The Slow, Steady Pace of DSL
Welcome to the world of digital subscriber
line (DSL), where everyone talks a good
game but they won't tell you their scores.
Like the golfer decked out in the latest duds
and armed with the best clubs but who
refuses to divulge his handicap, even the
most DSL-happy incumbent local exchange
carriers (ILECs) and competitive LECs
(CLECs) become curiously silent when asked how many lines they have
installed or customers they are serving.
"They won't even tell me that kind of information under nondisclosure,"
says Claudia Bacco, the Dallas-based senior DSL analyst for research
house TeleChoice Inc. "All I'm left to believe is that [ILECs and CLECs]
have a lot fewer customers and lines installed than they will admit."
To get a reasonable handle on the state of DSL deployments, it is
necessary to sift through all of the fanfare and announcements of trials,
says Craig Driscoll, research analyst for The Yankee Group Inc., a
Boston-based market research house. While getting actual customer
names remains an elusive task, he says, it is easy to see why users are
interested in the new service. DSL promises to provide users with
high-bandwidth connections--up to 8 megabits per second (mbps)--over
existing copper lines at about half the cost of traditional access services,
such as T1 and frame relay.
The Pelorus Group Inc., a Raritan, N.J.-based market research house,
projects the total number of DSL lines installed this year is 325,000.
While this represents a healthy increase over 1997's estimate of 95,000,
ILECs and CLECs truly will have to step up to make the 1999 projection
of 1.2 million lines, says Pelorus Group President Al Fross.
Despite the lack of data, market watchers had no problem reaching a
consensus that current DSL dominators include: U S WEST !nterprise
Networking Services on the ILEC side; GTE Corp. with Covad
Communications Co. for the independent telcos; and NorthPoint
Communications Inc. and Rhythms NetConnections Inc. leading the
CLEC charge. But even when talk turns to the companies that were voted
most likely to succeed, talk turns out to be more about lagging than
bragging.
Slow But Steady
For example, U S WEST, which first announced its MegaBit DSL
services in January, didn't begin its rollout until early summer. While it only
offered three plans--including asymmetric DSL (ADSL) and
rate-adaptive asymmetric DSL (RADSL) at speeds from 192 kilobits per
second (kbps) to 704kbps--market watchers agree the pricing is
attractive (between $40 and $125 per month).
Even though the ILEC now is embarking on the largest ADSL
deployment to date, targeting at least 50 percent of its customers, market
watchers remain skeptical.
GTE Network Services, the ILEC unit of GTE, plans to offer ADSL
service in 300 central offices (COs) in 16 states by the end of the year.
GTE also has teamed with America Online Inc. (AOL) for trials in
Birmingham, Ala.; Phoenix; Redmond, Wash.; the San Francisco Bay
area; and northern Virginia. GTE Internetworking is providing the
integration between the participating telephone companies, GTE's local
telephone operations and Bell Atlantic Corp.
Yet even the scale of these announcements falls short of the initial
expectations of many market watchers. The ILECs, true to their long
heritage, remain focused on precision over precociousness.
"I'm disappointed that there hasn't been the big wave [of rollouts] that was
expected," says William Rodey, a vice president for the ADSL Forum
trade association and Westell Technologies Inc., an Aurora, Ill.-based
maker of DSL equipment. "But the ILECs are conservative by nature.
Their controlled introductions don't quite make as much noise as we'd like
to hear, but perhaps that will change by year-end."
That's not likely, however. U S WEST isn't turbocharging its timetable,
according to a spokesperson. At least plans now call for service to reach
as many as 40 cities by the end of the year, while the original plan called
for only about a dozen cities in its 14-state territory.
DSL Deployment Map
An equally deliberate pace categorizes DSL efforts at Bell Atlantic, which
has endured more trials than Perry Mason. The last ILEC to announce
deployment, Bell Atlantic also was the first ILEC to test ADSL back in
1995 when it was being considered as a delivery mechanism for
video-on-demand (VOD) services. The ILEC now reports its DSL
service will be available on 2 million lines by the end of the year, with a
total of 7 million by the end of 1999.
While that sounds substantial, the ADSL Forum's Rodey says it
represents only 18 percent of the Bell's total access lines. Bell Atlantic
started DSL service in Philadelphia, Pittsburgh and Washington last
month. Parts of northern New Jersey are slated for this month with
Boston and New York next year, according to a company spokesman.
While categorizing this "measured approach" as a bit exasperating for
wishful customers, Charles Carr, a Denver-based industry analyst for
research house Dataquest Inc., says he gives Bell Atlantic credit for letting
its current integrated services digital network (ISDN) customers upgrade
to DSL at no charge. Pundits blame part of the slowness in all of the Bells'
DSL rollouts on their fear of losing the hard-fought base of ISDN
customers.
In its InfoSpeed offering, Bell Atlantic is offering customers three flavors
of ADSL, which are available through either BellAtlantic.net or various
other Internet service providers (ISPs). Unlimited access of up to
640kbps costs $69.95 or $39.95 through an ISP, 1.6mbps costs
$109.95 or $59.95 and 7.1mbps costs $189.95 or $109.95.
Image: Go the Distance
The Pacific Bell unit of SBC Communications Inc., meanwhile, last month
extended the initial rollout of its FasTrak DSL service to some 5 million
customers in 200 communities serviced by 87 COs throughout California,
according to Mark Gallegos, the San Ramon, Calif.-based product
manager, DSL marketing, for SBC.
While acknowledging that PacBell is not the first to deploy DSL service in
the state--a silent tip of the hat to CLECs Covad, NorthPoint and
Rhythms, which have made the Bay Area in particular a DSL
hotbed--James Callaway, public affairs president for PacBell, calls the
rollout "the real deal" as it brings DSL service to 65 percent of
Californians. PacBell offers three flat-rate packages, starting at $59 per
month plus Internet access charges for 384kbps downstream and
128kbps upstream.
While the ILEC does not differentiate between residential and business
customers, Gallegos says the other two offerings are 1.5mbps
downstream and 384kbps upstream for $339, and 384-by-384 for
$99--an option more likely to attract small to medium-sized businesses.
BellSouth Corp. also favors ADSL service, which it began offering last
month in New Orleans. Plans call for six more Southeastern cities this
year with a total of 30 over the next 14 months with 23 more in 1999. But
not all customers in the "Big Easy" will find it easy to get ADSL service.
This has been true in many other rollout areas, where customers with
phone lines connected to digital loop carriers (DLCs) between their
premises and the CO are out of luck and those who are further than
18,000 feet from the CO are out of range.
A Limited Range
Perhaps DLC should stand for "disappoints lotsa customers" when it
comes to DSL deployment because its presence forces ILECs and
CLECs alike to pick and choose where they can offer ADSL service. As
a result, most carriers must do as BellSouth has done: focus on areas
where there is an abundance of copper and a lack of DLCs, with
multiplex telephone lines for customer access.
Unfortunately, DLCs typically are found in newer neighborhoods, where
offering DSL service is the most desirable, says George Churchwell,
president of Tech 2000 Inc., a Herndon, Va.-based company that helped
train both U S WEST's and Bell Atlantic's DSL implementation and sales
teams. There are ways to dodge the DLC dilemma.
"Bell Atlantic was more selective in its approach than U S WEST," he
says. While U S WEST chose to qualify potential DSL customers over
the telephone when they called to ask about the service, Bell Atlantic is
strongly advertising the new service only in areas where DLCs are not a
problem.
"While ADSL is looked upon as the most glamorous DSL technology
because of its ability to offer up to 8mbps of bandwidth, its acceptance in
the market will be somewhat limited due to issues with DLCs and loop
lengths," says Sassan Babaie, vice president of marketing for Pulse
Communications Inc., a Herndon, Va.-based supplier of DSL equipment.
"I predict ADSL won't be able to follow through on its expected uptake,
so [service providers] will look for a pragmatic, not exotic, DSL solution."
Kumar Shah, vice president of marketing for AccessLan Communications
Inc., a DSL equipment startup, says the interference caused by these
conditions is a huge impediment to ADSL deployment. As an alternative,
he suggests symmetric DSL (SDSL), which today represents 60 percent
of all DSL loops and should continue to do so through at least the end of
next year.
ILECs also are cautious because they are finding not all their copper is
qualified for DSL transmission, says TeleChoice's Bacco. "DSL is very
complicated to deploy," she says. "Especially if you're not aware of all the
issues involved with bridged taps, load coils and binder groups. Load
coils block ADSL's high frequency while bridged taps reduce the signal
and create reflections on the line."
Because of the potential for cross-talk, especially with ADSL, service
providers must keep on top of which wires carry which signals. For
example, T1 service should not be put in the same binder group as
ADSL, says John Reister, director of marketing for Copper Mountain
Networks Inc., a California-based DSL equipment maker with offices in
San Diego and Palo Alto.
A New Animal
If ILECs need any incentive for straightening out these problems, they
need look no further than a handful of CLECs that are powered by a
double-barreled threat: pockets full of venture capital (VC) cash and all
their eggs in the DSL basket.
While Covad and NorthPoint have garnered a lot of interest for their
proposed nationwide DSL rollouts, darkhorse Dakota Services Ltd.
proclaims equally gaudy intentions--namely, 1 million lines sold by 2000.
"I didn't say which month in that year," says Theodore Lasser, CEO of
the Wakesha, Wis.-based CLEC, with a chuckle. But Dakota is on the
way, currently offering SDSL and RADSL in 27 COs throughout the
Ameritech Corp. region with 73 more COs under way. The goal is to
offer DSL services in 78 cities by June, he says.
Such aggressive plans have gotten the attention of the financial community,
which has backed many CLECs with millions of dollars. In fact, Lasser
says the odds of Dakota making its million mark are more a factor of
dollars gained than deployments made.
"We know we can do it," he says. "Our plan depends on funding from the
financial community."
Unlike some other CLECs, Dakota doesn't worry that its regional ILEC
will slow its rollout by blocking its access to the raw copper.
Ironically, the biggest challenge that HarvardNet Inc., a Portsmouth,
N.H.-based DSL service provider, faced in the early going was its
regional ILEC's lack of a DSL offering. "When we first offered service
some two years ago, we had several people come to us and say, 'We
talked to Bell Atlantic [then NYNEX Corp.] and they said DSL doesn't
exist,'" says CEO Bill Southworth. Truth be known, the ILEC didn't have
such an offering at the time.
Even without the ILEC's DSL efforts to validate its efforts, HarvardNet
has deployed in 35 COs and is now on course to add 15 new COs each
month toward reaching 200 by the end of 1999, Southworth says.
HarvardNet's strategy is to blanket the Northeast, especially its home
state and Maine, two markets that he says traditionally have been
underserved by telco services.
Calling its $60-per-month SDSL offering "a cable [modem] killer,"
Southworth says HarvardNet is doing extremely well serving "secondary"
markets in New Hampshire such as Manchester, Nashua and Portsmouth
as well as Concord, Mass. While he questions just how much of a real
presence noisemaker Covad has in the Boston area, Southworth says
HarvardNet has several customers that have been receiving its DSL
service for two years.
"CLECs are going to prove the DSL market for the ILECs," says Ron
Young, vice president of marketing for Diamond Lane Communications
Corp., a Petaluma, Calif.-based maker of DSL gear. "The first to market
is usually the winner, but if DSL performs up to expectations, there is
plenty for everybody. Let's not forget: The ILECs are the Baby Huey of
the playground. When they finally want to play, everybody better make
room."
Gaining Mass
While the lack of any region-wide DSL deployments by the ILECs have
given fast-starting CLECs an early edge in some markets, Westell's
Rodey says it's just a matter of time before the race heats up.
"The ILECs know how to create mass-market demand," he says. "Look
at what they've done with Yellow Pages, cellular, fax, voice mail. DSL is
just as compelling. Real amounts of money will not be made in DSL until
there is a battle for market share."
According to the Pelorus Group, that battle won't begin until at least
2000, when 2.45 million DSL lines are expected to be in place.
Competition will further intensify the following year with the addition of
more than another 1 million lines.
"For CLECs, it's extremely easy to go from a trial to a small deployment
in a specified area," says Pulsecom's Babaie. "But going from small to
large deployment is another matter. Very, very few CLECs have the
resources to scale that way." While the ILECs, all large and
facilities-based, are a given, he estimates that perhaps as few as 10
CLECs have the big-game potential.
Survival requires more than providing DSL service. Understanding that
"DSL by itself is just another access technology," Jim Greenberg, chief
network officer for Rhythms, says his company's focus is on providing
value-added applications with DSL service. The Englewood,
Colo.-based company offers DSL service in the California markets of
Los Angeles/Orange County, Oakland, San Diego, the San Francisco
Bay area and San Jose on its way to 10 markets by the end of the year.
In addition to high-speed access, customers receive embedded
applications, remote backup and central archiving. Rhythms expects to
add Boston, Chicago, New York and Washington before the end of first
quarter 1999 and plans to total 30 markets by the end of 2000,
Greenberg says.
These kinds of value-added offerings are the future of DSL, says
Dataquest's Carr. "Otherwise, DSL becomes just another commodity
service. This is the trap DSL must avoid to reach its full potential."
It's not enough to bang the DSL drum loudly, which arguably no CLEC
has done better than Covad. This summer the Santa Clara, Calif.-based
CLEC announced its service is available to 1.1 million homes and
businesses in the San Francisco area. The gap between having service
ready for that many customers and how many have signed up remains
confidential, according to vice president of marketing Louis G. Pelosi. He
did say Covad plans to raise that number to 5 million by the end of the
year with service starting in Boston, Los Angeles and New York.
Planning for Seattle and Washington is under way as part of being in 22
markets by end of 1999, he adds.
Even though Covad professes to be at the leading edge of DSL
deployers, Pelosi admits Covad is keeping an eye on all its competitors.
"Our founders are from Intel [Corp.], so we know that only the paranoid
survive," he says, alluding to the title of Intel executive Andy Grove's
autobiography.
Image: U.S. DSL Installations
Aside from achieving blanket coverage in major metropolitan markets
nationwide, Covad sees its edge in offering the broadest selection of DSL
speeds and flavors, Pelosi says. Covad's TeleSpeed offering delivers
ADSL, ISDN DSL (IDSL) and SDSL at speeds from 144kbps to
1.5mbps.
"If we only offered ADSL, we'd have to turn away perhaps four out of 10
customers," Pelosi says, due to issues of DLCs, line conditions or extreme
customer distances from the CO. Covad also favors giving customers
choices of DSL customer premises equipment (CPE). Unlike many
service providers, which only work with one equipment vendor, Covad
offers customers gear from Ascend Communications Inc., Diamond Lane
and Efficient Networks Corp.
NorthPoint has chosen to differentiate itself from regional rival Covad in
several ways. First, the San Francisco-based CLEC has chosen to
wholesale its DSL services, which range from 160kbps to 1.04mbps, to
ISPs for resale to their customers. While Covad is targeting small to
medium-sized businesses and telecommuters, NorthPoint is the first DSL
service provider to offer nation-wide service level agreements (SLAs) to
network service providers that use NorthPoint's network. They are
guaranteed 99.99 percent packet delivery with a maximum roundtrip
latency of 10 milliseconds.
The ability to offer such guarantees is necessary for DSL to be accepted
as a reliable business service, says Ann Zeichner, NorthPoint vice
president of sales and marketing. If NorthPoint fails to uphold
packet-delivery guarantees over a 24-hour period, customers will receive
a one-day credit for service. Customers can receive monthly status
reports.
Using DSL access multiplexers (DSLAMs) from Copper Mountain,
NorthPoint has deployed service in Boston, New York, Los Angeles and
the greater San Francisco area. Plans call for a total of 10 cities by the
end of the year with 10 additional in 1999, Zeichner says. Decisions on
future deployments will focus on populated business areas in top tier or
"football" cities, she adds.
As for any rivalry, friendly or otherwise, with neighboring DSL provider
Covad, Zeichner says there is competition between the two companies,
as they were formed within nine months of each other. While Covad has
enjoyed a CLEC's biggest advantage--first to market--she says
14-month-old NorthPoint's edge gained from its executives' prior DSL
experiences with WorldCom Inc. and UUNet Technologies Inc. will help
narrow the gap as deployments continue and mature.
"Let's face it, until that time we're all just pretty faces out there," she says.
"So whose story sounds better or more promising? Over time, how we
execute is how we will determine who succeeds." |
