What SBC's DSL announcement really means

SBC Communications recently announced that it will be investing $6 billion to enhance the delivery of broadband services throughout its service region. This move was widely reported as SBC spending $6 billion on digital subscriber line (DSL) or Internet access. In truth, that's not what was announced at all, and what's really going to happen may have more impact on businesses than on residential Internet users.

The modernization SBC proposes involves spending $1.5 billion on improved trunking facilities, the connections between switching sites. The goal is to improve the handling of packet services, particularly those based on IP. The remaining $4.5 billion will be spent on what service providers call "outside plant," the gear from your phone connection block inward to the carrier's serving office. This outside plant will be optimized through the introduction of new devices that SBC calls "neighborhood gateways."

Neighborhood gateways support xDSL copper attachments, which could be used for residential Internet access. But does this mean that SBC just plunked down $6 billion on DSL-to-the-Internet? If so, SBC is blind to market reality. The company's regulated business - the one buying all this xDSL stuff - will have to provide equal access to other ISPs. Even if SBC were planning to get into the Internet on a huge scale, it couldn't limit its xDSL customers to its own Internet offering.

So is SBC expecting to get a huge return on its $6 billion from selling access to the Internet only? That's what the Covads and Northpoints do now, and none of them seem to be topping Wall Street charts in revenue. There's got to be more to this, folks.

One of the real justifications for SBC's move is the fact that fiber-based neighborhood gateways limit the regional Bell operating companies' wholesale risk by reducing the number of copper loops that run all the way from the customer to the central office. With the neighborhood gateway in the path, there's no continuous copper for Covad, Northpoint and others to exploit with collocated DSL access multiplexers in the central office.

The second justification is the business user. It's more likely that SBC's digital broadband changes will empower the 5.5 million business sites that don't have T-1 or better services today. It probably also will help the 1.5 million more-or-less-permanent telecommuters. This raises the question of what SBC would expect to sell to these businesses. Not just Internet access, clearly, so what else?

Voice, the generator of 80% of today's total domestic service provider revenue, is one answer, but that again raises the issue of how new voice services might look in terms of real features. SBC's digital infrastructure will be available for wholesaling to competitors. So if competitive local exchange carriers figure out how to create a kind of business "super-Centrex" service set, they could force SBC to respond with its own new-generation voice feature offering. Good news to the new-generation voice players, at least those that have ATM capability.

IP virtual private networks (VPN) have to be the other major service opportunity. Given an ATM access network, the IP VPN service goal elevates the "how to link ATM to IP" issue significantly. That may favor vendors that have announced support for the IETF's Multi-protocol Label Switching standard that, among other things, helps to link the two competing network technologies.

Linking legacy services to an ATM access network is also going to break new ground. The major interexchange carriers (IXC), which all based their new-generation integrated-access services on ATM, obviously knew this was going to happen all along. Maybe SBC's move and the inevitable comparable steps from the other RBOCs will jump-start even Sprint's troubled Integrated On-demand Network.

These IXC offerings only go through the ATM access network to get to the customer. What if the customer wants integrated services and not just integrated access? Will ATM access pull ATM transport networks along just to make the connection more efficient? If so, Williams Communications' decision to base its network on ATM looks really smart.

SBC can make its new access infrastructure pay for itself on business. So what about the Internet? Well, with the return on its $6 billion guaranteed by business, perhaps SBC will create a test bed of broadband residential users that can be used to explore what people really will be willing to pay for good Internet service - from SBC, of course. If well-heeled power-Webnicks respond to upscale Internet offerings, SBC might even make the Internet profitable.