DSL start-ups begin to fold before turning a profit, while Bells sit pretty
By Peter J. Howe, Globe Staff, 12/17/2000
If there's one thing besides love people can't seem to get enough of these days, it's high-speed Internet access.
And because they can use many of the same copper phone lines that already blanket America, digital subscriber line systems would seem to be uniquely well suited to meeting the demand for zippy Net service.
So how come almost no one, except the gigantic Baby Bells, seems to be able to stay in business selling DSL?
In the last 10 days, two local providers, HarvardNet of Medford and Digital Broadband Communications of Waltham, announced they are pulling the plug on their red ink-hemorrhaging DSL systems, laying off a total of 730 people.
Nationally, the scene also is growing bleak. Shares in DSL providers such as Covad, Rhythms NetConnections, NorthPoint Communications, and Network Access Solutions have plunged over 95 percent from their late-winter highs, raising questions about who will be next on the DSL death watch and dragging down shares of DSL gear makers such as Efficient Networks and Copper Mountain.
NorthPoint's revenue and stock collapse led Verizon Communications last month to back out of buying a 55 percent stake in the company. That triggered a $1 billion lawsuit from NorthPoint, which laid off a fifth of its work force this month.
If these companies have one thing in common, it is having been whipsawed by a dramatic turn in sentiment on Wall Street.
A year ago, investors seemingly couldn't throw enough money at DSL providers, seeing them as nimble newcomers that would run circles around the Baby Bells meeting businesses' and someday residential customers' insatiable demand for high-speed, always-on Net access.
Now the prospect that they will need hundreds of millions of dollars more before they start making money has led a newly profit-conscious Wall Street to shut off the spigot.
Failure to attract more financing pushed HarvardNet, for example, to abandon its DSL unit while it still has hope of salvaging its Web page hosting business as a viable operation.
After a slow start and despite continued complaints of bureaucracy-plagued ordering and service systems, Baby Bells have roared back to control about three-quarters of the 2.3 million DSL lines forecast by IDC of Framingham to be in service nationally by year's end.
Given the huge costs and limited profit margins of delivering DSL, it has become clear in the last several months that only giants such as Verizon, SBC, Qwest and BellSouth are deemed to have the deep pockets needed to sustain the business.
SBC, for example, is pouring $6 billion into its ''Project Pronto'' to roll out DSL nationally, already attracting over 500,000 lines.
''I felt like the opportunity and demand for broadband was still strong,'' said Valeri A. Marks, president of Digital Broadband, whose customers include a state program serving school and municipal buildings in 85 Massachusetts cities and towns now scrambling to find a replacement provider.
''We were still having great success on the sales front, but then things turned on a dime on Wall Street, and funding dried up,'' Marks said.
DBC had attracted nearly $160 million in funding, including loans from Net equipment provider Cisco Systems, before it concluded that its DSL business - which had attracted fewer than 700 lines in service - had no sustainable future.
''It's a very capital-intensive business, and the competitors are very big and have very deep pockets,'' said HarvardNet president Mark Washburn. ''The markets have gone from a position of, `What will you do for me next year?' to `What will you do for me this quarter?'''
Besides some investors who may lose their shirts on DSL, the looming shutdown of service is creating a nightmare for Net access-dependent companies that fear they could be shut off for weeks with the holiday season slowdown.
''We have a potential one-month blackout where we have to go back to using a modem,'' said Manny Cabanas, whose four-person Linx Test Systems business in Brookline faces losing its HarvardNet DSL service on Jan. 15. The company helps design networking systems for computer chip manufacturers.
After a frantic search for alternatives, Cabanas said the company has signed up to get new service installed by XO Communications, the former Concentric, but it will take six to eight weeks.
''Effectively, we are going to be crippled, or Internet-challenged,'' during a busy time of the work year, Cabanas said.
In assessing how things went awry for non-Baby Bell DSL providers, industry executives and analysts cite not just Wall Street's refusal to keep pumping in money to cover operating losses and network buildout costs, but other factors including the following:
Overly aggressive expansion and marketing plans by some companies that, in the words of Christopher J. Oliver of New Hampshire-based Vitts Networks (a still-healthy DSL and networking services provider), tried for ''success by growth, instead of growing by success. Some of these guys overbuilt and got way out ahead of their funding.''
The dangerous dependence of many DSL providers on the services of their arch-foes, the Bells, in providing phone lines for them to use. The alternative would have been moving sooner as Covad has with SBC - and NorthPoint tried with Verizon - to co-opt their goliath competitors by becoming resale agents for them.
Excessively optimistic projections about how many phone lines actually would be able to support DSL. The service generally works only within a three-mile range of a phone company ''central office'' switching station and only then if the line to the business or home is free of blockages such as ''load coils'' or stretches of fiber-optic cable, which thwart DSL.
Years of delays by the Federal Communications Commission in approving so-called line sharing rules that leveled the competitive playing field between the Bells and independent DSL providers.
For years, the Bells had the built-in advantage that they could use a portion of their existing copper phone lines into homes and businesses to sell DSL, which adds electronic gear such as splitters and transmitters to exploit unused capacity on voice lines for 640-kilobit-per-second and faster Net service.
Companies such as HarvardNet, DBC and Covad, in contrast, had to pay for a new, premium-priced phone line when they wanted to offer a business or homeowner DSL.
It was not until this June that the FCC finally approved a plan to let non-Bells use untapped bandwidth on existing voice lines from Bells to add DSL.
In Massachusetts, state regulators did not complete issuing a pricing and deployment plan until late September, a four-month lag state telecommunications commissioner Paul Vasington defends as ''faster than we've ever moved'' on implementing a federal mandate locally.
Judy Reed Smith, president of Atlantic ACM, a Boston telecommunications consulting firm, said she thinks many DSL providers were doomed from the outset from their dependence on a shotgun marriage with deep-pocketed Bells who sought the same customers.
''It is very hard to make a business model that relies on your competitor,'' Reed Smith said. She said FCC delays in approving line sharing ''absolutely'' aggravated the situation by giving Baby Bells a crucial head start on grabbing business.
Verizon appeared before Massachusetts regulators this fall arguing that its success promoting local competition had earned it the right to enter the state's $2 billion long-distance calling market.
But Covad, DBC, and other DSL providers were quick to dispute that, saying Verizon regularly gave its own DSL operations more advantageous treatment than rivals.
Verizon contends apparent differences in installation times can be explained by business practices such as competitors not scheduling installations on Saturday and less rigorous work by them to ensure phone lines can handle DSL.
''I don't think there was any intentional sabotage'' of rivals by Baby Bell companies, Reed Smith said. ''It's just from their perspective, if you are not making money from it, that is not where you are going to put your technological people working on overtime.''
HarvardNet's Washburn said, however, ''For the most part, I think the service Verizon provided was adequate to get the job done,'' but competing on price alone with Verizon was nearly impossible.
Digital Broadband's Marks said getting cooperation from Verizon ''was challenging,'' but she refuses to blame her company's DSL demise on the incumbent phone giant.
One bright spot in the demise of HarvardNet and Digital Broadband is the potential for new customers for DSL survivors as well as alternative providers such as WinStar.
That company uses rooftop fixed-wireless devices to deliver voice and broadband data access to 3,500 buildings across the country, including business clusters in Boston, Cambridge, Burlington, and Woburn.
Locally, WinStar says it already has picked up several DSL refugees, including The Beale Cos. and Boston advertising agency Team H. WinStar expects to be able to add businesses leaving Digital Broadband and HarvardNet to its network within 20 to 30 days of signing them up.
While fixed-wireless rival Teligent has been widely reported to face a cash crunch, Winstar Boston general manager Gary Johnson said the company has cash on hand to last through 2002, and expects to reach positive cash flows next summer.
''I think there's still a lot of potential in the market,'' said DSL analyst Amy Harris of International Data Corp. in Framingham. ''I think we are just seeing a little bit of a shakeout.''
For all the complaints about problems working through Baby Bell competitors to reach customers, Harris blames the spate of DSL flameouts squarely on the investment community.
''Wall Street kind of precipitated this,'' Harris said. ''If they had had a little more patience and waited, they might have turned profitable. These guys were the darlings of Wall Street even as late as September and October. Then it came to a crashing halt. I do think a lot of it was out of these companies' control.'' |
