individuals are born with (Gartner, 1988), and there is increasing research on social and contextual influences on entrepreneurial activity (Aldrich & Fiol, 1994; Chiasson & Saunders, 2005; Fletcher, 2006; Jack & Anderson, 2002; Sarason, Dean, & Dillard, 2006), empirical studies remain conspicuously focused on the individual. By studying the effect of an individual’s social context on her entrepreneurial intent and activity, we hope to provide for a more socialised view of the entrepreneurial process.. We define user entrepreneurship as individual-level activity by which innovating users move beyond product or service invention to manufacturing and the establishment of new organisations to capture value created with their inventive activity (Baldwin et al., 2006; Shah et al., 2007). User entrepreneurship is different from user innovation, where users construct prototypes to satisfy their specific needs and even freely reveal and share their prototypes and ideas
7
with like-minded individuals but do not seek to scale up this activity by offering products and services to others on a commercial basis. I was a Propellerhead user so I spend a lot of time in their community and then I started thinking about the software, how there were limited possibilities of using it for live sessions. I started asking other users if they had the same problems that I had experienced or already had found a solution to these problems. When I realised there was not a solution to solve these problems, I decided on developing this extra software…. It later provided the basis for me starting my firm” (User entrepreneur and lead user in the Propellerhead online community, Dec. 2007) At some point in life most people have experienced highly specific needs for a given product or service. Often that need is met by an existing, perhaps slightly customised product or service, but in a few cases, when an appropriate substitute is not available in the market, the individual goes on to create a unique solution to her special need by herself. This inventive activity is known as the ‘Do it yourself’ phenomenon (Gelber, 1997). Occasionally such inventors may be prompted to share their invention with other individuals sharing the same specific need or interest. If the interest is widespread enough, established manufacturers may discover a promising new niche and start servicing it (Von Hippel, 1988). In some instances, however, the user innovator may eventually start manufacturing the developed solution herself, thereby becoming a ‘user-manufacturer’ in the terminology of Baldwin et al. (2006). Research has shown that user entrepreneurs are often embedded in communities of like-minded individuals who share both a passion for given products and technologies as well as engage in similar activities and practices (Jeppesen et al., 2006; Luthje et al., 2005; Shah, 2003; Shah et al., 2007). Also a degree of social capital exists in such communities of individuals engaged with similar practices or products (Coleman, 1988, Wellman et al.,1996). “The community has always been very positive about our work, which as I said before all started with the user developed mods we did….If we say that our GUI design started from the possibilities given in the community by the fact that mods were around within this specific community, I would say, that our firm wouldn’t 8
have been today, or even may not have existed at all if the online community hadn’t been there” (User entrepreneur and lead user in the Propellerhead online community, April 2008)
“It was not so much having the community as such helping us in establishing the firm….it was more making the right contacts within the community” (User entrepreneur and lead user in the Propellerhead online community, April 2008)
Communities of practitioners are important because they provide a platform for social learning and experimentation (Bandura, 1977; Brown et al., 1991; Lave et al., 1991) and facilitate experience sharing and feedback for innovative activity. Such communities may be physical (as, e.g., car hobbyist fairs gathering around a certain brand) (Muniz and O’Guinn 2001) and virtual (as, e.g., software user and developer communities). Webhosted communities, while lacking in physical interaction, may offer advantages in knowledge sharing, as discussion threads can be traced back over long periods of time, and tools are available for quick distribution of user insights and experiences (Dahlander et al., 2006). This transparent proliferation of user experiences and information can provide well-embedded individuals with greater access to new ideas and opportunities (Shah, 2003). As one of our interviews for scoping the study revealed; web-facilitated user communities may provide an important platform for user entrepreneurship. “We know that some users like Peff, Peter Tools, Marco R and Ed Baumann [lead users in the community] have used their involvement in our online community as a launching path for going into business. […] and that is very positive for us….They [users in the online community] may see business opportunities by being involved with similar users. […] they sometimes use our communities to recruit beta testers” (CEO, Propellerhead, Dec. 2007)
Recent theoretical and empirical work has begun to explore the contingencies under which user entrepreneurship is more likely to occur (Shah & Tripsas, 2007). Shah and Tripsas focused on external contingencies that influence the user entrepreneurship process. In their post-hoc model of the user entrepreneurship process, users are gradually
9
drawn to entrepreneurship by the positive feedback received from the user community for their inventions. The community provides valuable feedback and guidance and may even propose amendments to the invention (Franke et al., 2003; Lakhani & von Hippel, 2003). This was documented by our interviews with user entrepreneurs: “I definitely used the online community to develop my original product idea as well as my firm. I use the community to market my software as it complements the Propellerhead Reason product [the key product of the firm that hosts the community]. Two times I got really qualified beta-testers involved through the Propellerhead community” (User entrepreneur and lead user in the Propellerhead online community, Dec. 2007)
By virtue of becoming an insider-member of the user community, the innovating user is able to benefit from information asymmetries that arise from information stickiness and selective disclosure. The privileged access to user feedback, combined with user’s deep knowledge of the specific need, increases the potential for breakthrough innovations. “First I got some ideas from the Propellarhead community and then I asked the community to help me beta-test my product as well as used the community for marketing” (User entrepreneur and lead user in the Propellerhead online community, Feb. 2008)
Low opportunity costs enable user entrepreneurs to overcome resource shortages and devote time to inventive activity. Fragmented markets offer small sheltered niches to work in without attracting the interest of incumbents. High uncertainty about market developments enables socially embedded user entrepreneurs to capitalise on information asymmetries (Lave et al., 1991). Such conditions are met, for example, in many extreme sports communities where typically a relatively small and socially cohesive group of enthusiasts regularly gather to practice their hobby, show off new tricks and share experiences and tips on how to perform even more audacious feats. Such a community provides an ideal setting for user innovation and entrepreneurial activity, as the social and intrinsic incentives for experimentation with novel equipment are high, opportunity costs
10
often low and the niches too small for mainstream manufacturers to get interested in (Luthje et al., 2005; Shah, 2003). As another example, Shah and Tripsas (2007) describe how motherhood and the rearing of small children provides a strong glue for social bonding, which often prompts individuals to share tips and experiences, sometimes in the form of specialist devices and equipment designed to address special needs, such as the need for specially designed child seats for underweight children.1 In a related study, Baldwin et al (2006) studied the path through which user inventions are transformed into commercial products. In their model, users’ innovative activity begins when a new ‘design space’ is opened. If the opportunity cost is low, users keep inventing until ‘user-purchasers’ emerge. This may prompt some user-innovations to become usermanufacturers. As the niche gains momentum, opportunities for serial manufacturing increase. At the same time, the design space gets mined out, and the creation of new radical improvements grows increasingly difficult. This causes the rate of innovation to decrease, making investment in manufacturing methods more feasible. Eventually, the conditions may become ripe for established manufacturers to muscle their way into the market, leaving user-manufacturers the choice of either withdrawing or specialising into increasingly narrow niches. Both the Shah and Tripsas as well as the Baldwin et al models are concerned with the question of when and under which conditions user entrepreneurship is more likely to occur, thus emphasising field-level contingencies. While useful for their purposes, little is known about the process of user-entrepreneurial gestation itself. None of the existing theoretical models of user entrepreneurship address this critical aspect. Clearly, not all extreme sports enthusiasts start their own firms, and not all parents start manufacturing whatever specialist devices they may have invented for their offspring, even if they had 1
It is a strong testimony of the draw of user inventive activity that one of the major patent categories of patent agencies is ‘devices facilitating everyday life’.
11
received positive feedback from peers. Even though most enthusiasts in the rodeo kayaking industry would be able to access the specialist-insider knowledge concerning latest trends and fads, and even if many do tinker with their kayaks, only a small percentage of these enthusiasts are converted into manufacturers (Hienerth, 2006). As we see it, much of the recent work on user entrepreneurs has focused on those individuals who have started manufacturing their inventions, thereby ignoring the process of selection that sorts out user entrepreneurs from specialist users. Existing post-hoc process models, while useful in explaining the process that selected user-entrepreneurs may pass through while ‘transmogrifying’ from an enthusiastic hobbyist into an enthusiastic manufacturer, are silent about the determinants that trigger such transition. Since information concerning recent trends and discoveries often flows freely within such communities and is simultaneously sticky across communities (Brown & Duguid, 2001), membership within such a community opens up information asymmetries for potential user entrepreneurs to exploit. However, because of information overflow and scarcity of attention (Hansen & Haas, 2001), not all members of the community enjoy equal access to the information that flows within the community. A user entrepreneur notes: “…yeah yeah I used the community to get my name mentioned. Many people were interested to hear what I was doing” (User entrepreneur and lead user in the Propellerhead online community, Feb. 2008)
Some individuals occupy more visible positions than others within the community and are therefore able to see more potentially valuable information (Lave et al., 1991). For example, queries posted by high-status individuals in communities would attract more responses and feedback than queries posted by low-status individuals. Greater feedback would not only offer comparative information advantages relative to other members of the community, it would also confer a sense of influence, thereby enhancing perceptions of 12
self-efficacy in terms of, e.g., resource mobilisation (Jack et al., 2002). Finally, we suggest that in an exploration-focused user community, where topical issues constantly arise from, and are shaped by, interactions within the community, agenda-shaping activities will be an important determinant of user entrepreneurship (Aldrich et al., 1994; Garud, Lant, & Schildt, 2007; Hunt & Aldrich, 1998). Agenda-shaping activities, we suggest, enable some individuals to influence the collective choice between alternative developments, mobilise momentum behind their own ideas, and hence, drive the shaping of opportunities within the community (Hung, 2004; Sarason et al., 2006). The propensity of an individual to become an user-entrepreneur is the result of an evolving dialogue between the individual and her social context, one which structures both opportunities and entrepreneurial roles. User entrepreneurs are thus not predetermined, but rather, constructed by user communities (Burton, Sorensen, & Beckman, 2002; Dobrev & Barnett, 2005; 1986) Based on these arguments, we create the conceptual model illustrated in Figure 1. The model provides a framework for developing the hypotheses.
---------------------------------------------------------------INSERT FIGURE 1 ABOUT HERE ---------------------------------------------------------------HYPOTHESES Lead User Characteristics and Entrepreneurial Intent “Before the firm was actually established, so, I was a power user as they say, not really a businessman” (User entrepreneur and lead user in the Propellerhead online community, Dec. 2007)
13
We define lead users as individuals of a given product or service who combine two characteristics: (a) they expect invention-related benefits from a solution and are thereby motivated to invent; and (b) they experience the need for a given invention earlier than the majority of the target market (von Hippel 1986). Lead users, whose needs foreshadow market trends, who possess specific product knowledge, and who would benefit significantly from product innovation, are often active in developing product modifications, add-ons, and sometimes completely new product categories (Franke et al., 2003; Jeppesen et al., 2006; Morrison et al., 2004; Morrison, Roberts, & Von Hippel, 2000; Von Hippel, 1988) Morrison et al. (2000) studied library software users and discovered that users inventing add-ons or changes to the software scored high on lead user characteristics relative to other users in the same community, with the impact of characteristics being moderated by the ability of users to exploit their resources and those of the external environment. In a subsequent study, Morrison (2004) developed the measurement of the lead user concept and showed it to be positively associated with innovation adoption. Franke and Shah (2003) demonstrated that inventors exhibit lead user characteristics more strongly than non-inventors. We suggest that lead user characteristics will also predict entrepreneurial intent. Because lead users perceive needs earlier than other users, they will have a clearer vision of where the market might be heading. Based on their own experience, lead users are more likely than other users to see clearly the value creation potential of their inventive activities. Because of their success in product invention, they will have confidence that they can commercially launch their products for sale to others. While the lead user construct as mentioned has been employed as a measure of the skill level as well as position relative to the average users in terms of being ahead of the
14
market in terms of needs and ideas, this appears to us as a somewhat inaccurate interpretation of the lead user construct. Since users in the widely used Morrison et al. (2000) scale are asked to self-assess their position and skills, lead user characteristics as determined from the lead user scale reflect rather a self-perceived level of skills and position relative to the rest of the user populations than the lead user’s objective level of skills and position. We conjecture that:
Hypothesis 1: Individuals, who perceive themselves as having lead user characteristics, will have greater entrepreneurial intent than other users in an online user community.
Agenda-Shaping Activity and Entrepreneurial Intent Online user communities provide useful social networks for joint action. This social collective will define the technological and economic opportunities available for members of the community (Orlikowski, 2002). The agendas of the community are negotiated in an ongoing interaction between users, as some opportunities are taken up and others ignored. Through evolving practice, the community’s knowledge base is constantly constituted and reconstituted through individual-level interactions. This process of creating alertness to a certain topic will determine both the existence, as well as the attractiveness of opportunities thus constructed. Depending on their network position and ability to shape the shared agenda, members of the community will be differentially able to influence this dialogue. “So, the reason behind this post is to see what the interest would be. Would you purchase the disc? Would you tell your Reason (key product around with the community gather) cohorts to purchase one too? Let me know what you think…” (Part of post from the Propellerhead online community send by lead user and user entrepreneur, Feb. 2005)
15
Aldrich and Fiol (1994: 649) suggested that by influencing processes of social construction, entrepreneurs can develop new meanings and shape institutional norms. Institutional norms will then determine the feasibility, appropriateness and perceived value of alternative courses of action. They contended that entrepreneurs may shape institutional norms through both cognitive and socio-political strategies. Cognitive strategies for collective agenda shaping in the dialogue include, for example, the development of collective knowledge base via symbolic language and behaviours, as well as encouraging convergence around dominant designs. Socio-political strategies include, for example, the development of trust by maintaining internally coherent stories and narratives, as well as by promoting perceptions of reliability by mobilising collective action. These strategies suggest that the ability to influence the discourse within an online user community constitutes a powerful determinant of entrepreneurial opportunity (Phillips, Lawrence, & Hardy, 2004). “…and so I started asking people if they like the idea and how they would like this software to be made, what features they would like to find in this kind of product (User entrepreneur and lead user, Dec. 2007)
In online user communities, discourse takes place within online forums that host several discussion threads. A new thread will define a new issue in the forum, and therefore, open the potential for shaping or modifying the collective agenda. Opening new threads within an online community, therefore, is a key agenda-shaping activity. We propose:
Hypothesis 2: The greater agenda-shaping activities of a given individual within an online user community, as reflected in the number of new discussion threads opened by the individual, the stronger the entrepreneurial intent.
16
Entrepreneurial Intent and Entrepreneurial Action “We were reassured from the community response we were having that we would be relative successful when starting up our firm” (User entrepreneur and lead user, Dec. 2007) Consistent with theories on planned behaviour (Ajzen, 1991), we predict a mediating relationship between entrepreneurial intent and entrepreneurial action. Starting up a new firm is a major undertaking, which involves careful planning, significant financial and social commitments, and career-altering trade-offs (Bird, 1988; Shapero, 1982). Such decisions are not taken by accident and without at least some consideration of available options. These notions have prompted a significant stream of research on the determinants of entrepreneurial intentions in its own right (Boyd & Vozikis, 1994; Douglas & Shepherd, 2002; Krueger, Reilly, & Carsrud, 2000). Common to studies across this research stream is the notion that entrepreneurial intent precedes entrepreneurial action. We therefore expect entrepreneurial intent to mediate the effect of lead user characteristics and agenda shaping on entrepreneurial activity:
Hypothesis 3: Entrepreneurial intent will mediate the effects of an individual’s lead user characteristics and agenda-shaping activities on entrepreneurial activity.
Social Standing and Entrepreneurial Activity Even though entrepreneurship is typically portrayed as an individual act, it takes place within a social context and is therefore subject to social influences (Aldrich et al., 1994; Hite, 2005; Larson & Starr, 1993). A given individual’s social position, and hence, social identity and likely directions of her agency, is continuously evolving and subject to
17
ongoing social construction. Social roles and related role expectations are determined by, and embedded in, a constantly evolving network structure, including roles conducive to entrepreneurial agency (Downing, 2005). A given individual’s position in the wider social structure will inevitably constrain her agency, because different network positions may confer differential access to insider information and resources, and they may confer differential opportunities for influencing others (Granovetter, 1985). This social structure may be an online user community where users seek advice, solve problems, share experiences, develop new products and product versions as well as meet to socialize. This is exemplified by the statement from a user entrepreneur commenting on the role of the social in the online community: “I always felt that if I have this status (in the community) people will allow me to do this and will also like the idea that they, they love someone who…I, I, When I started the company a lot of people were saying, wow this is great, because they knew, knew me, they felt like, wow this guy is, is taking a major step…so, when I initially set up my company I wrote something on the community board and a lot of people wished me luck and started listening to my music and using my products. So, it helped.” (User entrepreneur and lead user, Feb. 2008) Individuals participate in online communities for a reason, for example, to seek advice, to solve problems and to socialise with people with similar interests and practices. In some cases we find that individuals see commercial opportunities as well as gather resources from this social setting and set up firms to benefit economically from their inventions developed through collective involvement or to benefit from their skills developed and sharpened by community participation by selling customized services. The collective base of the community serves as a structure for the demand side activities of entrepreneurship by constructing entrepreneurs’ reputations and providing a market for diffusing the product or service of the entrepreneur. Communities can therefore serve as a social structure that provides resources and opportunities for participating creative individuals (Hargadon & Bechky, 2006).
18
We propose that in online user communities, an individual’s social standing will be an important influence on her actions. In online user communities, users tend to identify with a shared mission and share similar values, group intentions and group identification (Bagozzi & Dholakia, 2006). For the online user community to develop shared activities, users need to be mutually responsive to each others’ actions; share a joint commitment to the online activity; and be committed to supporting others within the community (Tuomela, 2005). Greater display of such values and activities will confer greater visibility within the community and enhance contributing participant’s social standing within the community (Himanen, 2001). High-status individuals will be recognised by other members of the community, and their postings in the online medium will be more intensely followed, increasing the chances of responses and feedback. Such visibility, and concomitant privileged access to information, as well as influencing opportunities, will enhance high-status individuals’ ability to recognise new opportunities as well as to mobilise resources and actions for their pursuit. A user entrepreneur remarks: “Through the years here on the forum I’ve thrown out a lot of RNS files and other stuff to help folks with problems they’ve encountered or solutions to things they couldn’t figure out in Reason (the main product around which the Propellerhead community gather). These tricks and tipis files seemed to be valuable to a lot of people” (User entrepreneur and lead user Propellerhead Online community, April 2008)
In online user communities, conversation takes place in online forums. Users post messages identifying issues, signalling problems and proposing new courses of action. An individual’s social standing will be reflected in how others react to her postings. The higher the status of the individual, the more responses her postings would generate. Summarising, we propose:
19
Hypothesis 4: The greater an individual’s social standing in an online user community, as reflected in the number of responses received to her postings, the greater will be her entrepreneurial activity.
Finally, we expect an individual’s social standing to not only exercise a direct influence on entrepreneurial activity, but also, moderate the effect of entrepreneurial intent on entrepreneurial action. Whereas lead user characteristics and agenda-shaping activities endow the individual with qualities that enable her to perceive opportunities and thereby raise her entrepreneurial intent, the step from intent to commitment involves careful considerations of trade-offs and risks. The individual’s perceptions of self-efficacy are likely to play an important role in this stage (Ajzen, 1991). We suggest that a high social standing within an online community of like-minded individuals is likely to enhance such perceptions. A user entrepreneur argues about the association between setting up his firm and his role in the online community: “I’m absolutely sure that my special social position in the community was important, because it make things easier…They remember your name and they know your specialities, they yeah, wll read your post, how do you say, your advices and they will take note of your statements” (User entrepreneur and lead user, Feb. 2008)
A high social standing, as expressed in a high volume of responses to messages posted by an individual, will increase her confidence that she will be able to leverage the community for advice and for resource mobilisation. High-status individuals in an online community can use their status as a form of social capital and leverage it for entrepreneurial purposes (Renzulli & Aldrich, 2005). We therefore propose:
Hypothesis 5: An individual’s social standing in an online user community, as reflected in the number of responses received by her postings, will moderate the effect of entrepreneurial intent on entrepreneurial activity, such that at a high social standing the effect of entrepreneurial intent on entrepreneurial activity will be stronger.
20
METHOD Research Setting As we were interested in how an individual’s entrepreneurial intent and activity are shaped by community involvement, we focused on individuals embedded in a public unrestricted online community. The community embraces a large number of individuals that are mainly end-users of music software. The Swedish music software firm Propellerhead has hosted the community on their website since 2001 to invite users of their software to socialize and solve problems. This is a suitable research setting for several reasons. First, we could track the evolution of all interactions in the community from its inception in 2001 to the point when the survey was distributed. Second, because of some transparency in the software, particular individuals make customizations and introduce additional functionalities and features that occasionally are commercialised. Third, our interviews and investigation of secondary material showed that several individuals had started or were in the process of establishing a firm based on their engagement in the community. We were therefore able to test our hypotheses about how individuals use the community to make an entrepreneurial transition. The choice of using an online community as a platform for studying entrepreneurial intent and activity offers detailed ‘objective’ data about interaction patterns and community structure, which are often impossible to obtain in non-virtual communities. The individuals of the community are involved with creating, producing, processing and recording music through using the software. Individuals can use the community to ask questions and receive help from other individuals. Many individuals are using the community to foster the creative act of creating and arranging music by using the software in novel and different ways than originally intended. A growing number of individuals in the community are occupied with the technical development and design of 21
the software products themselves. Creative contributions from the individuals in the community span between producing artistic outputs as music to developing new pieces of software applications, design and content. Throughout its history, individuals in the community have voiced ideas for new areas of development for Propellerhead, discovered bugs, and suggested new product features and functionalities. Some individuals are even innovating in the community by developing new functionalities themselves (Jeppesen et al., 2006). Whereas individuals of the community have benefited greatly from having their need related problems solved and the sharing of user generated inventions, the main benefactor of economic gains from the problem-solving and inventive activities of the individuals of the community have been accrued by the community hosting firm Propellerhead. However, recently, individuals have begun to set up firms in order to appropriate benefits from their inventions as well as problem-solving services.
Data In this research we adopted a multi-method approach and triangulated between three types of data (Denzin 1978). Before initiating the quantitative study, we conducted 35 interviews with individuals in the community and employees at Propellerhead. The questions were geared at getting a contextual understanding about how the community works and the products of the firms. At the firm, we conducted interviews with the CEO, product developers and the community moderator. With individuals in the community, we did interviews with individuals that had become entrepreneurs based on their experience in the community. Interviews guided our quantitative data collection, and we also used interviews to assist us in interpreting quantitative results and check the face validity of our findings. 22
To test our hypotheses we assembled information from multiple sources. As a unique feature of the study, we collected data from the community’s message archives (web log) of all interactions in the community from its inception in 2001 to May 2007. We gathered the complete information including the sender of the message, timing, subject, message content, whether it was a first in a conversation thread and who the recipient was. Our dataset comprises some 280,000 messages. This dataset, the like of which has not been previously available for the study of user entrepreneurship, enables us not only to monitor an individual’s network position, but also, her social standing and legitimacy, as expressed in communication content and response volume and latency. Rather than solely relying on self-reported data, this dataset allowed us to get objective data on the individual’s involvement in the community since its inception. Table 1 shows the evolution of the community by illustrating the total number of active participants and the total number of messages by year.
---------------------------------------------------------------INSERT TABLE 1 ABOUT HERE ----------------------------------------------------------------
Independently of the web log, we collected data on individual characteristics on a sample of the individuals in the community using a survey instrument that we administered online at the community’s website and received 280 answers. The survey employed multi-item scales to measure a wide range of theoretical constructs. We pre-tested the survey with a small number of individuals in the community before distributing it. Following the suggestions by Armstrong and Overton (1977), we compared early and late respondents and their scores on the dependent and independent variables, and
23
found no significant differences. We discovered that those individuals responding to the survey are slightly more active than the general population in the community. This selfselection of being part of the survey may therefore bias the results. To control for this problem, we use the web log data to retrieve the number of posting of all individuals in the community and control for the self-selection of responding to the survey by using the Heckman procedure. By combining web log and survey data we avoid single-source bias (Doty & Glick, 1998; Podsakoff & Organ, 1986). Single-source bias becomes a problem when the observed association between variables is due to artifactual covariance, caused by social desirability bias (Podsakoff and Organ, 1986). Our dependent and some of the moderator variables were collected from the survey and the other measures we constructed from the web log. Using these data sources are also suitable given our interest in how entrepreneurial intent and entrepreneurial activity is linked to community involvement. We therefore had to get actual data of their activities in the community (from the web log) and information about their entrepreneurial activities (from the survey).
Variables Dependent variables Using the data from the survey instrument, we developed two dependent variables measuring entrepreneurial intent and entrepreneurial activity respectively. We asked the survey respondents to state whether they had been done a list of entrepreneurial decisions. We coded an individual with a dummy as having entrepreneurial intent if they say that they have ‘Made written plans to start a business’ based on their experience in the Propellerhead community. Entrepreneurial activity was captured with a dummy if an individual report that they ‘Sent invoices for products or services offered by your 24
company’. We derived alternatives measures for the dependent variables using different items in the survey. Our results are robust for using other definitions of entrepreneurial intent and entrepreneurial activity. 2
Independent Variables We based our scale on lead user characteristics from the work of Morrison et al (Morrison et al., 2004). Because we perceived methodological issues with that scale, we extended their scale and used eight items in the online survey to make a more comprehensive lead user construct. This variable had a Cronbach alpha of 0.75. As noted above, opening new discussion threads is a powerful means of influencing the collective agenda of an online user community. We therefore measured an individual’s ability to agenda shaping in the community by measuring the total number of new threads started by the individual divided with overall activity of the individual. This measure thus reflects an individual’s ability to initiate new discussions that have not been discussed before in the community. Following our argument above, we assessed an individual’s social standing by the average number of responses when posting a new thread. We reasoned that individuals who are recognized by other community participants are likely to receive more attention. The study by Hansen and Haas (2001) echoed Simon’s notion that attention is a scarce resource in instances with over-abundance of information. With a large number of new conversation threads initiated every day, there is a great variety in number of responses that an individual receives. As with academic journal articles,
2
The factor analyses of different entrepreneurial decisions revealed two major factors entrepreneurial intent and entrepreneurial activity. The three items for the construct entrepreneurial intent ‘Seriously considered a business idea, in view of potentially starting a company’, ‘Looked for resources to start a new business’, and ‘Made written plans to start a business’ exhibited a high degree of internal reliability, with a Cronbach alpha of 0,875. The two items for the entrepreneurial activity construct ‘Sold products and or services through your company’ and ‘Sent invoices for products or services offered by your company’ had a Cronbach alpha of 0,94.
25
messages posted by high-status individuals are more likely to get read, and therefore, more likely to solicit responses.
Control Variables Burt (1992) argued that individuals that act as brokers between individuals have an information advantage and a higher likelihood of discovering new ideas. We measured the degree to which individual’s ego-network consisted of disconnected alters. To do so, we constructed the full network from the web log data and derived a measure of structural holes using Burt’s aggregated constraint. Individuals that have been active in the community for a long period of time have had more time to recognise new ideas. We therefore controlled community tenure, measured as the number of days since the individual’s first posting in the community by using the web log data. Individuals who claim to have been inventing new ideas may be more inclined to become entrepreneurs. To tease out this effect, we asked in the questionnaire whether or not the individual had been inventing. We control for education and age by using information from the questionnaire. We excluded gender as a control as the community only included 1,44% (!) females.
RESULTS The descriptive statistics are presented in Table 2. We mean centred the main effects in the analysis to avoid multicollinearity when testing the hypothesized interaction effect.
---------------------------------------------------------------INSERT TABLE 2 ABOUT HERE 26
----------------------------------------------------------------
Our dependent variables are binary variables reflecting whether an individual have entrepreneurial intent or have done entrepreneurial activity based on their experience in the online community. As explained in the methods section, we have an issue with that respondents to the survey are slightly more active in the community. To control for that effect, we use a Heckman probit procedure where we estimate the likelihood of self-select and being part of the survey by the individuals overall number of postings in the community (Heckman, 1979). Our analysis is therefore based on 280 individuals that responded to the survey (the uncensored observations), while controlling for that these individuals have self-selected to respond to survey. Table 3 shows the results from Heckman probit regressions predicting entrepreneurial intent to test Hypothesis 1 and 2. Model 1 is the baseline model and Model 2 includes the two independent variables. As can be seen in Table 1, the hypotheses receive good support. First, when the independent variables are introduced into the equations, the Wald Chi Square statistic is significantly enhanced. Consistent with Hypothesis 1, we observe that the respondent’s lead user characteristics statistically significant influences at p < 0.01. Hypothesis 1 is thereby supported in our data. Also, Hypothesis 2 proposing a positive relationship between agenda-shaping activities and entrepreneurial intent is also supported at p
