The study, which analyzed participation behavior among 26,539 members of 66 private online communities, also found that consumers prefer fully transparent and branded communities to non-specific, non-branded ones.

"Big public communities may attract more eyeballs, but they may not be the answer for marketers who are looking for deep engagement with customers," said Julie Wittes-Schlack, Communispace vice president of innovation and research.

The more intimate the community, the more people participate. Results indicate that 86% of the people who log on to private, facilitated communities with 300 to 500 members made contributions: they posted comments, initiated dialogues, participated in chats, brainstormed ideas, shared photos, and more. Only 14% merely logged in to observe, or "lurk."

By contrast, on public social networking Web sites, blogs, and message boards, this ratio is typically reversed, as the vast majority of site visitors do not contribute. In a typical online forum, for example, just 1% of site visitors contribute, and the other 99% lurk.

"Everybody is talking about communities now, and so the question is no longer 'should we have one?', but more 'what kind should it be?' and 'how can we design it to truly engage people and fulfill our objectives?'" said Communispace President and CEO Diane Hessan.

When potential members were considering whether to participate in a community, they were 30% more likely to log on when the welcome notice disclosed the company sponsoring the community. Branded sites had an initial log in rate of 71%, compared with 55% for unbranded sites.

In addition, of the 66 communities analyzed, parent communities, as a group, had the highest levels of participation. In general, the research found that the stronger the "social glue"--or common interests and passions among members--the greater the participation.

The research found that although members of women's communities participated more frequently than men, men seemed to have more to say when they did participate: 4.8 weekly contributions for men compared to 4.1 for the women.

Notably, educational background and household income were not related to community member participation, as the passion around a community's purpose appeared to be the main influence on participation.

This news story is important. I recommend that everyone read it.