Could ‘open DSL’ save the ISPs?

When the Telecom Act of 1996 was signed, there were more than 3,000 ISPs, a half-dozen of them major providers. Today, there are fewer than 1,000 ISPs, and it’s arguable that no major ISP remains outside bankruptcy protection. Are the predatory regional Bell operating companies, seemingly preparing to take over the carrier world, eating the ISPs alive? Actually, they might unintentionally be throwing the ISPs a lifeline.

The RBOCs need to promote DSL, with or without favorable regulatory treatment. With cell phones eroding second- and even first-line voice market share, the copper loop is in danger of becoming irrelevant in voice communications. For the RBOCs, having the loop become useless is worse than having to share it via wholesaling and unbundling. So DSL is here to stay.

The issue is how DSL can make money. AOL, the largest retail provider of Internet service, is clearly no financial powerhouse, and it provides only dial-up speeds. The average price spread between AOL’s service and DSL Internet access is only about $20 per month. Even assuming good economies of scale, it’s hard to get the return on investment (ROI) for DSL alone to come out above 11%, and that’s probably 50% too low for financial success. Moral: You’ve got to sell something over DSL. That’s where the ISP opportunity comes in.

AOL’s experience shows that the Internet alone probably can’t jack up returns enough. The question is how the concept and technology of the Internet can be extended to earn more revenue.

There should be no one in the industry that knows more about the Internet, its customers and its economics than the ISPs. They should have taken the lead in developing commercial applications that would generate incremental revenue from Internet services. Instead, they sat back and counted customers while companies such as Microsoft, IBM and Cisco developed Internet applications that generated profits for them but not for the ISPs. Think about it: Practically all the so-called Internet applications, including voice on the ‘Net, are based on the assumption that users pay nothing for the application except their standard Internet access fee.

The RBOC DSL positioning will give the ISPs another chance. The Federal Communications Commission clearly is going to relax unbundling rules to permit the RBOCs to accelerate DSL deployment, which will make it nearly impossible for ISPs to provision their own DSL to the user. But who cares? The ISPs should let the RBOCs figure out the technology of DSL and focus on projecting profitable services over DSL.

Most services that buyers are interested in fall into three categories: application service provider (ASP), VPN and content services. There are many hosting companies selling ASP services. Suppose an ISP packages ASP services such as Exchange hosting with VPN services and enhanced quality of service? Suppose an ISP that blocks VPN usage from residential portals sells a quality- and availability-enhanced VPN offering? Suppose an ISP creates support for interactive gaming hosting providers by offering per-game VPNs? Or maybe hosts games and other content itself?

One important step in this enrichment process is to stop trying to make the Internet the zero-cost alternative to every other kind of service. You can’t charge for something you’re already giving away. I can do VPNs and voice over IP on my broadband connection without paying a nickel extra. Would I pay $5 per month to keep that capability? Sure.

The trouble is that the ISPs are still living in the 1990s, when Wall Street valued an ISP by the number of customers it had and not its profit or ROI. The ’90s are over, gang. ISPs have to create a profit model for their services. They have experience with the problem, at least by having experienced what doesn’t work. The RBOCs have no experience at all. If the ISPs move quickly, they can gain an important lead in a critical area – profitable residential and business IP – before the RBOCs figure out how to be players in the Internet game at all.

The ISPs need to do this, because the RBOCs, like the proverbial mills of the gods, grind slowly but very fine. ISPs can be ground up, too.