Highway To Hell
Scott Woolley, Forbes Magazine, 02.19.01
EVEN BY THE LOW STANDARDS for service and credibility set by many firms riding
the Internet bubble, the Digital Subscriber Line industry is a mess.
DSL was supposed to thrust antiquated phone lines into the digital broadband era.
Consumers would be able to download Web pages at speeds 26 times as fast as
regular "dial-up" service, and phone companies would be able to compete with
cable modems.
Instead, bankruptcy courts are flooded with filings from NorthPoint, Zyan and
dozens of other firms providing DSL. Investors have been scorched. The three
biggest players Covad (nasdaq: COVD - news - people), NorthPoint (nasdaq:
NPNTQ - news - people) and Rhythms NetConnections (nasdaq: RTHM - news -
people) are all down more than 95% from their highs, wiping out $19 billion in
market value. Combined, the big three lost about $1.5 billion last year. Rhythms
expects to lose $395 million more this year and says it won't even show positive
cash flow until late 2003.
A cash crunch is rippling through the business, sparking feuds between big DSL
carriers and their tiny Internet-access partners. Covad gets stiffed on a third of its
invoices to the Internet service providers that resell its service. Desperate, Covad
and other DSL carriers are signing the ISPs' customers directly, despite contract
terms that bar any such contact. Caught in the crossfire are consumers; one class
action seeks $50 million from Verizon (nyse: VZ - news - people) for painfully
protracted installations.
The turmoil threatens to hobble a popular on-ramp to the high-speed Internet. People
like DSL once they get it. DSL grew 185% last year to more than 2.9 million phone
lines in North America, says research firm Cahners In-Stat. Roughly 65% are
residential customers; the balance, businesses.
The question is whether the torrid growth can continue. DSL had looked like a lock
to surpass its archrival, the cable modem, by the end of the year. (Cable modem
use grew 144% to 3.4 million lines last year, according to Gartner Dataquest.) DSL
is a long shot to seize the lead now, as the industry cuts spending to the bone.
Rhythms NetConnections says it is "refocusing" on 40 cities, down from 60, and
just fired 450 people, 23% of its work force. Covad canned 13% of its workers, is
halting its network expansion and plans to chop capital spending to $250 million,
roughly half of last year's level.
Only part of the industry's descent into chaos owes to the balkiness of DSL
technology. The bigger contributors are the frenzied rush to sign up customers at
any cost and archaic regulations that forced DSL players to adopt a wrongheaded
structure from the get-go.
First seen in the mid-1990s, DSL had two obvious sellers:the Baby Bells and ISPs.
The Bells could use DSL to expand into online access; the ISPs could offer DSL and
charge $40 or $50 a month instead of the $20 a month they get for poky dial-up
service. But the Bells were hindered by 16-year-old federal rules that bar them
from carrying voice or data across long-distance boundaries. Getting around the
restriction required an inefficient setup, with the Bells handing off data traffic to
other carriers.
ISPs, meanwhile, needed phone lines to offer DSL, but were loath to rent them from
their natural rivals, the Bells. Into this gap sprang a new, unregulated creature:the
nationwide DSL carrier, à la Covad, NorthPoint and Rhythms. They installed their
gear in Bell sites, rented Bell lines and partnered with ISPs.
And so was born an ungainly, three-tiered industry. The conceit was that the three
types of companies would work in harmony to let DSL proliferate. The reality was
that when glitches occurred, customers ping-ponged among the three, each of
which pinned the blame on the other two.
Worse yet, "The whole structure made zero sense from an economic standpoint,"
says Barry Diamond, chief executive of Internet Express, an ISP. Of a $50 monthly
fee paid to the ISP, the ISP pocketed $10 and passed on $40 to the DSL carrier; the
carrier paid $22 or so to the local Bell to rent the line. In each case, the amount left
over didn't cover the costs involved.
Eager to attract subscribers quickly, DSL carriers signed resale deals with
hundreds of regional ISPs. To get a new line installed, DSL carriers had to send
trucks to homes and offices to hook up the inside wiring, at around $225 per
account. The ISPs covered that cost instead of charging customers for new
hookups, and spent another $100 or so on other miscellaneous setup costs.
Recouping the $325 would take almost three years, given an ISP's $10-a-month
gross profit; yet the ongoing expenses of billing, hosting e-mail accounts and
technical support ate up the $10-a-month cushion and more.
Worse for the DSL carriers, they often had to dispatch truck visits two or three
times, losing money on their $225 payment from the ISPs. Their $18-per-month
gross profit on each account got them nowhere near positive cash flow, but the
stock market didn't care. Wall Street valued these firms at as much as 250times
sales. The only important number was line growth.
"We were highly incented by Wall Street to spend money like drunken sailors," says
Elizabeth Fetter, NorthPoint's chief executive. So everyone did.
As orders poured in, DSL's flaws loomed larger. Phone lines weren't designed with
DSL in mind. Gizmos called load coils, installed underground long ago to improve
voice quality, often filter out DSL signals, but the carriers installing the lines didn't
always know they were there. Since DSL works over distances of less than a few
miles, carriers used software to estimate-as the crow flies-the distance from a
customer to a phone-switching office. But the software often got it wrong because
lines take winding routes.
Now that the DSL push is sputtering, the new focus is on eking out a profit, but
there is no proof carriers ever will. While more bankruptcies are inevitable, Covad
Chairman Charles McMinn sees "plenty of reasons for optimism." Like many in the
industry, he pins his hopes on "line sharing," which lets DSL carriers piggyback on
an existing copper line rather than lease a brand-new one. The incremental price
for sharing is a mere $6, typically, rather than a $22 fee.
That would nearly double the DSL carriers' gross profit on residential lines. And
shared lines often can be hooked up by customers, saving $200 or more on
sending out a guy in a truck. Covad says that 17,000 of the 69,000 lines it installed
in the fourth quarter were shared, up from just 400 of 67,000 in the third quarter,
and it now won't offer consumer service without line-sharing. Still an open
question: whether those savings will be kept as profit or get passed along to
consumers in the fight for subscribers.
The Bells, despite a shoddy service record, could ultimately dominate DSL,
especially if they can shed old long-distance bans and sell service without the
hassle of having to partner with long-haul carriers. But the Federal Communications
Commission has been slow to free them. Even if the FCC acts quickly, it isn't clear
that DSL, in such turmoil, can keep pace with cable. That's too bad; seems like
everyone involved should have known better. |
