Covad Communications Group, the nation's largest independent DSL provider, announced in February that it would be closing 260 central offices. And Covad's not alone in its DSL woes; rival NorthPoint Communications has already filed for bankruptcy protection, while Vancouver, Wash.-based DSL provider GST Telecommunications shut down its operations completely.
What gives? You'd almost believe nobody wanted DSL — and yet we know that isn't true.
According to Nielsen/Netratings, the number of residential Internet users with broadband connections in December 2000 was up 148 percent from December 1999, a bigger jump than any other type of access. And a survey by The Strategis Group shows that among new broadband installations, fully 60 percent are now DSL lines.
It makes sense. After all, we've known for years that science has hit the theoretical bandwidth ceiling for old-style modems. Way back in November 1998, PC World magazine trumpeted that 56Kbps V.90 modems would be "the last analog modems you'll ever buy." Traditional phone circuits just can't handle any greater throughput.
DSL circuits, on the other hand, are tailor-made for data transfer. Their all-digital design can push bits four times as fast as analog modems, and that's on a really slow day. Plus they run over the same standard phone wiring — so it seems obvious that if you've already got a phone line, DSL is the way to go.
So why, in 2001, are we still buying analog modems? Veteran modem maker U.S. Robotics isn't merely continuing its existing V.90 product line. The company is readying an entirely new analog modem technology for release this year.
But don't expect any real performance gains. While the new models will connect more quickly and upload speeds will be slightly improved, downloads will still max out at the same old 56Kbps. Truthfully, the new hardware's most promising feature is that it supports putting a data connection "on hold" for Call Waiting — provided, of course, that your ISP doesn't opt to hang up while you're on the other line.
Never mind how lackluster the improvements may be, though. The fact that U.S. Robotics continues to spend R&D dollars on analog modems, fully three years after PC World announced "the end of the line" for this technology, ought to tell you something.
Sure, being able to pump data traffic down standard phone wires at near-T1 speeds is a revolutionary idea. But from a business perspective, all isn't as rosy in the "DSL revolution" as the marketing may have you believe.
DSL providers will claim that their current financial woes only represent temporary setbacks. Their chief problem over the past year, they'll say, was that a few bad customers didn't pay their bills. That's true, but they were pretty big bills that went unpaid.
Though consumer demand for DSL is high, companies such as NorthPoint and Covad don't market their services to individual consumers directly. Instead, their business is selling DSL circuits wholesale to ISPs, who then offer the lines to their own customers as part of Internet access packages.
With the recent slowdown in the technology sector, however, many of those ISPs' best customers have run into financial difficulties themselves; or in some cases, they've disappeared completely. That's left many ISPs unexpectedly strapped for cash — cash that might otherwise have gone to their DSL providers.
Still, you'd think a well-managed, well-financed DSL outfit ought to be able to weather a few such hardships. And indeed Covad is already taking measures to control its losses, having announced it's dropping some of its most delinquent accounts.
But shedding some dead weight alone won't stop the hemorrhaging of money from Covad and its competitors. Provisioning and deploying DSL ain't cheap. It's so expensive, in fact, that telcos offering the service have been spending almost four times their revenue on operating costs.
A lot of that money goes to hardware, including the expensive switches and routers that manage the flow of digital data. But ultimately, the No. 1 expenditure for each DSL installation is the manpower it takes to make it happen.
According to research published by Infinilink, DSL providers must send technicians to a customer's site 2.7 times on a typical installation, with each trip averaging over an hour. After that, the customer might make two calls to technical support during the configuration process. All this can cost the provider as much as $500 per customer, up front.
That might be all right, were it not that the profit margins on DSL are so low. The problem is that right now, the DSL business isn't much more than a land grab. Independent players such as Covad and NorthPoint must compete with local telephone carriers like Pacific Bell, who are offering customers deep-discount incentives in a bid for market share.
The result is artificially low prices that barely cover the providers' costs. As it stands, it can take providers as long as 55 months to recoup initial expenditures. To put it into perspective, Covad hasn't even been in business 55 months.
The local carriers' plan is to just tough it out. Unlike the exclusively DSL players, they can subsidize their data packages with revenue from their voice calling services. Remember, anyone using one of those new U.S. Robotics modems is still paying in to the coffers of a local carrier. And features such as Call Waiting and Caller ID all help squeeze a few extra dollars out of those old analog lines.
But even the players with the deepest pockets can't last forever in this market. In fact SBC, the nation's largest DSL provider and parent of Pacific Bell, has already announced a 25 percent price hike. The new residential rate will be $49.95 per month.
That's great news for Covad and NorthPoint, since it puts SBC's pricing more on par with what they've been offering all along. But even with higher prices, the only real hope for any player in this market is that DSL will prove to be a winner-takes-all game. The real money comes when a provider can offer bundled services — including voice, data and even video-on-demand — all over the same wire.
To achieve that, however, means keeping your customers — and that's proving a tough battle for any DSL provider. The need to reduce costs has prompted staff cuts, which in turn has led to poor service across the board. Lead times for installations and repairs are both at all-time highs.
That this results in poor customer satisfaction should be a no-brainer. According to the Strategis Group study, despite the majority of broadband customers choosing DSL initially, only 43 percent are "extremely satisfied" with the service once installed. "Churn," the rate at which customers jump ship to other providers or services, averages 15 percent.
At this rate, the only probable fate for the DSL industry will be widespread consolidation. Smaller providers will continue to fall, leaving the major players to feast on their remains — just as Time Warner Telecom demonstrated when it purchased most of the assets of GST Telecommunications.
Eventually, should this economic house of cards manage to remain standing for a few more years, we'll likely be left with no more than two or three major players in the DSL game, total. How good do you suppose the quality of DSL service will be then? How competitive will the pricing be? Or will the telecommunications industry fumble the ball on DSL completely, the way it did with ISDN in the early 1990s?
Either way, the glow of the initial marketing hype for DSL is beginning to fade, and what once appeared to be a promising technology now seems to have an uncertain future. So for a business looking for online connectivity, is it time to overlook DSL as a viable option?
Absolutely not, according to Bill Woodcock, telecommunications consultant and founder of the Berkeley-based network solutions company Zocalo Engineering. "Buy it," says Woodcock. "And its competition, and its competition's competition. Buy some of each."
The key, according to Woodcock, is redundancy. "Never get stuck with just one option. Order all the options." In a world of unreliable providers and poor service, the only way to ensure your Internet onramp may be to hedge your bets.
It's almost enough to make that new analog modem sound downright appealing, isn't it?
