Just read a great article/presentation related to the challenges of unsegmented social graphs (linked/embedded below)
The case used in the presentation to illustrate the problem was a swim teacher with young students who was also connected to gay friends and often commented on their pictures on Facebook - unknowingly exposing her students to photos from wild nights out from her friends.
The more common case is just that we normally behave differently with different groups of people (family vs. co-workers for example). Mark Zuckerberg would probably say that people with nothing to hide wouldn’t need any segmenting. That’s just not realistic. We interact with people differently.
The problem is, that as methods to segment social graphs emerge, it’s entirely too tedious to go *back* and put everyone into buckets. FB lists or groups (not sure what they’re called) are just too much work.
And that is why, in my opinion, Google’s social site (Google Me?) could win. People loved the blank canvas that moving from MySpace to Facebook provided. “OK, this is my REAL network”.
As Facebook has increased in popularity, people have started making much looser connections. What was once family and friends is now friends, family, industry connections, coworkers, etc.
More than that, adding people to ‘buckets’ of contact types at the time of connection is 1000x easier than going back and trying to do it now (think about your Twitter following, would you actually go back and group them all? You’d get about 1/10th of the way through and give up).
That’s why on Sprout Social, we have people add contacts to lists/groups as they connect. The same approach could make Google Me a very likable web property for the overwhelmed, OCD like me, and just people who want some walls between groups of online connections.
Users have different social graphs on Facebook, LinkedIn and Twitter for that very reason. If Google Me offers a way to have that same distinction in one place, they will do amazingly well.