Online Community Summit: Marc Smith

I spent last Thursday and Friday at the Online Community Summit in Sonoma, California. Forum One, which organizes the event, brought together a fantastic group of folks — about 60 people, including corporate and grassroot community facilitators, foundations, venture capitalists, consultants, and researchers. Zack Rosen joked that he wanted to come to this gathering, because he always sees the same people at the other events he attends, and he didn’t recognize anybody at this one. I saw several familiar faces — Zack, Tony Christopher (who informed me of the event), Jerry Michalski, Jan Hauser, Paul King, Thomas Kriese, and others — and recognized the names of several other attendees, but most of the group were new to me.    (2E1)

I enjoyed meeting and talking with Marc Smith, whose work I’ve blogged on multiple occasions, and who kicked off the day by giving a fantastic overview of the work he’s doing at Microsoft. Marc is doing very, very good stuff, and I’m not just saying that because it overlaps with some of our own work and thinking.    (2E2)

NetScan performs a variety of data analysis on USENET newsgroups based entirely on postings, and presents those metrics in useful ways. A lot of the innovation is in the visualization (see his slides for examples), and the visualization software is freely available. Some key points:    (2E3)

  • About two percent of the 13 million USENET users post three or more days a year. That doesn’t sound like much, but it works out to a quarter of a million people.    (2E4)
  • Support communities need about 40 active posters to be sustainable.    (2E5)
  • The metrics you expect depend on the type of newsgroup. The alt.binaries.* newsgroups have lots of posts, lots of posters, and very short threads. Metrics describe the interaction, but to place a value judgement on these metrics, you need to combine them with qualitative analysis.    (2E6)
  • Marc posed a bunch of metrics to look for in a healthy community: retention of leaders, interaction, size and growth, topical focus, speed, and host participation.    (2E7)

I am amazed that more people have not done this kind of research. The opportunity for evolving our applications in useful ways is tremendous. As Marc sardonically stated, “There’s a little room for conversation in our UIs.” Marc showed a few possible directions in which to evolve UIs, but whether or not these features will show up in future applications remains to be seen. There’s a tremendous opportunity for Open Source developers to study this research and implement its findings in their own applications.    (2E8)

Marc also described Project A U R A, which I blogged about a year ago.    (2E9)