Conventional wisdom is that a remarkably powerful effect known as Metcalfe's Law is driving the growth of the Internet. The law says that the value of a network grows in proportion to the square of the number of users, which means that, once a network achieves a certain size, it becomes almost irresistibly attractive. But Metcalfe's Law actually understates the potential value of the Internet, and by a huge margin.
I'd like to suggest a new way of looking at the economics of the Internet. I think my approach can explain why forecasters have so consistently underestimated its growth. (And, believe me, they have: In 1995, estimates for on-line commerce in 1998 were $2 billion to $3 billion, while the real number turned out to be more like $13 billion.) My approach not only should add to the urgency that businesses feel about moving on-line but also helps identify which on-line strategies will work and which will fail.
It helps to first understand the two laws of networks that have been around for some time. First is what might be called Sarnoff's Law, after the pioneer of the broadcast industry. This law says that the value of a network grows in proportion to the number of viewers.
Second is the law named after Bob Metcalfe, the inventor of the Ethernet computer-networking technology. He reasoned that 1,000 people on a network can have roughly one million different conversations, so he said the value of a network grows in proportion to the square of the number of users. The n² value explains the growth behavior of networks, such as phone systems or electronic-mail systems, that are mainly used for one-on-one communication. The n² effect says that, given the choice of joining a large existing network with many users or an incompatible new one with few users, new users will almost always decide that the bigger one is far more valuable. The result is often explosive, accelerating growth once a network establishes dominance. This behavior explains why there is now one global e-mail system, while just a few years ago there were thousands.
There's an additional law at play with the Internet because it facilitates the formation of groups, in a way that Sarnoff and Metcalfe networks do not. The number of groups that can be formed over the Internet isn't the Sarnoff n or Metcalfe n². It's 2n by the time you add up all the possible two-person groups, three-person groups, etc. So, the value of the Internet grows in proportion to 2n. Let's call this effect the Group-Forming Law.
This law is so powerful because 2n gets impossibly large very fast. There's the old story about the king who rewarded a wise minister by offering him anything he wanted. The minister said all he wished for was two copper coins on the first square of a chess board, four on the second, eight on the third, and so on—a progression based on 2n. The king protested that the minister should ask for gold or pearls, not copper. But, by the time the 8,192 coins were placed on the 13th square, the king realized he'd been had—264; is more than 18 quintillion, which, if memory serves, is more grains of sand than exist in the world. (The story has it that the king had the minister beheaded for being a wiseguy.)
All three laws, in fact, apply to the Internet. Services such as news sites that are aimed at individuals benefit from additional users in a linear, Sarnoff way. Services aimed at facilitating transactions, such as many commercial sites, benefit in an n² Metcalfe way. Services aimed at building communities, such as AOL, benefit in a 2n, Group-Forming way. What's important is that the dominant value in a typical network tends to shift from Sarnoff to Metcalfe to Group-Forming as the scale of the network increases. So, as the Internet continues to expand, investments in Group-Forming networks are likely to produce the biggest returns.
As the scale increases, what's important also shifts. When Sarnoff's Law dominates, content such as TV programs is king. When Metcalfe's Law kicks in, transactions are king. When the Group-Forming Law takes hold, communities are king. The value in a Group-Forming network is constructed jointly, whether through discussion groups, through joint plans to buy something in bulk at low prices, or through some other means.