I don't follow this reasoning. Social networks do tend to rely on network effects, but must they?

It sounds cyclic. Social networks need lots of users because they can't charge their users directly, and they can't charge their users directly because they need so many users.

I don't know that app.net has to replace Twitter for everyone. Can't they be smaller and yet successful, precisely because they charge users directly?

No, social networks need lots of users because their value is related to the number of connections. If I don't know very many people on a social network, it isn't very valuable to me.

Therefor, to offer the most value to users, those building social networks want to get rid of barriers (including payment) that might prevent more people from signing up. That, then, gets reinforced by the cyclic relationship between the fact that indirectly monetizing users tends to pull in relatively little per user while marginal costs of hosting an additional user are minimal - but the initial dynamic is an artifact of the nature of social sites to begin with.

I think you're taking a huge leap in your reasoning. Sure, I want lots of people I know on the network. But beyond the roughly 200 of those, I don't care if the network has a hundred thousand users or a hundred million.

Before facebook, social networks could provide a lot of value by saturating small demographics. Example: If all Swedish teenagers are on playahead.se and can all talk to each other there, they don't gain much from network expansion. The reason those networks need to grow beyond the clique where they're successful is that their business model can't sustain itself on a small number of paying users.

If app.net saturates the demographic of "people who care enough about Twitter's new API to chance $50 away," I could talk to 50% of my Twitter circle even if the total number of users on app.net is only 1% of Twitter's.

Most people I know couldn't give two hoots about talking to random people on the internet, even interesting random people. All they care about is their friends (and maybe famous/notable people). If your social network contains < 100,000 people, chances are it doesn't contain many friends, so most people will discount it. Now, us techies are used to a degree of anonymity, and are used to interacting with people we haven't met, so I don't think it is impossible that app.net will be successful in that demographic. I think it is akin to Netflix: some people just want to watch a movie, it doesn't really matter which one so long as it is a decent one; these people might like Netflix. But others only really care about that specific movie their friend said was good; chances are Netflix doesn't stream it, and they will be disappointed.

Nah facebook stopped being useful once my dad and aunt would read and post stuff there. It turned from a hangout at a friends place into sunday supper.

Well, the effect doesn't preclude the possibility that there are connections you specifically want to avoid. My point was just that the value of the social network is the connections - Facebook would be even less useful if none of your friends were on it in the first place. Managing who sees what in a way that jives with social dynamics is still something these networks are figuring out. Charging $50 is far more likely, however, to keep out my unemployed friends than my well employed parents.

The value of a social network is never in the technology. It's always in the users.

join.app.net is making the same mistake as diaspora, facebook, myspace, etc. You can't charge someone real money when you aren't the one providing the value.

??? This doesn't even make a tiny bit of sense.

Thanks, you just get me the exact term I'm looking for. It's hard to explain to "smart" people without using any computer-ish/science-ish/math-ish terminology

And yet 12,000 people payed to get a network going. Myself included.