31 May 2005

Metcalfe's Law overshoots the mark

Early in March, the New York Times published an article - Researchers: Metcalfe's Law overshoots the mark - about a paper written by 2 researchers refuting Metcalfe's Law, which is quoted as the rule of thumb for computing the value of communication networks. I recently read their academic paper again because of some work I am preparing for a client and a conference.

Andrew Odlyzko and Benjamin Tilly - researchers from the Digital Technology Center at the University of Minnesota - posit several arguments to substantiate that this rule is a "significant overestimate". Their paper - "A refutation of Metcalfe's Law and a better estimate for the value of networks and network interconnections" - explains why network growth rate is much slower than Metcalfe thought. And...why the dot-com and telecom boom failed.

They reviewed several quantitative models of networks...some dating back to 1974...and compared them to each other. Metcalfe's Law seemed to have an intuitive appeal, and people used it generously to project growth calculations for networks and to demonstrate "a linear relationship between the perceived size and value of a network".

I always wonder why business people grab onto "rules" like this without understanding their foundation, and without understanding the consequences of adopting it into practice. The sheer number of consultants and sales people using a "rule" like this and quoting it can turn it into a standard of practice, which is exactly what happened. This is what I call a borrowed strategy.

Perhaps it time to rethink our business models and make them unique to our businesses...and not borrow them from somewhere else.

No comments: