> The network effect works even if the network is relatively small.Let's do some wild guesstimation here.
Sturgeon's Law says that 90% of everything is crap. Let's apply that to social connections and assume that any given person only cares about interacting with 10% of any given crowd. Let's actually be even more generous and assume people are an order of magnitude pickier than that: for most people, 99% of other people are crap.
Meanwhile, Dunbar's number says that a human only cares about socializing with up to 150 people. Dunbar's number itself is kind of weird pseudoscience, but it seems reasonable to assume that there is an upper bound (or at least diminishing returns), so let's be generous and go an order of magnitude higher and say that a given person won't care about interacting with more than 1,500 specific people.
Putting these together, it suggests that being a member of a network of 1.5 million individuals ought to suffice to give anyone all the social connections they could ever want.
Of course, it took silly reasoning to arrive at this precise number, but I think there's an element of truth to it. The limits of the human ability to socialize mean that, beyond a certain point, the size of the network itself is no longer the limiting factor on the ability to find people to socialize with.