Social Networking Software

We've seen it all before. Cool Technology In Search Of An Application.

Social Networking software and services, like Friendster and Tribenet , have been getting a lot of attention lately. These services allow their users to take advantage of the "six degrees" concept to expand their social and business circles. For example, you might not know Joe Smith, but you might know Mary Jones, who knows Joe Smith. This gives you a connection to Joe, and you can use that to make an introduction for business or social purposes.

Sounds really cool. And the underlying theory is sound. Research has shown that it takes a remarkably small number of connections to link almost any two people in the Western world. Odds are that it takes six or fewer connections to link me to, for example, Ronald Reagan, or Paris Hilton , or Edith from Tulsa, Oklahoma. So, if everyone belongs to Friendster, and we're all diligent about keeping our links current, then if I ever wanted to contact Ronald, Paris, or Edith, I could find a chain of introductions which would get me there.

I just don't see how translating this theory into a real business is going to work, despite all the venture capital being thrown at it.

Problem 1: Social Network Software Is Only About the Network

By itself, Social Network software doesn't really do anything. Of course, the venture capitalists like this stuff because of the winner-take-all characteristics of the network effect (i.e. the first Social Network to get big will own the entire market), but the problem is that services like Friendster are nothing but network. There's nothing else there. They don't really do anything.

Aside from the cool feature of linking you to friends-of-friends, Social Network software doesn't provide any services which you can't get elsewhere, and with much less hassle. Message boards? Matchmaking? Personals? Want ads? Shared calendars? Interest groups? Let's face it, this is not unique, and most other places which offer these services are quite a bit simpler to get started with.

If anything, existing services which offer message boards, matchmaking, personals, want ads, etc. are competing directly against the Social Networking upstarts, since people tend to cluster around areas of interest anyway. The Social Network aspect is an interesting add-on, but not anything really special.

Problem 2: The Network Effect is Second Order

The network effect means that the value of a network (of any kind) goes up exponentially as the network gets larger. A phone network with ten people is useful, but a phone network with ten million people is far more than a million times more useful. It works in reverse, too. If only one person in the world has a telephone, then the telephone is useless. If two people have telephones, then suddenly the gizmo does something.

With a Social Network, the network effect is second order. By that I mean that the usefulness of the Social Network depends not on the number of possible connections between members, but on the number of possible connections which are second order or higher. In other words, if only one person in the world is on a Social Network, that network is useless. If two people are on a Social Network, that network is still useless. If three people are on the Social Network, then it is useful only if two of those people aren't directly connected.

The net result of this is that, where the classic network effect goes as the square of the number of members of the network, the network effect in a Social Network is much smaller, and actually decreases as the number of connections per member goes up. If everyone on a Social Network has lots of direct links to other network members, then the number of second order and higher links is much smaller.

Problem 3: Supernodes Are Less Likely to Join

The people who have the most social links (called "supernodes" because they have many more connections than most network members) have the least to gain from using Social Network software. This is simply because someone who is a "social supernode" (someone who knows everyone) already has a lot of direct social connections, and therefore fewer indirect connections to gain through the Social Network software.

To make things worse, is is harder for a "social supernode" to start using Social Network software in the first place. That's because someone who knows everyone is going to have to take a lot longer to enter all his connections into the software.

So the people who know everyone are going to have a harder time using Social Network software, and less to gain from it. This means they're less likely to join. Unfortunately, the supernodes are also the most critical to making the "six degrees" effect work, which means that if they don't use the software, then the Social Network program is dramatically less useful for everyone else.

Problem 4: Social Network Software Takes the Social Out of the Network

At its core, Social Network software tries to automate something which is a very human interaction: the Introduction. Fundamentally, having machines perform social functions for us (as opposed to merely mediating them) doesn't work, because humans are social creatures, and without the human-to-human contact, there's no point.

Without Social Network software, if you want to meet Joe Smith, you might ask all your friends if any of them know Joe. Mary Jones tells you she knows Joe, so you ask Mary to introduce the two of you. Mary tells Joe that you'd like to meet him, and Joe knows that Mary is vouching for you. This is a highly social interaction, and in some cultures, it can have considerable nuance.

With Social Network software, you search for Joe Smith among your "circle", and you see that he's there, so you send Joe an E-mail telling him he's a friend-of-a-friend and introducing yourself. Much more efficient! Efficient in the same way that artificial insemination eliminates all that unnecessary sweat and noise.

Without Mary's explicit introduction, Joe doesn't really know anything about you, so why should he care that you're a friend-of-a-friend. Aside from the sheer novelty value (which is mostly what Social Network software has going for it right now), there's little reason why Joe should pay any attention to you. Whereas, if Mary had introduced you, Joe has an obligation to Mary to hear you out, since Mary has (presumably) decided that you have a good enough reason to want to talk to Joe.

So Why All The Fuss?

Social Network software is hardly the first idea to become wildly popular among a relatively small group of technophiles who declared it the Next Big Thing, only to see it slip beneath the waves (anyone remember PointCast? I do. I even remember companies selling special hardware and software just to manage all the network traffic which PointCast was going to generate).

Social Network software is much the same. It has caught on among a small group of technophiles--conveniently, many of these people travel in the same social circles, so their social networks are fairly well connected amongst each other. Once the novelty value wears off, and once people realize that this doesn't really do anything that they couldn't do before (and what it can do is fairly limited), the buzz will die down.