Rethinking Metcalfe’s Law
Enter a room of investors and entrepreneurs and it would be hard to leave without having heard the term network effect. It’s network effect this, that, and everywhere.
Network effects have been considered the be-all and end-all of business models. In the “if you could pick only one” of competitive advantages, many would scream for network effects. Incubators and venture investors have decided to exclusively look for network effect businesses and I don’t blame them.
Network effects reference Metcalfe’s law to explain the concept. It’s not a real scientific law. It’s a law from empirical observation that states an additional node in the network has exponentially greater value. So 5+5 doesn’t equal 10 but something like 20.
Think of it like Facebook. It got more valuable to you when more of your friends went on it.
Scaling networks can appear limitless thanks to the internet. Combined with human optimism and exuberance, we can argue for Metcalfe’s law taking effect in all kinds of companies (i.e. network of homeowners, car-owners, dog food, some bloody software). But what if people are basing everything on an incorrect understanding of an empirical law?
The article that inspired this thought is here.
Consider the effect of Ziph’s law, which is what the Long Tail is about. Practically, the 1st place is valued at K and the second place is valued at 1/K…so it’s worth half as the 1st and so it goes. This shows a view that each additional increase to the network is less variable than the first (i.e. the 80/20 of book sales, songs, movies).
Ziph’s law (i.e. the long tail) is why e-commerce stores will thrive online as niches. It’s why Amazon’s limitless online catalog trumped the physical limitations of brick and mortar shelves. Amazon was able to capture every niche audience not covered by the 80/20 selection of brick and mortar.
It’s not that the 1000th item had zero value. Physical stores just couldn’t justify the opportunity cost of shelving niche products. But online businesses with zero inventory costs can.
Metcalfe’s law is predicated on the belief that each additional growth in the network is considered exponentially more valuable (about a quadruple increase in value for every doubling). But the question is whether each network is of equal importance.
We have weak networks and strong networks in our lives. That’s the same for a community. Some of the networks the platform crated will be valuable and some won’t. They aren’t all equal.
Think about a platform business like Facebook with the view of Ziph’s law over Metcalfe’s law. The first 20-50 networks (i.e. friends or followers) may make up the majority of the value for the network with continuously diminishing value for every additional network. Now, it doesn’t mean they are devoid of value and the platform’s ability to supply you with a near-endless supply of possible people to add to your network is valuable. But that’s to a degree as each addition will be less and less valuable to you.
That is made clear by Ziph’s law. But people forget this phenomenon when looking at tech companies and saying they have the chance to get massive because of Metcalfe’s law. There is always a limit to how valuable the network can be.
If each additional network in my life is log(n) instead of n^2, wouldn’t that be applicable to the overall platform as well? Not all networks have the same size and they have a different point of diminution.
Metcalfe had first used the law to showcase a crossover point where the network could become large enough for it to be sustainable. But after reaching this point, it seems irresponsible to think the curve will continue to hockey stick up instead of curving out like an S-curve. All good things come to an end and that’s the wonderful thing about them. Some will last longer than others but it feels we assume too many will last a while.