This article appeared in the November 27, 2000, issue of the Industry Standard.

The essential beauty of the Internet – its openness – is also its essential flaw as a platform for business. Companies achieve profits by erecting barriers around markets, locking customers in and keeping competitors out. But since the Internet was designed for communication rather than commerce, its structure frustrates attempts to raise barriers. That’s why one dot-com after another has flopped, despite massive investments to build brands, seize first-mover advantages and acquire customers.

But the days of the open Internet may be numbered. A lot of companies are working hard to fence off areas of the Net – to, in essence, unweave the fabric of the Web. They know that if they can use the Internet’s infrastructure to create private networks, they’ll have the chance to be the toll collectors of the new economy. And that’s where the real money lies.

Look at Napster. At first glance, peer-to-peer networking seems to represent the apotheosis of openness. Everyone’s connected to everyone else; no central authority exerts control. But that’s an illusion. Peering, at least in the Napster model, is a closed system. You need to download proprietary software, sign up for membership and connect to a proprietary set of servers – and you can communicate only with other members.

Napster has been able to maintain its rebel image by giving away its service (parading Shawn Fanning around in his Red Sox cap doesn’t hurt, either). But as Napster’s recent deal with German media giant Bertelsmann makes clear, giving kids free access to tunes is not the company’s ultimate business model. Bertelsmann and Napster want to entice the other major record labels to join them in turning Napster’s free service into a fee service. The recording companies’ hope, no doubt, is that by creating a closed system for online music distribution, they’ll end up with more control over the music business than they have today.

Other companies also hope to use peering technology to create closed systems. The secretive startup InfraSearch (its Web address is Gonesilent.com) wants to use its p-to-p search software to create private knowledge-sharing networks. You’ll pay a fee to exchange information with other members.

And the drive to partition the Net isn’t limited to p-to-p. Caching companies like Akamai are building vast systems of proprietary servers – private networks within the Net’s public network. Instead of charging consumers to access their servers, they’re charging Web sites and media companies to store and deliver information. While Akamai and its competitors have so far concentrated on distributing content, their desire is to use their private networks as platforms for a host of services.

Microsoft also wants to play in this game. It first succeeded by gaining control over the PC desktop, establishing its operating systems and applications as gateways and billing consumers and companies for the right to go through them. (Indeed, the company’s entire strategy is encapsulated in its founder’s name: Bill Gates .) It is now trying desperately to gain similar control over the Internet. That’s why it keeps retooling its MSN software, and why it’s pouring money into its .Net initiative to become the dominant provider of online applications.

In a recent interview, Tim Berners-Lee, the inventor of the Web’s hypertext language, expressed hope that the Web would not end up being chopped into competing enclaves. “There is always an incentive for one company to try to … change standards and leave other companies inoperable,” he said, “but there’s a tremendous incentive for the community as a whole to prevent that.” We’ll soon see how strong those contending incentives really are.