Metcalfe's Law in Reverse

by Jakob Nielsen on July 25, 1999

Current attempts to split the Web into many isolated mini-networks undermine the long-term potential of the Internet which depends on universal interconnection:

  • AOL trying to prevent customers of Microsoft, Prodigy, and Yahoo from communicating with AOL customers through "instant messages". This is reminiscent of email in the 1980s when everybody had five email addresses listed on their business cards; one for each of the main services.
  • Universal Studios fighting movie websites that want to link to film trailers on Universal's site.
  • A similar story last year where Ticketmaster did not want Sidewalk to bring potential ticket-buyers to the Ticketmaster page that sold tickets to events the users had read about on Sidewalk.
  • Owners of sports teams trying to prevent other websites from covering the sport so that fans will be left with no other coverage than the "official" sites sponsored by the National Basketball Association and the like.
  • Cable modem services preventing users from downloading video clips beyond a certain length from sites outside the span of the cable service. In the future, cable services may slow down or prevent any access to unaffiliated sites. The long-term implications is that MP3-like video-on-demand services may not be available to cable modem subscribers except for a small amount of tightly controlled streams from their own servers.
  • The proliferation of distribution deals where portals give featured placement to those services that pay them the most and not the ones they would really want to recommend to their users. Of course, portal users are not customers , but just eyeballs, which explains why user needs get so little respect on the Web today.

All mis-guided attempts to achieve short-term advantage by locking in users or preventing the natural growth of the Internet.

Having other sites link to specific pages (so-called deep linking) is considered attractive by most clued-in Web strategists: many sites run affiliate programs to attract as many inbound links as possible. Users who follow a narrowly targeted link are likely to be hot leads with an interest in buying a specific product or service. It is unproductive to insist that such interested users be dumped at the home page and required to find their own way around the site. Links are one of the most effective Web marketing methods, so it is extremely short-sighted to prevent other sites from pointing to you.

Attempts to build walls around isolated sites will fail in the long term because of Metcalfe's Law which states that the value of a network grows by the square of the size of the network . So a network that is twice as large will be four times as valuable because there are four times as many things that can be done due to the larger number of interconnections.

Because of Metcalfe's Law, the largest network always wins over smaller networks, even if the smaller network has some larger initial value due to some special-purpose feature or benefit. As the networks grow, the square factor ultimately tips the hand in favor of the large network. And since the Internet is the largest network of them all, it will ultimately win over any proprietary network.

Metcalfe's Law provides much of the explanation of the success of the Web relative to earlier hypertext systems like HyperCard, Intermedia, and NoteCards. They were all much better than the Web and had features ten years ago that we are still sorely missing on the Web. But the Web was universal and the other systems were proprietary. You know who won.

The law is usually quoted in terms of growth of the network, but we can run Metcalfe's Law in Reverse and use it to characterize the effect of cutting a network into pieces:

The value of partitioning a network into N isolated components is 1/N'th the value of the original network.

This new law follows directly from the original Metcalfe's Law. Each of the new components has a size of 1/N'th the size of the original network. Thus, its value is 1/(N 2 ) of the original value. At the same time, there are N of these new mini-networks, so the over-all value is N * 1/(N 2 ) = 1/N

The value of the full Web is currently about $300 Billion according to an analysis published by Cisco. In three years, the value will likely be around $1 Trillion. Let's assume that the various attempts to split the Web succeed to the extent that it is split into 5 parts soon and 10 parts in three years.

  • The current value of the Web would be reduced from $300 Billion to $60 Billion - for a loss to society of $240 Billion
  • Each of the current "mini-nets" would be worth $12 Billion
  • The future value of the Web would be reduced from $1 Trillion to $100 Billion - for a loss to society of $900 Billion
  • Each of the future "mini-nets" would be worth $10 Billion

The short-term interest in partitioning the Web lies in the hope of gaining supreme dominance of one of the resulting mini-nets. Capturing most of $12 Billion can surely be more attractive than capturing a small part of $300 Billion, even if your actions lead society as a whole to lose $240 Billion.

In contrast, the long-term prospects of proprietary strategies are dim, with the value of each mini-net dropping over time from $12 Billion to $10 Billion.

Example: Proprietary Instant Message Services

Let us use the Reverse Metcalfe's Law to analyze the potential impact of a proprietary communications system for AOL. Assume that AOL succeeds in capturing 20% of the world market for Internet services. Further assume that the remaining 80% of the market can agree on a single, open standard (hopefully one defined by the Internet Engineering Task Force). Relative to the theoretical value of a single, universal network, we find:

  • The value of AOL's share is 0.2 2 = 4%
  • The value of the other companies' share is 0.8 2 = 64%
  • The loss due to the partitioning is the remaining value that is captured by nobody: 100% - (4%+64%) = 32%

In other words, AOL ends up owning a 4% share (which could be worth a good sum, of course) and a third of the full potential is simply lost to society.

In contrast, assume that AOL joined in the single open standard and then succeeded in capturing the same share of the new value as the hypothetical 20% I have assigned as its over-all share of the future Internet. Then AOL would realize five times as much value as they do under the proprietary scenario. The only downside is that they would have to compete on quality of service to win their 20% share instead of simply being guaranteed a proprietary 4% share.

See reader comments on this Alertbox, including answers to the questions "who is this Metcalfe?" and "what is the true value of the Web?".

I'll present my newest usability guidelines in the tutorial on Fundamental Guidelines for Web Usability at the annual Usability Week conference.

The conference also has a full-day seminar on Emerging Patterns for Web Design.

Share this article: Twitter | LinkedIn | Google+ | Email