Metcalfe's Law: Double the Network, Quadruple Value

Metcalfe's Law says a communications network's value grows with the square of its connected users (n²), because the number of possible pairwise connections does. Add users linearly and value rises quadratically, so doubling a network roughly quadruples its worth. This is why each new member benefits everyone already connected.

The counting trick behind the square

The idea rests on a simple combinatorial fact. In a network of n members, each person can connect to the other n−1 people. Counting every unique pair without double-counting gives the formula for the number of links.

connections = n × (n - 1) / 2

This is the triangular number, and for large n it behaves like n²/2. That n² term is the heart of the law: value is assumed proportional to the count of connections, and the connection count grows like the square of the user count. You can watch this happen by feeding a user count into the network value calculator and reading off both the link total and the relative value.

Here is why the growth feels so dramatic. Each new arrival does not add one unit of value; they add a connection to every single person already present. The 101st user on a 100-person network creates 100 brand-new links all at once. That accumulating marginal benefit is what makes the curve bend upward so sharply.

10, 100, 1,000: watching value explode

Concrete numbers make the scaling vivid. Plug a few sizes into n(n−1)/2 and the jumps are striking:

Notice the pattern between consecutive rows. Going from 10 to 100 users (10x more people) takes connections from 45 to 4,950, roughly 100x more. Going from 100 to 1,000 multiplies links by about 100 again. Tenfold growth in members produces hundredfold growth in possible links. That is the square law in action: multiply users by k and value scales by roughly k².

The headline phrasing follows directly. Double your users and (2n)² = 4n², so value quadruples. A competitor with half your user base does not have half your value; under Metcalfe's Law they have only a quarter of it. That asymmetry is exactly why network-based businesses fight so hard for scale.

Fax machines, telephones, and social networks

The classic illustration is the fax machine. A single fax machine is a useless slab of plastic; it has nobody to call. The instant a second one exists, one connection becomes possible. The millionth fax machine is enormously valuable precisely because it can reach the 999,999 machines already deployed. The value lives in the connections, not the device.

The law is named for Robert Metcalfe, who co-invented Ethernet with David Boggs at Xerox PARC in 1973 and went on to found the networking company 3Com. He first sketched the idea around 1980 as a sales argument, framed in terms of compatible communicating devices like telephones and fax machines, urging businesses to connect more equipment. It became associated with users on the network after technology writer George Gilder publicized it in a September 1993 Forbes ASAP article, attaching Metcalfe's name to the now-familiar n² form.

Modern social platforms are the textbook case. A messaging app with ten friends is mildly useful; the same app with a billion accounts is nearly irreplaceable, because almost anyone you'd want to reach is already there. This self-reinforcing pull, where growth makes a network more attractive and attractiveness drives more growth, is the network effect that lets dominant platforms entrench themselves even against technically superior rivals.

The criticism: maybe it's n log n, not n squared

Metcalfe's Law has a serious and respected objection. In a 2006 IEEE Spectrum article titled "Metcalfe's Law Is Wrong," Bob Briscoe, mathematician Andrew Odlyzko, and Benjamin Tilly argued that n² wildly overstates real network value. Their core point: the formula assumes every connection is equally valuable, but it isn't. Your link to a close friend matters far more than your link to a stranger three continents away you'll never contact.

Because the most valuable links tend to form first and later ones add less, value runs into diminishing returns. The authors proposed that network value scales closer to n × log(n), which still grows faster than the user count but far more modestly than the square. They warned that the inflated n² estimate had helped justify the reckless growth-over-profit thinking of the dot-com bubble. The debate is not settled; a 2013 analysis by Metcalfe himself fit the n² model surprisingly well to Facebook's user and revenue data, and later papers have pushed back on the critique.

One precision point worth nailing down: n² is quadratic growth, not exponential growth. People often misdescribe Metcalfe's Law as exponential, but exponential means the user count sits in the exponent, like 2ⁿ, which grows incomparably faster. A square law is a polynomial; it is fast, but it is in a completely different and gentler league than true exponential growth. To compare how different growth rules behave, you can experiment with values in the network value calculator, and curious readers exploring other counterintuitive scaling rules may enjoy Zipf's Law, which governs how word and city-size frequencies fall off.

Whether the true exponent is 2 or something closer to n log n, the qualitative lesson holds: connected networks gain value faster than they gain members, which is why connectivity has been one of the most powerful economic forces of the last half-century.

Frequently Asked Questions