Our hosts for this session are Cameron Neylon and Deepak Singh, who have themselves explored the many social networks developed especially for scientists and bring to us the question: Why aren't these more successful? Some examples are Nature Network, MyExperiment. Others? Please add others that you know of, and your thoughts on them.
The obvious elephant in the room is Facebook; none of these sites have anywhere near the reach of a site like Facebook (or Ning or MySpace or ... ). FB was built for an existing community, originally Harvard undergraduates, then undergraduates of a small group of universities, before going public to everyone. For networks built specifically for an existing community of scientists, why has success not followed? Cameron has also brought up his FriendFeed profile to show all the data that it can aggregate for him, and how it allows him to keep up with people (subscriptions to friends' accounts, profile views of others from Facebook and Flickr, etc.). And LinkedIn is also mentioned as a networking site that has a useful purpose and a strong user base to keep it popular.
Among the networks we're discussing today: Nature Network, for example, does bring return visits from the users in the room due to the community of people found there that cannot be found elsewhere in that particular aggregation. Cameron theorizes that it is less about the science in the community than about the people there; how much scientific debate is occurring there? Some folks report that the forums and the Ask the Editor feature serve this purpose, though the rest of the site may be somewhat sparse when it comes to this.
Other than socializing, what are some of the uses of social networks for scientists? Several reports from the room indicate that solving a problem that no one in your immediate area (lab, library, etc.) has any solutions for is a great benefit of being a member. It can also facilitate introductions to people working in specific fields or at specific institutions, if that's what you seek; Deepak poses the question of whether we need specific scientific social networks to do this, or whether sites like del.icio.us or LinkedIn would be sufficient. As a room, we're not yet convinced of the usefulness of specific social networking sites for us.
Cameron has five ideas for successful social networks to follow: 1) there must be a problem that it solves, 2) the successful tool must be the best performer among many (unless it's the first of its kind, but even in that case, imitators will eventually come along if you experience any measure of success), 3) the tool must be as reliable as possible (anything less will be out of business very soon), 4) it must provide "at least one killer feature," and 5) it must be pre-populated with some users, as empty sites do not yield return visits. Other ideas that people mention are: tools that don't lock all content posted there behind copyrights, and tools that aim for the best in usability. Failed social networks often used the Build It and They Will Come mantra and thus didn't fulfill one or more of the items on this list, particularly 2 and 5.
To end the session, Cameron showed us Ravelry, an example of an extremely successful social network for a particular cohort of users, in this case knitters. With 260,000 or so current members, Ravelry manages itself by initiating a waiting list so that gluts of members don't overwhelm the site. It was built out of existing technology, for the most part. Cameron's point is that people aren't specifically looking for other people here, but perhaps for patterns or yarns that they use and want to see the resulting projects. Audience members agree that they use everything from Amazon to Metafilter to Omlet (a maker of urban chicken enclosures called Eglus that helps answer the questions of new chicken owners; seriously!) for similar reasons, an important lesson for builders of social networks in science, knitting, or something else.
I'll be back tomorrow morning with another set of sessions, including a follow-up to this session.