The GroupSwim Diving Board

Finding an expert isn’t easy.  Sometimes they make themselves available and sometimes they don’t.  Sometimes they say they are an expert and sometimes they don’t identify themselves.  Generally speaking, there are 2 ways to find them; explicit or implicit.

If experts are explicitly identified, then they have either named themselves or someone else identified them for others; this method has pros and cons.  On a positive note, they are usually easier to find as they have been “published” or “promoted”.  For example, company X lists Bob Y as the expert on integration.  More often than not, the expertise will be genuine.  On the con side, this method often relies on things like tenure and/or academic distinctions.  I don’t know about you, but these people may not experts in the way I need their expertise.  They may be book smart or “talk real good”, but couldn’t execute their way out of a paper bag.  One example that comes to mind are PMPs (Project Management Professionals).  The process for getting PMP certified is difficult and time consuming.  In theory, people who pass this test and go through the process should be top notch project managers.  Well, I’m here to tell you this is not necessarily true.  I’ve met many a PMP who couldn’t manage my kid’s birthday party much less a complicated project.  My point is you can’t always rely on explicit expert identification.

The other method is implicit.  Experts are identified implicitly through their actions, reputation and performance.  We all know these people in the places we work, schools we’ve attended and so on.  They simply know their shit and have the answer when you need it.  There isn’t a formal way to find these people.  You can find them by asking around and working with them.  The con for this method is it is often difficult to find them since they aren’t obvious or promoted.

GroupSwim finds experts using the implicit method.  The software takes the content in a collaboration site, and analyzes the tags, the rating, and the people submitting the content to figure out who the experts are by tag or topic.  For example, Joe H. is an expert using the Gruntmaster 2000.  He submitted a couple of documents he created.  He answered a bunch of questions about the Gruntmaster that people asked.  And, he contributed to several wikis and templates for using the Gruntmaster effectively.  Our software monitored all this activity.  It observed that most of Joe’s contributions were read, downloaded, watched, forwarded, voted positively on, etc.  We can determine through the group that Joe belongs to that the information pertaining to the Gruntmaster 2000 that he provided is valuable.  Over time, when we see these patterns emerge consistently, we can attribute authority on the Gruntmaster 2000 to Joe.  We’ve done our best, and continue to refine, the technology to perform this service.  The benefit here is when someone using GroupSwim searches on Gruntmaster 2000, Joe’s names comes up as an expert.  His expertise is very likely to be authentic based on the groups positive reaction to his contributions.

While neither way is perfect, we’ve decided to follow the implicit approach that relies on earned expertise.  Nobody gets to be an expert until their peers and colleagues say so, through their actions.