Academic journal article Research-Technology Management

A Pervasive Model for Participation in Voluntary Forums: Participation in Online Challenges and Forums Follows a Predictable Pattern across Industries and Public Forums, at Scales from Dozens to Millions of Contributors

Academic journal article Research-Technology Management

A Pervasive Model for Participation in Voluntary Forums: Participation in Online Challenges and Forums Follows a Predictable Pattern across Industries and Public Forums, at Scales from Dozens to Millions of Contributors

Article excerpt

Internet technology and familiarity with social media make it easy to broadcast messages to hundreds and thousands of employees, suppliers, or customers; the optimistic visions of The Wisdom of Crowds (Surowiecki 2004) and Open Innovation (Chesbrough 2006) suggest that the ever-expanding reach of social media could be used to access mass intelligence to power innovation. It's the rare business leader who has not at least considered these new models for collaborative work. For that leader, it is tempting to imagine being freed from the limitations of physical meetings and scheduling problems to augment traditional business teams with the collective knowledge and experience of huge virtual teams.

Yet participation in such virtual teams takes a very different form from traditional business teams, with major consequences for the business leader's expectations and project management. Our standard of comparison is the business leader's mental model for team members' performance, most likely a model in which there are a few low performers, a broad peak of "typical" performers, and a few high performers (Figure la). This expectation underlies a great deal of managers' behavior and is also the basis for most corporate incentive and promotion programs. But the distribution of contributions from people in a voluntary participation forum is dramatically different; with rare exceptions, a few people contribute a lot and a large majority contributes a little (Figure lb). This pattern is seen in a wide diversity of companies and public forums and has a strong basis in theory, suggesting that it applies whenever (1) participation is not forced, and (2) there is some form of sustained positive feedback for participation.

These different participation curves originate in the very different circumstances of the two venues. The bell curve of typical team participation is a consequence of constrained behavior when a defined team of people is set to work on a given project. People vary in their talent and knowledge, and when team members each work about the same amount of time, these variations add up to the bell curve. But large virtual forums are voluntary in practice. We know that in physical "town hall" meetings, only a few people will raise their hands and speak while the majority listen (or daydream). There is even less social pressure to respond to an electronic announcement when sitting at one's desk, so active participation becomes a personal, voluntary decision. In voluntary events, the equal-time constraint is removed, and the dominant factor in contribution becomes the amount of time a person chooses to devote to the task. Most people will ignore the summons and contribute nothing, some people will contribute a little, and a few will contribute a lot.

[FIGURE 1 OMITTED]

My five years of experience crafting and running more than 200 new-ideas challenges inside Pfizer, plus analysis of large datasets generously shared by colleagues in other corporations and a review of the literature, reveal that this is a common pattern in voluntary online forums. The factors affecting this pattern appear to be that the contributor is distant (in space or influence) from the forum sponsor and that participation is voluntary.

[FIGURE 2 OMITTED]

When these factors are in place, the pattern of participation is remarkably consistent across diverse industries and business purposes and across groups of participants. Participation data for forums are commonly visualized in a rank-frequency graph, with a person's rank on the horizontal axis and level of contribution on the vertical axis. "Rank" is found by sorting participants such that rank=1 for the most-active contributor, and rank=N for the least-active contributor (where N is the total number of contributors). Using logarithmic scales for both axes shows the pattern in full detail (Figure 2).

This pattern holds for huge public social collaboration systems, such as Wikipedia and Digg (Chi et al. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.