Tuesday, April 18, 2006

Building network economics

I've been working with a number of start-ups in the NYC area recently, and most of them are launching community based sites, hoping to capture some of the value created by networks of users. While most of these ideas are interesting once the network is built, the one part that seems to be overlooked in many cases is how to build to that critical mass.

Most people talk about the network value of successful sites. However, before a site is successful, the network value is much smaller than the non-network value. Until a critical mass of users exist, the only value to a new user is the "selfish motive". This is the reason you are willing to be the very first person on a community site, even if you think no-one else would ever join the community.

A great example is flickr - one of the poster childs of the web 2.0 wave. Now that the site has a great community, and active content posters, the value of the community feeds itself, and helps them bring in even more people. But at the beginning, when there were no other people uploading and tagging photos, they still had something useful - an intuitive and quick way to upload and share photos - that was the selfish motive for people to use the service.

With that in mind, I think it's important to clearly define your customers' selfish motive for using your service/site/network, and at the beginning, THAT needs to be your primary marketing message. Only once you achieve critical mass can you switch to really pushing the value of the network itself. Push the network value too early, and you risk people expecting more than they find, and not sticking around to help build the network itself.

Warning - equations and low level geekspeak below!

A few months ago, there was a lot of attention paid to Metcalfe's Law, and later, to it's big brother, Reed's Law. Both of these talk about how the value of a network grows faster than linearly with the number of people on the network (or correspondingly, in the community). Metcalfe's law is usually quoted as saying that the value of the network is N^2, where N is the number of members of the network.

Now, this is an approximation based on the assumption that N is large. More precisely, Metcalfe's Law states that the value to each member is: a + b*N, where a is MUCH greater than b. Therefore, the total value is simply

N*(a+b*N) = a*N + b*N^2

For networks with a large number of N the value to a new member is driven much more by the b*N portion of the value, than by the a. However, for those first early adopters, there absolutely has to be a selfish reason to join - the a.

Update: This post had been in draft form for a while, and it looks like Tom Evslin and his readers beat me to the punch, both in time and in content.