This. There is a definite finite life cycle of social sites. Would do investors well to study the rise, domination, fall, and practical disappearance of such sites. BBSs, usenet, Compuserve, AOL (freaking bought Time-Warner!), MySpace, many others - all out to take over the world at one time, all legitimately worth vast sums at their peak, all now little more than rotting corpses. Know when to invest, know when to cash out, know when to stay away.
How many here remember sixdegrees.com? (1996) That's the first social networking site I remember signing up to...
Of course you must be "different", but the difference between each successive evolutionary step from Six Degrees to Facebook has been tiny.
While it is getting harder as the size of the successive networks have gotten bigger (SixDegrees reached a low number of million users), because of the network effects, it's not so much about features as about finding a way of building the network to a tipping point, and social networks are incredibly fickle and subsets can migrate to new services because they are not the dominant network.