Do social networks have an optimal size?

A few weeks back, The Economist carried a somewhat sceptical article about Facebook and its valuation (Social graph-iti: There's less to Facebook and other social networks than meets the eye) - which by itself was not too much of a surprise since it reflected some of my own skepticism. However, buried in the article was a comment that caught my eye:

But unlike other networks, social networks lose value once they go beyond a certain size. “The value of a social network is defined not only by who's on it, but by who's excluded,” says Paul Saffo, a Silicon Valley forecaster.

I'd always assumed that the network effect applied to social networks as well, so this took me by surprise. I'm still not sure what to make of it: the first part of the comment attributed to Paul Saffo makes sense (value defined by who's on it) - but I don't know that I agree with the second part because it adds a connotation of elitism, and because I think larger networks should actually be more valuable than small networks.

So lets think about that for a second... is there an optimal size, somewhere between a network of one, and a network containing the whole known universe (everyone on the planet)? I guess a network of one (just me) isnt very useful, and as we factor in more people into our network it gets more useful. But how big is good, or best?

When I blogged on this train of thought on my internal blog at work, a couple of people responded with interesting comments: JM pointed out that research he'd seen indicated that its possible to have a strong relationship with a network of about 15 people, and extrapolating it out to two degrees of separation suggested a size of about 150 people; AM responded that he tended to focus on quality (of relationship/trust/reputation) rather than quantity for his network, which meant that he preferred to keep his network small - a point also made over at café salemba.

I tended to agree with both of them to some degree... but felt that while for a first approximation the "strong and small personal network size" was a good estimate, it didnt take into account the ability of technology and social networking tools to leverage a much larger network without a corresponding increase in difficulty in terms of managing those relationships.

However, a big part of the value of social networking is not just that it reflects a set of relationships, but also the element of *discovery*, so I would argue that looking at networks in terms of only size or quality of the network itself is a mistake - the ease and quality of relationships that can discovered and added, is more important than a simple metric of size or membership. So going back to Paul Saffo's comment, I would instead argue that better filtering and search mechanisms are more important than the principle of exclusion, i.e. assuming you have a way to filter out what you dont need, the bigger the network the better it is. How big? Well, I tend to think you'd actually want your network to include *all* the relationships and people that you know, even if you don't need them right now, so that you can remember to reach out to them when you need them. To say that yet another way, your social network is an asset - a social asset - whose value you should grow always, and be able to mine when needed. To be precise, I'm thinking that there's no lower bound for an optimal size - the optimal network size would be as big as possible (with the caveat that these are people you actually do know, of course).