CC BY
10
WHY PARTICIPATE IN THE COMMUNITIES OF INNOVATION? STUDY OF MOTIVATION DRIVERS FOR MAJOR PARTICIPANT GROUPS
(Part I of the COI series)
by ALEXANDER POHL, DANIEL MÜHLHAUS,
ROLF WEIBER UND MARIA VOLA1
Communities of Innovation (COIs) are a phenomenon, that gained attention in the recent years as a pragmatic alternative to the in-house innovation (R&D departments). COI approach claims to yield excellent results in the open innovation sector, which is usually explained by COI strong developmental dynamics. More and more companies are trying to obtain their share of innovation capability by encouraging and participating in the COIs. While so far only several discrete faucets of COIs were analyzed in the existing literature, the general dynamics of a COI are still largely unexplored. For example the so-called direct network effects that evolve because the number of contributors has considerable influence on motivation to participate and on the quality of interactions. In this particular essay we take a closer look at this and other intrinsic and extrinsic motivation drivers based upon the segmentation of participants in the general organizational context. The motivational framework described here aims to support companies in fostering such communities or otherwise participating in COIs, amplifying the already proven positive effects.
Keywords: Virtual teams, network economics, IT and new organizational forms, Communities of innovation, online communities, open innovation, open source software, technological innovation, collective innovation process
1. INTRODUCTION: COMMUNITY CONCEPTS AS VALUE DRIVERS FOR OPEN INNOVATION PROCESSES
High R&D costs, shortened product lifecycles and a high flop rate in market launch of innovations have led many companies to move away from traditional innovation processes (West and Gallagher, 2004). Whereas the development of new products and services was previously driven almost entirely by the firm's internal resources, such as a powerful R&D department or a progressive management, companies are now increasingly trying to involve their customers in the innovation process. Chesbrough (2002) describes Open Innovation as the integration and systematic development of external resources to generate innovation capability. Currently, a growing number of firms are realizing that opening up their traditional, firm-centred innovation processes holds enormous advantages. There has been several noteworthy attempts to tap into this innovation resource (Gassmann and Enkel, 2006; von Hippel, 2005; Füller et al. 2006).
So-called Communities of Innovation (COIs), where users interact with each other and collaborate to generate solutions to specific problems, exist for several decades. However COIs are now gaining particular importance due to the spread of a relatively new medium - Internet. Ideas and concepts are disclosed for assessment at an accelerated speed and can be used by other interested parties to develop their own solutions. Dynamic learning and evolutionary processes are set in motion in an environment, where the requirement fit of the concepts developed is continuously examined and an interactive improvement process (extensions, adjustments and corrections) acquires perpetual nature. Hienerth (2007) describes this as follows: «The overall community can be seen as a dynamic system of adaptive learning that produces innovations from the ideas of community members». Thus, a COI combines the process of identifying customer/user requirements with the process of generating solutions, accompanied by dynamic and collective learning. Therefore, we consider integration of COIs into the firm's R&D activities to be an approach, that yields the greatest value on one hand, and is the most reliable in the field of Open Innovation on the other, which is why COIs are also of particular interest to firms at the moment.
Especially bright example of the new-age COIs are the so-called Open Source Software Projects (OSS). As early as 2000 and 2001, IT firms were disclosing open source projects to a large number of contributors, for example, IBM with Eclipse and Sun Microsystems with Java and its development tools (cf. Henkel, 2004). Both firms also support many open source projects organized through Apache.org. Furthermore, IBM and Sun provide suitable staff for the projects, e. g. to create seminal technologies for the development of web applications. However the success of community-based innovation projects has also been confirmed in other industries. Shah (2009) investigates the successful COIs in the sports (i.e. windsurfing, snowboarding) and children toys and articles (i. e. jogger.cart) industries through the 20th century. Fuller, Jawecki and Muhlbacher (2007), for example, show huge innovation and creativity potential found in online basketball communities, for example in designing new sports shoes. The great success of computer games, such as Half Life, is based to a significant extent on the large number of creative extensions originating from COIs. Some of these are integrated into new versions or, alternatively, made available as downloads (cf. Jeppesen, 2004). Furthermore Fuller et al. (2006) highlight the beneficial nature of the COI concept in the car industry.
However, there is also a number of examples where community-based innovation projects have failed. On the subject of OSS projects, Krishnamurthy (2002, p. 1) notes that «a lot of OSS programs do not generate a lot of discussion». Healy and Schussman (2001, p. 16) go one step further and say that a large part of open source software projects fail and «are fallen deadborn from the PC». But there are also other areas in which unsuccessful COIs are not uncommon. The relatively well-known «Oscar Project» (www.
M <
24
► M
theoscarproject.org), for example, which was initiated in 2000, was considered to have failed and now has been reactivated.
All in all, we can only agree with Christley and Madey (2007, p. 1), who stated that «while research of a few, large, successful projects have provided insights into the nature and practices of the open source software community, it still leaves open the question about the thousands of other open source projects which are neither large nor highly successful.» In order to be able to detect failure of a COI at an early stage, first of all we have to understand how a COI works. Only then critical developments can be identified in the right time and the course of corrective action can be developed and implemented. If companies are aware of how COIs function, they can derive further approaches to achieve success or obtain additional sources of support, effectively making use of COIs for their own innovation purposes.
2. ANALYSIS FRAMEWORK AND RESEARCH DEFICITS
The existing research work on COIs can be divided into conceptual and empirically oriented papers. In the conceptual sector (e. g. Lerner and Tirole, 2002; Harhoff, Henkel and von Hippel, 2003) we identified three major directions of thought (cf. von Krogh and von Hippel, 2006): Firstly, researchers analyse the reasons participants take part in such communities, focusing on the question of why they supply knowledge without payment. One of the thesis is that providing assistance and disclosing self-developed concepts are the strongest motivating factors (e. g. Hars and Ou, 2002; Lerner and Tirole, 2002; Hemetsberger, 2003). Furthermore, there is a whole series of studies, whose central theme is coordination and structure of community-based development processes (e. g. Scacchi, 2004; Bonaccorsi and Rossi, 2003). Finally, a third direction of the research deals with the effects of such innovation achievements on existing competitive structures (e.g. Economides and Katsamakas, 2006) and on the behaviour of commercial participants (e.g. Mustonen, 2005; Dahlander and Magnusson, 2005). The empirically oriented research work on COIs appears to concentrate on the individual facets and functional elements of COIs. In particular, the existing studies investigated participants' motives (e. g. Shah, 2009; Lakhani and von Hippel, 2003; Ghosh et al., 2002; Hars and Ou, 2002), the coordination mechanisms in a COI (e. g. von Krogh, Spaeth and Lakhani, 2003; Shah, 2006) and COI success factors (e. g. Crowston, Annabi and Howison, 2003).
A central research deficit is the fact that each study is only a snapshot examination, where the various aspects of COIs are viewed largely in isolation. Interdependencies between, for example, the motives of the individual contributors and the evolution of a community (extent of communication, solution concepts generated) are not really taken into consideration. However, particularly due to the collective nature of the interactive processes in COIs, an isolationist approach is not applicable in explaining their dynamics because the participants' motivation depends to a large extent on the behavior of the other contributors. The research work investigating the dynamics of this kind of «group phenomenon», however, focuses entirely on macro variables, such as contributor fluctuation, changes in the volume of contributions, or modification of concepts (e.g. von Krogh, Spaeth and Lakhani, 2003; Butler, 2001). The motivations and activities of the individual contributors have not been taken into account, which can be considered a substantial research deficit in the context of the network structure of COIs.
Considering these deficits, we decided to make an emphasis on the design principle and the functional elements of COIs, and describe it rather from the micro-perspective. The intention of our research is to investigate the main elements, structure, and success factors of a COI in order to create a set of specific recommendations for companies and initiatives for fostering the communities of innovation of their own.
3. GENERAL DESIGN PRINCIPLES OF COMMUNITIES OF INNOVATION
3.1. Major elements of a coi
A community of innovation is generally understood as an informal, open community of non-commercially motivated participants with the objective of jointly developing new and improved products and solutions within a defined area of application.
In order to describe the basic principles of a COI we must first of all distinguish between the micro level and the macro level. While the micro level focuses more on the COI participants and their interactions, the macro level is comprised of general structural features of the community, such as the critical mass of active contributors and users of the solutions developed, and general large-scale effects of a COI as an entity. The micro level incorporates four major elements: Participants, Process, Solution or a Product, and Diffusion or Market (i. e. diffusion of a COI product on the market).
Solution / Product
- quality
- scalability
Contribution
Motivation (perc. success)
Process
- quality
- intensity
- frequency
Diffusion / Market
- # of users
- profitability
Figure 1: Major Elements of a COI
Participants in this context are viewed as a sum of skills, roles, and motivations. Process element here incorporates the frequency, intensity and quality of the activities of individual community members. The two other constituents - Solution/Product and Diffusion/Market - are very important also as tangible success indicators, however are not the centerpiece of a COI as much as in a commercial enterprise in a motivating sense. The question arises then, what are the other factors that motivate COI participants instead of monetary rewards? Since the formal goal of a COI is to generate innovative solutions under the
circumstances where the members receive no direct payment for their contribution, the participants' perceived success (i. e. the extent to which they are able to fulfill their individual objectives) is an important aspect. In effect, perceived success (defined individually by each participant) provides motivation, which ideally leads to overall success of a COI, also on the macro level. Success factors that interconnect between the two levels were not investigated at all in the literature as of yet.
3.2. Coi as a critical mass system
the micro and macro levels cannot be considered separately because there are strong interdependencies between the two. If, for example, a large number of members with complementary knowledge and skills contribute very actively, the pool of information available grows. A high level of communication activity therefore leads to considerable advantages for community members (Hagel and Armstrong, 1997). The structural characteristics of a COI and their stage of development thus have direct influence on the activities, creating additional motivation to participate, both for existing and potential COI members (snowball effect). Essentially, this means that individual activities have a direct influence on the community and this in turn provides feedback to the contributors.
COIs differ radically from firm-centered innovation processes by the fact that success can only be obtained in collaboration with others, that is, COIs posses a so-called direct network effect (cf. Scacchi et al., 2006). Direct network effects appear if the benefit to the individual participating in a social system depends directly on the number of members in the social system (cf. Weiber, 1995). Direct network effects involve so-called feedback because individual benefit to existing members increases with every new contributor (positive feedback). Similarly, however, decreasing membership means that the remaining members lose benefits (negative feedback). Because of the direct network effects COIs as a social system type are disadvantageous in the initial phase. As the number of members in the development stage (a so-called installed base) is relatively small, the system runs the risk of members leaving again because the individual benefit stemming from the number of contributors is marginal. The next stage of COI development is a phase of stability - when the number of contributors rises beyond this critical level (so-called critical mass) - due to the increasingly positive feedback. COIs are a form of Critical Mass System because sustainable advantages are not achieved for the individual contributors until a certain number of active members is exceeded and a large number of diversified ideas and concepts are available (cf. Markus, 1987; Marwell and Oliver, 1993).
The critical mass nature of COIs is also highlighted in some studies on existing COIs. Scacchi et al. (2006), for example, notes on the subject of OSS projects that there are direct network effects here that lead to some projects growing exponentially in terms of number of contributors, lines of program code and functionalities generated. Lancashire (2001, p. 5) notes in this context of some projects experiencing «snowballing support». The need to have a large number of contributors is also pointed out by Raymond (1999), for example in his comments regarding bug-fixing process, where he notes, that «given enough eyeballs, all bugs shallow». Mockus, Fielding and Herbsleb (2002) also consider a large number of contributors to be important for feedback in evaluation of concepts. If, however, there is only a small number of interacting contributors in the long term, the entire community may collapse.
Unlike the applications from the information and communications systems sector, otherwise described as critical mass systems, (cf. Schoder, 2000; Weiber, 1992), the number of contributors to a COI is subject to certain limits (cf. Butler, 2001). It appears, for example, that many successful OSS projects are driven
M <
27
► M
by a small nucleus of developers (cf. Mockus, Fielding and Herbsleb, 2002), which also confirms findings in other areas in which effectiveness and performance of teams support the argument that the optimum potential lies in a limited group size (cf. Shepperd, 1993). Thus, for some fields, such as development of concepts, we can assume that too many participants cause the community's inner stability to diminish, and communication structures are weakened in such a way that the effort required to coordinate all contributions is too great and there is a kind of «information overload» (cf. Jones, Rafid and Rafaeli, 2004).
4. COI AS AN ORGANIZATIONAL ENTITY: WHAT MAKES IT WORK?
In order to map the functions of a COI adequately, we will go back to the four central elements. The nucleus of a COI is formed by the various participants who are actively engaged in a community-based innovation process. It is evident that collective learning and development processes can only be achieved by means of balanced interaction by the participants (cf. Sawhney and Prandelli, 2000). For this reason, it is very important to understand what motivates these participants to contribute and which participants are particularly important for the innovation process at the various stages of COI development. An important motivating factor of a COI is perceived success, which has different faucets and can be related to the quality of the innovation process, the quality of the solution generated, and to diffusion of an innovation («marketability»). Other than in case of classical, firm-centered innovation, where employees are paid for their contribution, it is essential to understand here what the participants perceive as a success and what they consider a failure (Crowston, Annabi and Howison, 2003). In the course of our research we discovered, that the main motivating factors strongly correlate with the type of the participant, defined by their roles and responsibilities, within a COI. We identified several types of participants, and were able to draft the structural dynamics of a COI, depicting the interdependencies between those:
Commercial firm
Nomination, control
±
Promotion, commercialisation
Contribution, disclosure, withdrawal
Project leader 4"
Coordination
Development, improvement
Developer
Supports documentation, feedback and bug finding
Community of innovation
«Process quality»
(Determines the quality of innovation)
Use, promotion
«Solution quality»
(of innovation)
Users
Trial, use
Facilitator
Supports implementation, use, functionality addition
«Diffusion»
(of innovation)
Figure 2: Interdependencies of the COI elements
28
► M
Firms were included as well- as commercial participants, to illustrate that firms can control the course of a COI process and therefore can also influence its success. In the following sections we discuss the functional elements of a COI as well as investigate the following questions:
• What makes a COI successful?
• Who or rather - which groups of participants are of central importance?
• Why do these participants contribute to a COI?
4.1 Success factors of a community of innovation (What)
Upon reviewing several essays that provide analysis of success factors of COIs we were able to derive three general dimensions with corresponding variables. Specifically, Table 2 contains summary of each paper's central findings. In their investigation into COIs in so-called marginal sports Franke and Shah (2003) distinguish between: a) process-related; b) result-related; and c) COI impact-related success factors. For example, a process-related success factor would be satisfaction of the participants with the attained mutual assistance, while result-related factor would be the quality of innovations produced on the basis of their novelty factor or their market potential. An impact-related factor would be the diffusion of the solution concepts. The authors come to the conclusion that there are strong dependencies between these three dimensions. They argue that the interaction within the community leads to improved solutions, which is then crucial to their diffusion. This statement supports our thesis, and corresponds to our COI elements model.
Crowston, Annabi and Howison (2003) identify in their paper success variables of COIs in the open source software sector, based on the opinions of programmers and interested parties from the «slashdot» community. The individual aspects identified relate primarily to the dimensions «process», «project result», and «consequences for the contributors». This work corresponds almost exactly with that of aforementioned Franke and Shah, except for the latter aspect. «Consequences for the contributors» is the motivation-related factor, and is interconnected with the perceived success.
Another useful differentiation criteria is external versus internal success (this is where macro versus micro perspective enter into the analysis). We define external success primarily as capability of generating a functioning solution and distributing it on the market. Internal success is measured by achievement of the objective pursued by the participants in the course of the innovation process. Again, these two are interconnected, i. e. external success becomes internalized, if it is a primary objective of an individual participant.
Summarizing the literature analysis the three «success dimensions» of a COI can be defined as «process quality», «solution quality», and «diffusion» - corresponding to the main elements of a COI in our model. Following are the suggested variables of each «dimension»:
й^ВЯТй
