You said that Sites that are useful for “online friends” and don’t require too many people are the easiest to get off the ground (but have other issues, like they might be too niche) – I looked at the Skype example, requiring 2 minumum but the opposite of too niche, and am wondering what're the factors that decide whether a useful site that requires only a few people to “get off the ground” is niche or not…

Thanks!

We recently launched a Tell-a-Friend widget (http://www.socialtwist.com) to help websites/blogs to get their content promoted by visitors. It is catching on well, but we have been thinking about this critical mass issue. Today, 80% of sites implementing our widgets are from only two countries US and India (both places, we are doing some marketing). For a product like this, should we just focus on select locations and not face fragmentation, as you suggest? Although we are getting several requests for multilingual support.Should we simply work on critical mass at few regions, before.

Thanks,
Dave

Provacative stuff. I would add that one further consideration is the “density” of the starter population from which the network arises.

Social sites that are of the “organize my friends” variety — Facebook, eg — grow because they leverage existing online and offline social networks (ie, people invite their friends.)

These populations in effect have 99% density. That is, just about anyone that a user knows may be fair game to invite to the new network.

But in building social networks around shared interests — golden retriever fanciers, ultramarathoners, rare coin collectors, or in the case of our sites at http://www.healthcentral.com, people with diabetes, breast cancer, or migraines — the density of the target population is way, way low. It’s highly unlikely that a person diagnosed with type II diabetes, for instance, knows anyone else in their existing social networks who suffers from that disease.

Thus a single individual can’t be a huge viral catalyst for these kinds of social nets, for the simple reason that they can’t just email all their friends with “hey, i found a great diabetes site!”…their friends won’t care because very, very few of them have the disease.

Ironically, while the 1st type of social net is a “nice to have” application for people to organize their social lives, etc., the 2nd kind of net — one of shared interests — is critical: Whether it is trying to learn about stamp collecting or trying to cope with MS, connecting with a network of people who have the same goals and interests is huge — there is so much shared information and support to be gained and given compared to one’s own already-existing social nets.

Thus in the case of building these 2nd type of networks, the challenge isn’t just to get a user excited about the benefits of the social net — these are quite obvious and quite powerful. It is figuring out how to grow those populations to the level that they are valuable. One way is to leverage the already existing, smaller node points of existing communities that can help spread the word. Another is through transference: People come to the site via searching for basic information — side effects of a drug, for instance — and discover the social net that exists there. But the density factor means that these kinds of crucial nets grow far, far more slowly than the “organize my friends” type of net.

The good news is that, because they intrinsically are more valuable to each individual, these “shared interest” nets are likely to be more lasting once built.