REMEMBER WHEN just getting an e-mail was exciting? When logging onto a Web site and watching the image appear pixel by pixel was thrill enough

"/>

How broad is your bandwidth?

Serving up high-speed hope to home Web surfers.

REMEMBER WHEN just getting an e-mail was exciting? When logging onto a Web site and watching the image appear pixel by pixel was thrill enough to keep you at your monitor for hours? Been a while since you felt that way, hasn't it? But if a curious mix of corporate competitors great and small have their way with you, the Internet may soon be fun again.

Alternatives to the slug-speed dial-up Internet connection have actually been around for the better part of a decade, but for all but the most passionate surfers, they've been too technically demanding, too expensive to install, and too damn much trouble to get up and running. That's changing—not quickly, but fast enough that by the end of this year, a significant number of Websters will have shucked their old analog modems for a high-speed link to the Internet via fiber-optic cable, digital subscriber line (DSL), even satellite or wireless technology, at a comparable or better price per byte to what they're paying now, and (sometimes) greater reliability as well.

One reason people are getting more willing to take the high-speed plunge is that big names in the telecom field are starting to sell such service to the masses. Recently AT&T, which last year bought TCI's nation-spanning cable TV system, started offering Pugetopolitans low, low monthly rates and cut-rate installation and hardware to get new subscribers for its branded cable-modem system ATT@home.

Local phone companies can provide a similar service called "digital subscriber line," though up to now they haven't exactly gone out of their way to tell anyone about it. But if you call your US West (or GTE or whoever) customer service office, the operator may cautiously acknowledge that if you live in certain service areas, your telco might be willing to set you up with high-speed Internet access—sometime, for a rather stiff price, if asked nicely.

The only reason they've gotten that accommodating is that, thanks to an unintended consequence of the federal Telecommunications Act of 1996, a number of feisty new players are invading the telcos' turf to offer DSL service—cheaper, better, and with a human face.

WHICH TECHNOLOGY serves the customer better? At first glance, there seems to be no contest. Cable systems operate at light speed along the same fiber-optic line that already carries a bazillion bytes of digital TV through your neighborhood; DSL depends on a pair of twisted copper wires installed and maintained by—shudder—the phone company, the same kind of line that limits your current modem to a beggarly 56 kilobytes per second throughput. Cable can pump data at speeds verging on 1.5 megabytes per second; most home DSL systems deliver at only half that rate, and, if the distance from the telco central office to your house is more than a few miles, don't deliver at all.

But this time out, old technology has some hidden advantages over new. Cable can shoot you data at a meg-and-a-half per second—if you're the only person receiving on the channel. But as soon as anyone else on the same line starts up- or downloading, your share of the pipeline shrinks in proportion. It's like taking a shower in an old house: If enough other people turn on a tap, the flow in your stall can drop to a dribble. The only way your cable company can cut the congestion is to cut the number of users per "node": the box somewhere in your neighborhood where copper co-ax lines from up to a thousand cable households meet the fiber-optic spine of the system. Installing a new node is laborious and expensive, so the cable service provider has a financial incentive to load as many new customers as possible on each one, while keeping existing customers from getting too irritated by the constantly diminishing levels of service. When a new node is installed, relief is temporary, because the new "bandwidth" immediately starts filling up in its turn.

DSL providers (and users) don't face this dilemma. Once you have a functioning DSL line to your location, it is yours: Whatever its rated capacity—384 kilobytes per second and up—you won't sit down to surf one morning to discover your home-page takes a minute and a half to load because everybody else on the block is doing the same thing.

The biggest impediment to a mass market for DSL has been the phone company itself. DSL technology has been around since the 1980s, when a Bell Labs scientist figured out how to send megabytes of data over a humble copper phone line. But by the time the bugs had been worked out of a consumer version of DSL, the phone companies were making a bundle by providing high-speed T-1 data lines (same principle, different technology) to business, and saw no reason to promote DSL, which could only cut into their existing high-speed business.

Enter Congress, with the Telecommunications Act of 1996, which required the Baby Bells to allow competitors to sell local phone service. There was no way for a newcomer to compete profitably with the big kids in providing standard phone service. But DSL was something else: US West and its ilk found themselves bound by law to lease phone lines to upstart companies like Northpoint and Covad, so that Covad and Northpoint could in turn lease said lines to customers tired of waiting for the biggies to sell them high-speed Internet connections.

The new technology also meant a new lease on life for local Internet service providers, who have found themselves losing customers to national-brand outfits like AOL, AT&T Worldnet, and the like. One such ISP is Seattle's Speakeasy, which started out in 1995 as a humble Belltown coffeehouse-with-Internet-access and is in the process of transforming itself into a national provider of high-end, high-speed Internet service for demanding users.

HOW CAN A GRASSROOTS outfit like Speakeasy, the brainchild of Mike and Gretchen Apgar and Mike's brother Tyler, compete with multinational telecoms for the high-speed Internet consumer dollar? Because beneath the brand names and glossy animated logos, the infrastructure of electronic communications today is increasingly anonymous, distributed, and commodified. All you need to break into the business is money: to lease hardware and software and bandwidth, to market your service to potential customers, to hire personnel to sign them up and keep them happy.

It's that last part that gives upstart outfits like Speakeasy a level playing field to take on the communications megafauna. Mike Apgar's first computer-related job was in customer service, and he cut his techie teeth devising a program to expedite trouble calls for himself and his customers. Speakeasy promises its customers not just DSL but access to the kind of content that demands high-speed service: real-time gaming sites, streaming hi-fi audio, neo-video, and experimental film. But above all it promises competent, fast, reliable customer service.

In a commodified communications world where incessant corporate takeovers have reduced company names to little more than Post-It labels, customer service is the one opportunity most firms have to establish brand loyalty. Companies like AT&T and TCI try to do it with slogans ("It's better here," pipes US West plaintively), but old monopoly attitudes die hard when top management is 10 tiers away from the individual customer who wants to know why she hasn't been able to access her e-mail for three days. US West marketing guy Steve Bartholet says that, far from neglecting DSL in the Puget Sound area, the company is selling its service hard here. Maybe so; but when you have to call Denver three or four times just to get a brochure the local sales rep says doesn't exist, you have to wonder how hard the sell really is.

So far, according to Web sites that rate DSL and other Internet service providers, Speakeasy is doing something right. The company is signing up customers in 50 cities nationwide with little promotion beyond word of mouth. Of course, that's still only 1,000 or 2,000 new customers a month, while US West already claims over 130,000 in its 14-state service area and ATT@home has topped 40,000 customers in just the Northwest. But Speakeasy's not alone in taking on the telcos over their own wires: Two among many other smallish but highly regarded local providers are RMI.net and Fastpoint. Even with such competition, Apgar's not worried about keeping Speakeasy growing: Operating up till now on hard work and an SBA loan, his family-owned company plans to go public sometime this spring. "We're rapidly gaining significant market share because we're just listening to our customers and finding out what they want. Coming from the Internet space for the last four years, we understand that customer service and attention to network quality is what people not only want but demand; they are the minimum we've got to do."

There's no question that once fiber-optic cable becomes the default delivery system for data to individual households, companies like ATT@home are going to be formidable competitors for your business. But in the right-now real world, AT&T is still gagging from trying to absorb TCI at one gulp and continues to stumble badly in the reliability and customer-support departments. (Things got so bad in mid-February that the company apologized by e-mail and gave thousands of customers a free month of service in recompense.)

DSL has its own frustrations, but at least its infrastructure is bug-free and reliable, and competition between providers helps keep customer service attentive and is already beginning to bring prices down. Looming on the horizon are even faster, more flexible, more convenient technologies to connect you to the Web using geosynchronous satellites delivering data to a backyard dish or the distributed microwave arrays that now service cell phones, auto-computers, and Palm Pilots. But for now DSL seems to be the way to go.

Still hungry?

Read this very informative survey of broadband options from October's Scientific American.

 
comments powered by Disqus