About Tom Evslin

Video Profile of Tom Evslin

Follow Tom Evslin on Twitter


Add to Technorati Favorites!
Powered by TypePad
Member since 01/2005


« It’s Time to Go Nuclear | Main | As the Phone World Turns Part 4 – Building Network Value with Skype »

As the Phone World Turns Part 3 – Metcalfe’s Law

Bob Metcalfe, the inventor of Ethernet, is also credited with Metcalfe’s Law: the value of a network is proportional to the square of the number of endpoints.  Making Metcalfe’s Law work for you (or not) is the key to success (or not) in establishing a communication services company. A network with 50 endpoints is 100 times as valuable as a network with only 5 endpoints.  Huge networks have huge value.  But, and this is the problem, small networks have infinitesimal value.

The success of Internet phone companies like Skype and Vonage depends on how successfully they harness network effect so it’s important to understand how Metcalfe’s Law operates before looking at these two companies’ very different strategies.

Suppose you invented the world’s first phone.  Who’s your first customer?  There’s nobody to talk to.  This was a real problem after Alexander Graham Bell’s invention.  The first market for phones was the kind of closed network that a company might establish to link its offices because a single purchasing decision to buy a bunch of phones established an instant network and instant value.  A residential market didn’t develop until there were already some phones to talk to.  Then it developed a town at a time.  But value mushroomed when the towns were interconnected through a hierarchy of human operators and switchboards.

The math is intuitive.  The value of a network is actually proportional to the number of connections you can make using it.  To keep things simple, let’s assume these are all two party connections between people.  If you have one person in your network, she has nobody to talk to and the network has no value to her.  If you have two people, they can talk to each other so that’s one possible connection.  If you have three people, there are three possible two party connections.  But, if you have four people there are six possible connections; with ten people, 45 possible connections, 100 people means 4950 possible connections and so on.  If you like formulae, the number of unique two-party connections is (n)(n-1)/2 where “n” is the number of endpoints.

It took one hundred years between the time that fax was invented and people started asking “what’s your fax number?” rather than “do you have a fax machine?”  It took that long because of Metcalfe’s Law.  Fax machines can only talk to other fax machines and so there had to be lots of small networks set up before there was any value in buying a single fax machine.

It took roughly 20 years to go from “what’s email?” to “do you have email?” to “what’s your email address?”  Evslin’s corollary to Metcalfe’s law is that network nirvana is reached when it is assumed that everyone you meet will have an address on that network.

THE reason for the enormous value and success of the Internet is that it is a network of networks.  When networks are connected, there are huge increases in network value.  We can prove it with Metcalfe’s law.  If there are 10 networks with 100 people (endpoints) each which are NOT connected to each other there are 4950 possible connections on each network or 49,500 possible connections altogether.  Now suppose you connect these networks so that everyone can reach everyone else regardless of which network he or she is attached to.  Now there are 499,500 possible connection.  There is a ten-fold increase in value from connecting the networks even though NO new subscribers were added!  That’s what the Internet did by connecting tens of thousands of networks for email and other functionality.

Let’s get back to starting a communication company.  You cannot escape Metcalfe’s Law so you have to figure out how your service can get to having a large enough number of possible connections to attract anyone as a customer.

The easier approach is to make sure your network connects to an existing network.  That’s what mobile phone people did, for example.  It was relatively easy to sell the first mobile phone because it could connect with everyone who had a landline phone.  You didn’t have to wait until someone bought the second mobile phone to have someone to talk to.  Notice that features like text messaging and picture phones couldn’t evolve until there actually were a critical mass of mobile phones which supported them. (Why such features never evolved on landline phones is all about monopoly and regulation, the world’s most effective antidotes to innovation.)

Instant messaging, on the other hand, was an entirely new network.  ICQ (I Seek You), the first Internet instant messaging network, was invented in Israel by Yossi Vardi.  The software was available for free download and enough people downloaded it to create network value.  Yossi is famous for saying ''creating revenue is a big distraction."  He sold his company Mirabilis to AOL for close to $400 million at the beginning of the first dot.com boom.  It had never had a penny of revenue but it did have network value.

The first approach, add on to an existing network, is the safer one and can have revenue immediately.  But it doesn’t create any proprietary network value.  The second vendor of mobile phones used to connect to the landline network is not disadvantaged by the fact that the first vendor already has some subscribers since those subscribers can be reached from any phone.  That’s why mobile companies have to create artificial network value by providing “free” onnet calling.  This approach is the one Vonage is taking.

The second approach, create a new network, is far riskier.  If you want to reach critical mass within a single lifetime (unlike fax), you have to give initial access away since it has so little value when the network is new and small.  The Internet does, however, make it very, very cheap to distribute whatever is needed for access and MOST IMPORTANT, allows a new network of clients to be connected without any incremental investment in infrastructure. Successfully creating a network with millions of users who can all connect to each other is a home run because of the enormous value of the billions of possible connections.  This is Skype’s approach.

More specifics on Skype and Vonage strategies next week.

The first post in this series is everything you ever wanted to know about legacy access charges.

The second is about the cost of “free”.

The fourth is about Skype’s success in building a closed network.

The fifth is monetizing Skype’s network value with SkypeOut.

The sixth is about Skype reproducing the OLD telephony business model with SkypeIn.

The seventh is a summary of Skype features.

The eighth begins coverage of Vonage’s strategy.

A related post contains a very short abstract of what Skype founder Niklas Zennstrom said at VON (Voice On the Net) Canada and a way to download the slides of my talk there.


TrackBack URL for this entry:

Listed below are links to weblogs that reference As the Phone World Turns Part 3 – Metcalfe’s Law:



I'm quite impressed with the "math sense" that informs Metcalfe's Law. While I'm no mathematician, it seems logical to me, and I don't see any flaws in the thinking that went into this.


It didn't take one hundred years for the Fax machine to be widespread, it took very little.
There is an extremely well written article about it from Clay Shirky

"Zap Mail" have a look, I really love it

You could see in it also a very good forecast of the Future of VoIP.

As he says, what they didn't understand was that faxing was not a service, it was a product.

Exactly what VoIP is.
In spite of all the efforts to make it a service, like the old telephone system, VoIp is a product, it is an IP phone, and the infrastructures are already there, the Network is already there, it is called "Internet" and nobody should pay a further fee to have access to it.
The moment everybody will realize it (and believe me Metcalfe's law doesn't cover the sudden boom and the power of "Mouth to mouth") there won't be any more the need of a "service" to connect to the PSTN.

Clay Shirky has a great vision of the Market because he has a great vision of the human behaviour.

No computer, no mathematical law is able for the moment
(may be it will be someday, I do not want to put bounds to the artificial intelligence)to come to the same conclusions as a human brain, and also to understand the unpredictable of the human brain and consequently of the human behaviour...
But history tends to reapeat itself, because histroy is made by humans, and humans always repeat themselves..



First of all: the inventor of the telephone was Antonio Meucci, an italian engineer working in Cuba.
The only approach from Bell to this invention was his corruption of the Patent people and the theft of the invention.
This is all well documented and only very late the Americans admitted the truth, also if in a very "silent" way, in order not to damage the universal convinction of Bell being the inventor...

Please, have a look on the Internet, you will find the full story with all the details...

Regarding the small network in the network, that is the only possible way to begin.
You cannot begin with one million, you always begin with two.
But the good start is always with a clear vision of what you want to build.
And I think Vonage and Skype and the millions of others really have a clear vision of what they want to build.
Allow me to have a different one.

In Europe the cellular phones have reached sooner a much higher number of users than in US and not only because the Europeans are crazier when it comes to telephones, but because they have a common unique standard and in US to reach the same result you need a triple band phone.

I think it shouldn't happen the same with VoIP.
I wouldn't like to have an IP phone with all the existing codecs to be able to connect to all the VoIP providers.
At the end of the day it would be cheaper to use my analog phone, especially if I do not talk too much.

Use Skype who already has millions, one could say...

25 years ago IBM adopted MSDOS for its computers and later Windwos, making thus the fortune of IBM and Bill Gates'.
Now a day people think it was not really such a great idea...

A Monopoly is like a tyrannic state.
If you have a great chief it could have its advantages, but it is very dangerous to be in the hand of one.
It can dramatically change depending on which foot he wakes up in the morning.

We already saw and felt the consequences of a Monopoly.
Are we willing to destroy one to build another?
They say if you are not able to see your mistakes, you are condemned to repeat them again and again...


DG Lewis

Odlyzko and Tilly have a paper out (http://www.dtc.umn.edu/~odlyzko/doc/metcalfe.pdf) that claims Metcalfe's Law overstates the value of connectedness and that value of a network grows like n*log(n), not n^2. One implication is that interconnection of two networks is more valuable to the smaller network. This is consistent with your mobile phone example - the mobile phone network operator gets more value from being able to connect to the PSTN than the PSTN operator gets from having mobile phones connected to it.

My take on the implications of this for Skype is something I'm thinking through for a future article...

Post a comment

If you have a TypeKey or TypePad account, please Sign In.

Now on Kindle!

hackoff.com: An historic murder mystery set in the Internet bubble and rubble

CEO Tom Evslin's insider account of the Internet bubble and its aftermath. "This novel is a surveillance video of the seeds of the current economic collapse."

The Interpreter's Tale

Hacker Dom Montain is in Barcelona in Evslin's Kindle-edition long short story. Why? and why are the pickpockets stealing mobile phones?

Need A Kindle?

Kindle: Amazon's Wireless Reading Device

Not quite as good as a real book IMHO but a lot lighter than a trip worth of books. Also better than a cell phone for mobile web access - and that's free!

Recent Reads - Click title to order from Amazon


  • adlinks
  • adsense