India by gabyion


									India. AOL, China, VDSL2 .


Long time no see!

Have been shuttling to APAC/China almost every third week.

IVD (Integrated Voice and Data) is getting very popular among non-North America carriers as well. Main

1) New telephone lines: Despite all the wireless growth millions of POTS lines are installed every year! Why
just install telephone lines and come again to overlay DSL later?

2) DSL has reached the price sweet spot that even the coast of cable running from splitter rack to the voice
rack and the connectors required to hook them up seems to cost more.

3) "Lock and leave". The labor cost is high (even in China!)

Infineon of course has major designs in the bag and is leading the technology beyond the traditional
voice+data line card.

Probably we should chat sometime!
Imran Ashraf Hajimusa

For 706: Should the speed reported be a maximum rate, counting all the bits theoretically available
from the DSLAM to the customer. Should overhead be discarded, requesting a count of the IP
traffic carried, since that's what's usable. Or should the carrier report the actual speed delivered
over their network, measured from the consumer to the edge of the network

Which headline?
Reuters Summit-AOL sees business buoyed by broadband (Forbes)
AOL Stops Selling Whole Broadband Plan (LA Times/WSJ/AP)
Googling AOL Friday Night, I find dramatically different headlines back to back. It's the same
story DSL Prime wrote a few hours before as AOL Dying. The facts are the same: AOL has
stopped selling broadband and sees its future as a website/portal. CEO Jonathan Miller spun the
news as positive, looking to the increased margins on software and services. The Reuters story
reported his point of view. Anick Jesdanun at the AP saw it as AOL exiting access, its primary
business, as broadband replaces dialup. ―The reversal in strategy stands as another black mark
against the purported wisdom of the $160 billion merger between America Online and Time
Warner at the height of the Internet boom, a deal the companies had described as a perfect
marriage of new and old media with the means to deliver it. ― He quoted me calling this an
admission of defeat "Selling access, you had to buy something from them," Mr. Burstein said.
"That was a real service. Now they are essentially another Web site."
   Miller recognizes "the world is going broadband-centric," and I believe would agree AOL's
future depends on a profitable broadband strategy. I don't think he has one, at least for a company
of AOL's size and financial liabilities. So in DSL Prime, I was more blunt. I think it likely any
surviving AOL will be a ghost of the giant dominating the internet a few years ago. I pointed to
Mike Powell's inaction to save independent ISPs, an issue well understood by his staff and

AOL survival strategies
AOL needs to take bold risks or sink to a minor player at best, with much dismal news on the way
down. It just lost 2M subscribers, and the number would have been much greater without hugely
expensive promotions and massive price cuts. The remaining dial-up lines are now likely to move
even faster, as telco/cable competition is currently keeping DSL prices down. SBC's $32 DSL
price is aimed directly at them, and some DSL and cable offerings will be $19.95 to $25 this year.
SBC CFO Randall Stevenson says DSL makes money at $27 (with volume) because costs are
down, and chip, network equipment, and bandwidth costs will keep falling. Prices will follow if
competition continues (not a sure thing.) As part of a bundle, the natural cost of low end cable
modem offerings is $14.95 to $19.95, with one MSO already considering the move. AOL has to
pray Ivan Seidenberg, Ed Whitacre, Brian Roberts and a few friends take their jets to a high
security meeting and cut a deal to keep prices up (unlikely) or that they find more apparently legal
means to get around the anti-trust laws (signalling, as Roberts and Ackerman seem to be doing in
public statements.)

   What AOL could do:

Unlikely to work: Offering email with anti-virus and spam filters, parental controls, cheaply
obtained content like movie trailers, and similar in a $14.95 offering. Almost guaranteed to fail,
because SBC and Verizon are offering a similar bundle included, and everyone else will soon
pretty much match it. SBC is paying Yahoo $2.10 per month for the package, a high figure that
should come down, so it's easy to include.

    Turn on VOIP phone services, with free calls between AOL members. Yahoo BB is proving
that works, and Vonage offers the same terms even on their unpublicized $10/month service.
AOL's huge membership base would yield subscribers at low acquisition costs, and using
Windows XP and a $19.95 headset would be a great offer. A consultant friend recommended this
to them over a year ago, but the idea was rejected because "we need the phone companies for DSL,
and VOIP would hurt our Time Warner cable operations." AOL could roll nationally in six
months, perhaps joining the Time Warner/MCI/Sprint service. A $14.95 AOL for Broadband is
attractive to very few; a $14.95-$19.95 offering with 500 minutes of calls across the country would
be more interesting.

   Quickly cut deals for lots of really great video content and games, and put it out cheap. Create a
really compelling offering. But we're not talking "Sessions at AOL" or major sports (baseball
wants too much for the rights for AOL to have any margins.) Movies won't cut it either, because
Hollywood can sell them directly through Movielink, DVDs, Netflix, and more, so they won't give
AOL any interesting profit margins. AOL has plenty of cash and potentially so many users they
might find some deals, but I haven't seen them in the marketplace.
    Change their mind about abandoning access and put out an infrastructure, perhaps with a
partner like AT&T. David Dorman has a strategy before him of putting cheap DLCs in thousands
of CO's if he loses UNE-P, thinking the states will at least keep the price of the physical access line
reasonable. Any deployment like that requires massive scale, so working together with an AOL
before they lose their 20M base could make sense.

     Instead, AOL is claiming they can make it with more of same in software and services, perhaps
a little bit better. But the total market for AOL for Broadband, Yahoo, MSN, Real, the in-house
offerings of Comcast and the telcos, and all the other services over the internet is highly
competitive and a fraction of the size of the access market. I don't see how AOL becomes more
than a niche, and I don't think AOL expenses are sustainable as a niche player.

AOL's most likely future will be to continue losing money as dial-up disappears faster than they
can shed overhead. They have already taken $40B or so of writeoffs, mostly goodwill, and may
have concealed some future costs in there. But more likely is quarters like this one with "operating
profits up after special items." Special items in this case are predictable, and operating profits are
not the same as real ones, especially for a company with high development and acquisition costs
that may be capitalized.

Jim Coates, Chicago Tribune "It's only a matter of time before dial-up Internet access goes the way
of rabbit ears for accessing television programs." That's actually pretty discouraging, considering
how many people around the world don't take cable. But he's right about the trend: one-third of
U.S. net users, two-thirds of Taiwanese have already switched to broadband. Coates also writes
"the vast bulk of folks wanting broadband Internet service have superslim pickings--the
domineering Comcast Corp. cable TV service or the SBC Communications Inc. phone giant." It's
time for folks like Qwest to stop pretending that two companies constitute effective competition,
and reporters and legislatures to stop pretending to believe them.

SBC: Cable's behind in our territory
Competition watch
Cablevision: $19.95 in 2004 bundles
Winning customers, with $9.95 in sight
Cablevision’s $35 unlimited local and long distance service is much cheaper than the $60-65
Verizon and MCI are charging. (Don't you hate those ads that say the price is $49.95 when you
know the bill will be $65. Thank you, Cablevision, for telling the truth.) Cablevision has 22%+
penetration for cable modems, and now is looking for how to grow beyond the "early adapters."
With cable modems costing less than a provider $70 with full voice capability, they plan soon to
include a voice port on every new modem, and provide a free modem to any video subscriber who
buys voice service. They figure to recover the extra hardware cost by selling more services to each

$19.95 is a magic point that will drive take rates, according to their market research, and they
intend to feature it in several bundles. One will offer voice at $19.95, probably a bundle of
minutes. Another will offer data, at speeds slower than the 5-6 Mbps they are tuning their network
for. $9.95 is may be the ultimate weapon to get people to take a third service. That price is
ruinously low for a single service, but not crazy for an add-on to sell the bundle.

Bundling & Churn: Partly an "Urban Myth?"
Nearly everyone believes that bundling is the most effective strategy to hold customers, changing
everything because "they want one bill." The telcos are pointing at repackaged satellite deals as a
great savior. Clearly there's some effect, but my first guess is it's much less than claimed by xxx
experts in Matt Richtel's New York Times story.

   Randall Stephenson's data suggest otherwise. A first pass at the figures seems to confirm that,
with the SBC CFO reporting churn goes down 11% if customers add LD and even more if they add
other services. Actually, the 11% difference is very small, because the group of customers
choosing to buy SBC LD is presumably much less likely to churn in the first place. For one thing,
they probably like SBC service, or wouldn't have gone to SBC for LD. In addition, since few
customers expecting to move soon switch LD carriers, they were less likely to churn in the first

   High DSL churn rates suggest many customers still unhappy, and hence a risk to take all their
services to the cable bundle instead. SBC churns possibly 3% a month, with others higher. Again,
happy customers are far more likely to choose the phone company for broadband services, so
much of the reduction credit to the bundle strategy is introduced by positive selection.

    SBC is about to unveil a campaign to resell satellite video, with a marketing budget so high
they've told wall street to drop profit forecasts. I'll be reporting some remarkable ways to have
better service combining satellite video and DSL. Preloading the twenty popular moviesto the
home gateway with hard drive is one. But simply marketing a bundle is a very limited profit

Veterans of Past Murderous Campaigns Are Leading Haiti's
New Rebellion

Published: February 29, 2004

ORT-AU-PRINCE, Haiti, Feb. 28 — The armed men trying to seize power in Haiti are led by
death-squad veterans and convicted murderers, according to American officials and human rights

“The U.S. doesn’t really have broadband, and I blame the FCC. 200K and 768K is not broadband
today.” Masayoshi Son, CEO, Yahoo BB The third internet is fast enough to watch, I believe, and
the technology is ready at a modest cost. Germany is probably the most backward of developed
countries, with little cable and DT effectively imposing bandwidth charges on all DSL carriers.
Belgium, soon to be followed by Holland, France, and Britain, will soon start to catch up.

 Jennie was dressed all in black when her new iPod came, a silhouette ready for an Apple ad. She
started dancing before she opened the box, and will be "dancing all night" for a long time. An
everyday miracle of the 21st century, which DSL will bring to hundreds of millions.

India: next year will be really big
Tata, Reliant have backbones, international rates will be controlled
Indian regulator Praveen Sharma heard from John Gage of Sun ―Our customers in India need
better and cheaper internet connections. These are India’s growing industries: software, call
centers, engineering. ― Sharma replied immediately: ―We know the problem well, and will make
the necessary changes in our international connection system in a matter of months.‖ Those high
international rates are the primary obstacle to DSL, local providers tell me, forcing them to price
higher than the market can accept. Reliance, Tata/VSNL, and the government carrier, BSNL, all
have high capacity backbones across the countries, often supplemented with metro optical

200,000 lines is the short term objective of Tata/VSNL, a world class corporation with the
resources to quickly serve millions. The government carrier, BSNL, and Reliance, also have run
fiber to most major cities, and are anxious to add the last mile DSL connection.
50% is 12 cities Tata/Asia Netcom

India's BSNL looks to Korea
KT exporting services and network engineering
Texas Instruments India is one of the world's major DSL designers, and of course India has supplied much
of the leadership for Silicon Valley. Despite that, India is the greatest untapped market for DSL. Suddenly, I
hear Tata/VSNL and Reliance are ordering hundreds of thousands of DSLAM ports, with plans for millions.
Both have fiber backbones barely utilized to the major cities, and international alliances for bandwidth, until
recently the bottleneck. Tata/VSNL is negotiating to buy Dishnet, an early DSL provider with xxx customers
and past ties to Covad in the U.S.

    The dominant national wireline carrier, BSNL, now has a partner whose extraordinary achievements is
acknowledged even by the President of SBC. "You can't compare us to the Koreans," Bill Daley replied
when I pointed to them as far ahead of anyone in America. The Indians apparently came to the same
conclusion, and knew where to go for assistance in rapid growth. While India is a nation of a billion people,
internet users are highly concentrated. "I can reach half of the market with 12 cities," xxx of xxx calculates.
Most live in apartment complexes, with short loops similar to the Korean pattern.

   In Geneva last fall, I had the opportunity to join a conversation with John Gage of Sun and a senior Indian
official. Gage pointed out that his clients in India were handicapped by the high cost of international
bandwidth, and was re-assured that the government both knew the problem and was about to solve it.
Since then, xxx international fiber xxx. I'm half a world away, writing in New York, so I hope some of my
readers in India email me more information.

Verizon's Tauke: results are urgent
Tom Tauke of Verizon has advice for the FCC that is right to the point:
"The policy for wireline broadband is still stuck in the past. We need the FCC to be bold. We need
the FCC to be clear. We need the FCC to act with the determination and a sense of urgency at this
moment in time when change is so very crucial," he said. "So today, I want to renew my plea - we
need not only new wires and new rules in the U.S.-we need new thinking and a new commitment to
getting something done now!"

Editoral: "Get something done now"
Tauke's absolutely right. I agree the FCC should be "clear": if the bells want to keep the
enormously profitable local phone franchise, they must offer all their customers the internet at
world class speeds, and not leave America behind. The FCC should enforce the agreement Ed
Whitacre freely made in 1999 to offer broadband to 100% of SBC's customers. Bill Kennard
believed Ed's word was good on that one, a bad gamble. The FCC should be "bold", and consider
the fine Ireland threatened Eircom with until they agreed to offer DSL in every town over 1,500
people. Qwest, facing an $800M fine, might expand their coverage to the 90% level of Verizon and
Bell Canada. Or look to North Carolina, where every BellSouth CO in the state got DSL in 2001,
the first state in the union. BellSouth had no shortage of investment capital when North Carolina
suggested the state would fund competitors if BellSouth didn't find a way. The FCC needs "a sense
of urgency," and should pull everyone off proceedings like the rate hike the bells are requesting on
UNE's. Let the Bells first solve what Mike Powell has called his main focus "Delivering affordable
broadband to all Americans."

Ivan threatening comments about Tauke and

Telefónica Sí
2.7M lines, adding 400K/quarter
When I first reported Telefónica seeking bids on a million lines, people were amazed, and in fact
they built out slower than expected. Spain is hitting takeoff, and Latin America is recovering.
Telesp in Brazil is close to half a million lines, Germany over 200K, and CTC Chile more than
doubled to 121K. Telefónica Latin America reached 800K and Telefónica Europe 1.9M. Video is
set to grow rapidly.

Bill Gates ―SIP built in‖
―Everything sort of moves to Internet-based standards. SIP is something we really believe in and
we've got built in. The relationship between the PC and the phone will change. We're particularly
excited about IP/TV where if you want to send video over either an upgraded cable network or
DSL network we've got a whole suite doing the screens, and inserting the ads, measuring things,
connecting the billing systems. This IP/TV thing is a huge investment and we've got some pilot
customers working with us on it.‖

   Gates, now approaching 50, knows to project the future based on underlying technology and
economics. ―The costs of Gig-E just fits perfectly with the software work we've done. Actually
we're getting even slightly more response to the IP/TV in Europe than in the United States. ― In an
interview with Ovum’s Julian Hewett

Towards TV Quality DSL Networks
DSL can be more than email downloads
I asked last issue about how to measure DSL service reliability, which is crucial to offering video
over DSL. Bell Canada's and SBC's 3 Mbps service should be fine for watching a movie, which
requires between 1 Mbps and 2 Mbps sustained throughput with the current versions of MPEG 4,
Real, Windows Media Player, and other codecs. My amateur encoding of DVDs at 1.2 Mbps look
darn good, and at 2.2 matched the original to an untrained eye. Close to a million users around the
world already receive TV over DSL lines, typically as 4 Mbps MPEG 2's. The Japanese and
Koreans led the way with 7 Mbps DSL now going to 20 Mbps and more, and now we're seeing but
Bell Canada and now SBC regularly offer 3 Mbps.

"Why did DSL Prime ask about the most effective and inexpensive ways to measure actual
throughput last issue? As a telco field rep, all I test is sync. If I want to know performance, I'll
occasionally plug in my laptop and go to a speed test site. That's enough for what we're doing
today." A similar opinion came from a CTO, "I don't think it's worth testing the network regularly.
That's part of the service on other products, like T-1 and frame, but it would add cost. I think users
prefer the current 'best efforts' system. Imagine all the paperwork if every 1,000 user telco had to
file FCC reports on their service quality."

Error noted: An email from Nikos Theodosoupolous on another subject reminded me that we
disagreed ten months ago about the near term size of the VOIP market. Cisco had told him they
expected major growth from VOIP, while I was just back from VON conferences that were living
on hope, not orders. Once again, I'm glad to be wrong, and it’s wonderful to see VOIP taking off
after so many years. Cisco owns the enterprise market for VOIP, although Avaya, Nortel, Siemens
and others are crowding in. Currently Cisco hasn’t converted to SIP, although even their people
see that as essential.


From Peter McCauley as the DSL Forum meeting is about to begin ―With the work over the last
two weeks at UNH and this week in Brussels we should have a good ADSL2+ interoperability test
document at the DSL Forum. Just in time for the many ADSL2+ triple play implementations
starting this summer.‖

"What's with this n-ed stuff in your last issue? I'm guessing maybe you mean <deleted> and are
concerned that some people's spam blockers would prevent this from going through? My opinion
is that as a journalist, you should spell out your words. ;-)" If I printed <deleted>, many readers
wouldn't get the issue. But I think everyone realizes the word Notebaert used to describe his
service without extras.

"I am curious about higher speed ADSL vs SHDSL. What will happen in the market? The latest
ADSL spec includes a mode with a 3 Mbps uplink, and some DMT advocates tried to persuade the
industry to shift because SHDSL may cause more interference. The telcos resisted, with Verizon,
DT and others offering SHDSL because they needed a product that worked already. Ericcson,
Broadcom, and TI all announced 3 Mbps ADSL Annex J, but I've yet to receive a single report that
it's working in the field.
The same writer asked " Any feel for xDSL bonding and where it is going?" I think it makes plenty
of sense, but the companies active (Actelis, Spediant, Hatteras) haven't found a customer base.
Kermit Ross writes me that telcos have twice the copper they need in many locations, while line
counts go down.
Two notes came from the world's largest network vendor, whose ADSL effort has lately been
invisible. "Why aren't you writing about the European ADSL video vendors using our routers for
their backhaul?" wondered one. His colleague writes ― "DSL Prime should focus more on xDSL in
the business market. Europe is dumping frame lines by the thousands and switching to xDSL.
SMBs in the US are up to 60+% broadband. Revenue wise to the ILECs/PTTs this is a huge
change. For businesses it obliviously has huge implications in the type of apps they can run and
how their networks are built. Seems like a story worthy of some additional coverage:)" If I tried to
cover many stories about metro routers, I'd be stretching DSL Prime. Interesting things are
happening in the business market, and I wish I could cover more. Mostly, though, is that I never get
caught up even with my best stories, and wish I could write faster. I also sometimes wish I could
have a life. Too often, I'm working and responding to emails from Europe coming in at 4 a.m. New
York time.

Tiscali peach: 5M sterling in ads
Fair wholesale rates bring UK choices
Ben Carter describes a Tiscali ad that shows the advantage of speed, the image first "looks like a
naked bottom, but is slowly revealed to be a peach - emphasising the problems associated with
downloading pictures via a slow internet connection." That surely calls attention to DSL, as does
Tiscali pricing a low speed 150K DSL line at the same 16 pounds they charge for dial-up. They
have a million DSL customers across several countries. Britain's OFCOM regulator set the
wholesale DSL price, without ISP service, at about 60% of the BT retail price, allowing Freeserve,
AOL Europe and Tiscali to profitably go for market share.


   ·     eAccess in Japan reports 1.29M subscribers and projects a profit this year. "Higher
         network utilization resulted in significant decreases in backbone costs per subscriber and
         lower capex and depreciation."
   KPN will hit a million DSL subscribers Q2, Reuters reports from the CEO.
Geday of Globespan reported that 10-15% of DSL subscribers in Korea and at Yahoo BB are
upgrading their modems, resulting in a demand for chips beyond just the new customers.

Encouraging comment from a user at DSL Reports "The cable companies just aren't with it. DSL
stays capless, DSL eventually wins." I'm working on a piece "Verizon: Unlimited Means
Unlimited," about a strategy I think very smart battling cable.

Nearly everyone in the industry is occasionally available to the press, although the Times gets their
calls answered faster than DSL Prime for good reason. Mike Powell is a crucial exception. He
refuses most interview requests, including mine, since becoming chairman, which means my
stories about him are not as accurate as they should be. Powell was open and interesting to talk to
as a commissioner. The Denver Post calls the other hard to question figure still in office, Qwest
owner Phil Anshutz, a "reclusive billionaire." That’s probably a smart move for someone who had
to pay New York State $4M in a scandal and controls a company facing many indictments. Bernie
Ebbers and Steve Case were also unavailable. The results of MCI, AOL, and Qwest are yet more
evidence that folks who duck the press often have a story worth finding.

   Reporters on deadline, myself included, rarely make the time to do ―investigative reporting‖
even when they suspect much is being held back. In hindsight, the info was there for AOL, MCI,
and Qwest problems, but I and most others missed it. For the most important companies in my
coverage, like SBC and Verizon, I look deeper, but it takes a lot of work. A few reporters have
publications that can support them for a closer look, like Matt Richtel at the New York Times.
He’s still new to the beat, but his Utopia story – including the desperate move to block it
politically and the strong community support – stood out. His deadline AT&T Wireless reporting
had strong common sense that implied understanding. His impact piece after the deal went down
was dramatically different from most of other’s narrow industry focus. The head asked the right
question ―A $41 Billion Telephone Deal, but What's in It for Consumers?‖ and the lead had an
answer – an improved network – that reflects the opinions of several analysts. Many telco stories
affect all the readers of the paper directly, but reporters (myself included) far too often ignore the
results beyond the companies and their investors. Liz Douglas at the L.A. Times and Jim Hu as
CNET often brought a similar perspective.

stock market

mPhase's financial report reveals the company is chasing another opportunity beyond TV over
DSL. They've contracted with Lucent/Bell Labs for a nanotechnolgy product.

New Visual put out a press release claiming, "Embarq operates faster and reaches farther than any
other broadband semiconductor. And best of all, an upgrade to Embarq's technology does not
involve costly infrastructure improvements to the telecom network." They have provided no
evidence that has convinced me they have any real product with advantages, and have many
questionable actions and business associates in the past. Great caution advised - it's well past time
for a company like this to have easily tested and reproduced results.


Jim Moynihan writes "I am taking over coverage of the US wireline space at Merrill Lynch. Adam
Quinton is moving up to managment." Lead at Merrill is a tough, pressure cooker job, where if you
sneeze the market moves. When Quinton put a "sell" on SBC, it moved $8B. March 16 and 17th,
Merrill has a major conference, and I hope to see many of you there.
todo - mphase, TI,

To top