Business Life: The value of a network
First published in Business Life, January 2009
What is Facebook worth? The social networking website’s apparent value has fluctuated wildly: only founded in 2004, a deal with Microsoft in October 2007 seemed to value the company at about $15 billion; by the summer of 2008, Business Week was reporting that shares were being sold privately at prices implying that Facebook was worth less than $5 billion.
Business fashions come and go. But the uncertainty over the valuations of networking sites such as Facebook and MySpace is not surprising to an economist. Networks have a curious and controversial economics.
The value of a network, presumably, is something to do with the number of connections it makes possible. It is not hard to count them. A computer whiz called Robert Metcalfe even proposed, back in 1980, what later became known as “Metcalfe’s Law”: the value of a network is proportional to the square of the number of users (or devices) on that network.
Metcalfe reasoned thus: A single fax machine is useless. A pair of fax machines can make a single connection, and there are three ways in which a trio of fax machines can talk to each other: connections between A and B, B and C, or C and A. Four machines can connect six ways and five can connect ten ways. Because each new fax machine can talk to any of the existing fax machines, each new fax machine is worth more than the last.
Metcalfe was no slouch: he invented the Ethernet standard for computer networking and then when on to become a high-tech entrepreneur and venture capitalist. But while the mathematics of Metcalfe’s law is indisputable, its economic logic is debatable. That matters, because while Metcalfe conceived of his law as applying to small Ethernet networks, it has often been uncritically applied to much larger networks – notably during the dot-com bubble.
The problem is that not every connection is equally important. The first telegraph linking Europe and the US was hugely valuable. The ability to fax the next cubicle is not.
When Microsoft bought a slice of Facebook, Facebook had about 90 million users. Four thousand trillion two-way connections were mathematically possible, but few will ever be useful.
For economists, the value of a network also depends critically on the ease with which people can switch to another network. Fifteen years ago it would have been inconceivable for everyone to scrap their fax machines and switch to a new, incompatible generation of faxes. Yet it is not so hard to imagine Facebook’s users deserting in droves for the next big thing in social networking. And a network that can be abandoned overnight is a network whose value can be destroyed overnight, too.