Metcalfe’s Law, Meet Market Reality


Bob Metcalfe is one of the super smart people I have met as part of being a reporter with technology beat. The inventor of ethernet and 3Com is more often than not hits the bulls eye when he talks about network-related issues. Back in the day he said that usefulness, or utility of a network equals the square of the number of users. This became known as the Metcalfe’s Law, and has been the bedrock of this crazy ever growing organism called the Internet. As a theorem when applied to the network economics, it proved to be a be a good barometer of success of a technology, or a product or a service.

However, Metcalfe’s Law has never been tested in a situation where the network slows to grow – after all the Internet use is still growing. Today was a little different – EBay, a shining example of Metcalfe’s Law – failed to meet its financial goals and the stock markets punished it with a 20% plunge. (By the this might be the costliest penny on the planet – missing the profit estimate by a penny resulted in $17 billion in market capitalization.) In other words, this was the first time Metcalfe’s Law has been put to test in market reality environment. The EBay debacle makes me wonder if their network has reached its theoretical limit? So much so that despite being a near monopoly the growth has become difficult to sustain.

“Is this it?” asked Piper Jaffray analyst Safa Rashtchy in a research note following the earnings report, downgrading the stock to “market perform” from “outperform.” Mr. Rashtchy said it’s still unclear whether the slowdown in the fourth quarter was a one-time event or indicative of difficult structural issues, but said the shortfall increases eBay’s risk profile. (Wall Street Journal)

Now if we look at other online business models which are predicated on hyper growth of the network, aka Metcalfe’s Law, one wonders at what point will a Google, a Yahoo or even an Amazon hit the wall – and face the problems most offline retailers and rust belt companies face! If that is the case, then from an investor perspective, I am sure the highest returns are just after the early growth and first signs of maturity. I know I am rambling here, but this is all too curious. I have pretty smart readers, including Paul K, who would think more deeply about it, and hopefully post something more meaningful.

PS: 1. Suddenly this story in the, about being a better play than EBay starts to make sense!

PS: 2. By the way, 9 days ago, I wrote a little piece here about why EBay was raising its prices: “Also there could be another reason: slowing growth and unrealistic expectations are forcing E-Bay to jack up the prices.” A moment of indulgence. -)


Bill Hartman

Work the numbers here. If a network triples in size, it becomes 9 times as valuable, the square. But if it grows slowly compared to the existing size, it increases in value at essentially twice the number of nodes added (1% increase in size means 2.01% increase in value), which is multiplicative. Not so great, but hey, each addition adds twice its cost, so the whole gains twice the value added. More important, each 1% addition is 101 times as valuable FROM ITS OWN PERSPECTIVE than if it were to be created as a separate network instead. So the economic incentive to join the network is larger for a large slowly growing network than a small quickly growing one, unless one presumes that the small one will eventually outgrow the large one. Which is to say the attraction to join should continue to increase, not decrease, when growth becomes less extreme.

Comments are closed.