10:00 – 11:00 Session 4 / What Works? A Report from the Front Edge of Community Research
Description: What does current research tell us about effective strategies in the design and management of online Communities?
Introductory comments: Paul Resnick / University of Michigan
“How do you get participants in online communities to participate more and better?”
www.communitylab.org – Handbook of Social Design
If people can change their names then what is the effect on the community?
- You can’t trust new people
- People can get away with more
People on Digg game the system partially because of the design choices they made in the system.
One thing you need to keep a system clean is to have a barier to entry.
Obvious Idea: Have activities anyone can do, but only let the best people at those activities do more important activities.
Getting People to Participate More
Tested Theory: If you assign members of a community a goal then it works. The basic principals of goal setting work except that the more challenging the goal the better it tends to work online… as long as it’s doable. The goal also needs to be specific. The optimal goal was at the 90th percentile.. that becomes the most motivating.
One version of the experiment was done with groups. The theory was that people would participate less if they were in a group. But the reality is that they participated more knowing they were members of an anonymous group.
More contributions from people that were told they had rated rare movies.
If you tell the low performers how they performed to the mediate they will move up.
Think about how you can create groups that people can identify with within your community.
Welcoming Newcomers
Natural intuition is that users like to be welcomed. Users who’s first post got a reply were 12% more likely to return. The tone of the reply didn’t matter. Just being noticed made a difference.
Change in a community also improves productivity.
Introductory comments: Neel Sundaresan / eBay Research Labs
Ebay has three communities and the shape of each is very different.
Ebay, Skype, and Paypal
Realized that people, ala stumbleupon, is that if they came in for one reason, but saw something else then they would stick around and buy the something else.
The surprise factor of random product links actually results in more sales. It has the potential to beat the recommendation engines.
There is a huge difference between what people ask for and what they really want. Like people that search for Jessica Simpson on ebay, but they are looking for shoes.
The community is better at pointing out things that their tools can’t.
Community Computing Effort is their big research area now.
Thinking about there community problems.
Acquisition – Easiest – Marketing to get them there.
Activation – Medium- To get them on the site and perform thier first action.
Activity – Hardest – How do you keep people on the site and being active.
Observation: Stickyness in a lot of categories is based on reselling. People buy from folks with medium reputation and sell to rookies for profit on e-bay.
Discussion
- Does a real name mean anything to people’s participation?
- Real names put people complete online reputation at stake, but profiles only effect how you are viewed on that site.