DSL Prime Congratulates SBC

Congratulations to SBC for a reduction in churn and for the unveiling of naked DSL.

by Dave Burstein of DSL Prime and Future of TV [May 5, 2005]

Congratulations to SBC for Churn Reduction
Apologies for not seeing the figures
"SBC the last few quarters has seen a drop of 60 basis points in DSL churn, or 20 to 30 percent," CFO Rich Lindner answered Blake Bath in the quarterly conference call. Working from those numbers, SBC now has DSL churn of between 1.4 percent and 2.4 percent, down from 2-3 percent a year ago. So I was wrong to write, "None of the U.S. telcos release churn numbers for DSL, although they are clearly material to any analysis." In fact, two days before SBC had addressed the subject, after several years of refusal, like the other bells to answer my questions on the subject.

Average revenue per consumer, Lindner added, is in the low to mid $30's; ARPU including business customers is in the mid-30's. That's significantly higher than the $19.95 and $29.95 promotions publicized, indicating the terms of the deal prevent many consumers from taking advantage and extra fees are buried in the fine print. (People hate that.) ARPU is down slightly, but so are costs; Lindner, like Stephenson, notes that SBC has been able to lower prices while raising margins.

20 to 25 percent of the customers are paying $10 more to raise peak speeds from 1.5 to 3.0 Mbps. Lindner was happy this raised ARPU, but Verizon's "we promise everyone up to 3 Mbps" may actually be more profitable in the competitive U.S. market. The extra customers won (or not lost) probably are more significant to revenue, but I'd need more precise data to calculate that.

Lost customers are enormously expensive
Glen Campbell of Merrill points out that the "marginal" customer lost or gained is high-value. The usual figure for "value of a customer" in U.S. telephony and cable is $3,000 to $4,000, based on stock quotes and prices of companies and divisions sold. That represents an "average," but with the bulk of costs in telephony fixed, the additional customer disproportionately contributes to profit. The implication, as noted by Ed Whitacre of SBC, is that something that pulls away customers, like unbundled (UNE-P) losses is devastating to the bottom line. Campbell is modest in style, but has consistently provides more thoughtful work than the flashier analysts often in the newspaper.

$3,000 to $4,000 has long struck me as too high a value of the income stream per customer, incidentally, even if the stock market produces that figure. An old Graham and Dodd fundamentalist, I think the stock prices remain high, despite drops of half or more. However, a similar analysis of the income streams of other industries show them even more out of line, which would imply the entire market is way overpriced. I refuse to draw that conclusion, as after several decades watching market prices I've learned the connection between market levels and economic reality is stretched too far for me to rationally explain.

Going nude in San Antonio, Keller
Lindner "Expects we will do trials of naked DSL, especially bundled with wireless." Dick Notebaert of Qwest was the first Bellhead to suggest the income from selling DSL to people without landlines was more than the loss from the switch to independent VOIP, but expect SBC and others to move slowly on this. Verizon was applauded in the press for taking a very tentative first step in this direction recently, then had to correct the stories to explain it only applied in certain cases. I can't call Verizon and order DSL without phone service, it turns out. But if I switch my phone service to another carrier, they won't shut down the DSL. Even that option doesn't yet apply in Verizon in Keller, Texas, or another two-thirds of their territory. Shaygan Kheradpir's programming department is apparently the bottleneck—the systems aren't ready outside the old NYNEX region.

Correction

• Ray Le Maistre of course is with Light Reading, not Lightwave. LR has more verve—and more scoops—than just about any other pub in telecom, either print or online. Linley Gwennap, whose surname I misspelled, is one of the finest chip analysts.

e-mail

• "Dave, I can personally testify to two items you had. Optimum Online is much faster than Verizon DSL. I tried Verizon (without the 12-month commitment) and dropped it after two months, so I'm part of their unpublished 2004 churn stats," writes Steve Cherry, and old friend. His IEEE article on Swisscom and Microsoft IPTV, loaded with data, recently caused a stir and I've been referred to it by several people.
  
  
• "If Electriphy is using QAM for VDSL, How do they manage to be DSM compliant? My understanding of all the work that our friends at Stanford were doing was that DMT was a must to do any effective DSM in the loops out there?" asks an industry engineer. While DSM is a natural outgrowth of Cioffi's Stanford work on DMT, Metalink's Amir Leshem two years ago showed how similar principles could be applied to QAM systems.
  
  
• Dan Foster at Megapath checked in with some impressive results for an independent ISP in today's market. They did $88 million in 2004 sales, which they report as EBITDA positive. Even private companies release results when they are that favorable.

Wall Street

• The NY Times has a healthy perspective on the proposed merger between the New York Stock Exchange and Archipelago, suggesting the positive effects might be "marginally lower costs and marginally less fraud." Marginal is nearly always accurate for the positive benefits of a merger; SBC Ameritech and probably Verizon GTE haven't proven out after several years, with the increased complexities of the larger organization negating most of the "synergies." The same is almost surely true of Cingular/AT&T Wireless and Sprextel—the payoff is coming from fewer competitors driving down prices, not operational efficiencies.

Job ads

• For DSL Prime job ads, visit the DSL Prime website.

Copyright 2005 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.