DSL Prime: Statistics and Lies

While Point Topic and the DSL Forum have gathered great statistics describing the present, basic information about future plans is being withheld even by publicly-owned companies.

by Dave Burstein DSL Prime [January 5, 2004]

December 31, 2003 62 million to 64 million
Q3 55 million say Point-Topic/DSL Forum
Tim Johnson of Point-Topic's public numbers on DSL are at least as accurate as most of the research reports costing thousands of dollars, and the DSL Forum does us all a service promoting them. They show that DSL is handily beating cable worldwide. Johnson predicts 7 million more for Q4, with my guess a little higher. China points up from the 2.2 million Q3 figure, Japan is close to a million, France is looking for half a million as competition there sets the European model, and many are optimistic. DSL now costs only a few dollars more a month to deliver than dial-up, providing logic for the switch where telco pricing is competitive. The rankings, in thousands of lines:

Japan 9,228
USA 8,243
China 7,817
South Korea 7,069
Germany 4,252
France 2,429
Taiwan 2,374
Canada 2,027
Italy 1,672
Spain 1,433
UK 1,414
Brazil 837
Belgium 706
Hong Kong 660
Netherlands 643
Sweden 508
Denmark 416
Switzerland 383
Israel 358
Australia 333

On September 30, China may have only been in third place, but they have since passed the U.S. and should soon pass Japan as well. Chinese numbers from different sources conflict, however.

India and Mexico are the two countries with the size and economy they should be on this list. I'll be reporting some optimism from India, while I hope the subscriptions that just came in from Telmex are a sign they will grow as well.

The table is totally different when expressed as DSL per hundred phone lines.

South Korea 30
Taiwan 18
Iceland 17
Hong Kong 17
Belgium 13
Japan 12
Israel 11
Singapore 11
Denmark 11
Canada 10

While the Q3 growth (new lines added, in thousands) points out some trends

China 2,217
Japan 971
USA 762
France 390
Germany 387
United Kingdom 343
South Korea 258
Italy 237
Taiwan 232
Canada 159

All but Korea (at saturation) and Germany have a good chance to do substantially better in Q4. Verizon's partial recovery should help the U.S. total.

20 percent upgrade to VDSL in South Korea
Who says consumers don't want speed?
The service isn't even available in most of the country, but Johnson reports nearly 2 million South Koreans have upgraded to 10 million to 50 million VDSL speeds. The evidence is overwhelming: VDSL is the product of choice anywhere you have short loops (building, MDUs, and up to 3,000 feet). That's an significant percentage of city dwellers—40 percent in Norway, higher in Italy, higher still in newer Asian cities with many apartments. It also includes most new buildings, although fiber is soon the right choice. 20 Mbps to 30 Mbps ADSL2++ a plausible alternative with loops of 3,000 to 8,000 feet, although customers are looking for higher upstream speeds as well, according to all the cable market research as well as DSL Reports.

Bob Crandall's questions
Getting the facts on deployment requirements worldwide
Bob Crandall has the force of personality of a distinguished corporate insider, leads D.C.'s best procurer of economists at Criterion, and has direct support from Verizon for his work. His access to data should be extraordinary. Despite that, our conversation floundered because some very basic information about broadband deployment was hard to confirm. We were trying to determine what's actually needed for the next stage of deployment, to be able to answer his question to me about the cost.

Without basic facts, policy studies are flawed. So I'm writing to several major carriers asking:

• As of January 1, 2004, how many COs have DSL and how many are not covered? What percent of homes can receive DSL?
• What percent of your residential customers are within 8, 12, and 18,000 feet from those COs?
• Of those not within 8, 12, and 18,000 feet, how many are served by DLC remotes? Of the DLC customers, how many are not within 12,000 feet of a fiber-fed remote? How many are served by remotes with fewer than 300 lines?
• What specific other technical problems prevent offering DSL to what specific percent of home customers?
• Based on your 2004 capex budget, how will those numbers change? Can you make any projections further into the future? (Included especially because SBC has indicated a heavy build schedule for Q1, 2004 and I wanted to credit that.)

I'm not asking much, just the basic data you already compile for your own network planning, aggregated. This list is preliminary; I welcome suggestions on what we need to know to understand the network build. This is all about achieving Mike Powell's goal, "affordable broadband service for all." Getting the facts right is step one.

Copyright 2004 Dave Burstein.
The DSL Prime Newsletter is reprinted with permission.

"The power of the printing press belongs solely to those who own the presses"
—A.J. Leibling

The Internet is the cheapest printing press ever invented.