DSL Prime: FIOS Fantastic

Verizon TV product needs just one more feature to make it a cable killer. If Verizon fails to go for the kill quickly, cable can fight back and win, eventually. And, from Paris, the tale of a co-founder who admitted his embezzlement.

by Dave Burstein of DSL Prime and Future of TV [September 29, 2006]

"I screwed up, I will be judged and convicted."

Xavier Niel has the honesty to admit he did improper business in cash and accepted a $400,000 fine. Niel also has personally donated 10 million euro for broadband for the poor of Paris. His company is raising the speed of France's internet to 100 megabits. The many telcos that not do as well by their customers, and the ineffectual regulators behind them, also should apologize even if no laws were broken.

Xavier Niel's billion euro fiber build is the ultimate proof that competition can work, despite its failure in London, Chicago, Barcelona, and other cities. His robots will go through the sewers of Paris, bringing an active fiber to every apartment. Speeds will be 50-100 megabits in each direction, the price around 30 euro with many extras. The stock took a hit the first day, but recovered after Niel explained his plans. This isn't a "build it and they will come" plan. They are going where Iliad/Free already has 15 percent of the market and hence the volume to be profitable over 4 to 7 years.

Niel's 10 million euro gift to a foundation will provide a connection to nearly everyone in the building served by Iliad's fiber, even if they can't afford the regular fee. The network will be open to competitors, with Neuf Cegetel likely to buy wholesale capacity. This was a very happy day for Paris. One industry expert in Paris writes, "I'm going to start making fliers to pass around the neighborhood asking people to switch to Free so we reach 15 percent and get fiber."

Ginny Ruesterholtz, Verizon's new President, is leading her first big public meeting as I send this out. Fortunately, the topic is FIOS and the news is good. I pulled over what will be part of the first Fiber News.com if I ever finish it. Ruesterholtz, like Randall Stevenson at AT&T, is in her 40's and represents a new leadership generation at the Bells. She started at NYNEX managing cable installers, ran Manhattan business, and is respected by her peers. If she and Toben do their job well, the inappropriate $20 billion gap between AT&T and Verizon market cap should narrow soon.

What Verizon FIOS Can Do
250/125 GPON, 100/40 BPON, Promise of Open Set Tops
Verizon has extraordinary things coming, although nearly none are part of this morning's announcement, which mostly repeated old news with a few optimistic projections. What they have on the way is amazing, even if they didn't discuss it today. They even are preparing the first real answer to "how can your video be better than cable," with an on-the-record plan to open the set top to all the video over the net. If they follow through, they will have accepted the most important issue in net neutrality—honest consumer access to content of her choice.

GPON can go 250 down, 125 up. Soon to start early trials, Verizon will continue using another wavelength just for video. They probably won't offer those speeds at the beginning, but the folks involved are confident it will be reliable. Dynamic bandwidth allocation means the 2.4 gig down, 1.2 gig up is effectively shared, so that 99+ percent percent of the time any user needing speeds in the hundreds of megabits can access them. Many important technical issues, including interoperability, are making good progress in FSAN.

BPON can raise speeds to 100 down, 30 up using similar techniques bandwidth sharing techniques. Until recently, effectively shared This is important because Verizon will have deployed between 7M and 9M lines of BPON before they have enough confidence to switch over to GPON. BPON is 622 down, 155 up, split up to 32 ways. That's considerably better than the low end DOCSIS 3.0 (160/120), and similar to the high end DOCSIS 3.0 (1 gig/100 meg, shared to probably hundreds of homes.)

Babbio and Seidenberg have been talking about a cable-killing open set for two years, and now TV lead Joe Ambeault says, "this release of Home Media DVR does not support that capability. It's on the roadmap for future releases." That would be Verizon's killer app.

Two years ago at PFF Aspen, Babbio explained, "I need to get cable out of the house. I make my money on voice and data, not TV. We're considering offering everything." Ambeault's comment to Mike Robuck of Communication Technology is the first public statement they are moving ahead. If they do, I'll be able to watch on their multi-room PVR everything from FIOS video as well as my choice of channels from the Southern Baptist Church, Google, Amazon, Al-Jeezera or The Jerusalem Post. I also asked Brian Roberts speaking at IRG, "would you sell Comcast's program package over Verizon's fiber network?" Roberts was poised. "I saw Ivan at USTA in Las Vegas, and he told me to be ready for that question. We agreed to say 'the devil's in the details,' so we'll see."

Verizon will do 3 million more homes passed this year (to 6 million), and has the technical capability to do 4 million to 5 million next year. The major bugs are solved, so it's just a question of whether they spend the money. Contractors including Bechtel, Dycom, and MASTEC assert they have plenty of installers ready to jump in. That would take Verizon to only to 12 million homes passed in 2008, and 18 million in 2010. By then, they intend to have dumped enough lines (line losses to cable and wireless, sell or spin-off) to be down to about 25 million homes on their network. They will have rebuilt 75 percent of the network in 6 years. This is actually disappointing. If they hadn't tried to keep spending down, they could easily have reached 20 to 22 million.

The take rate on data should be excellent. Customers are enthusiastic in a way I've never seen in U.S. telecom. The TV is working fine, and numbers are likely to pick up soon. 10 percent of TV customers are likely to switch as soon as asked because they hate their current provider, which means Verizon (and AT&T) should be able to exceed the modest public projections for 2006 and 2007.

What they announced: Almost nothing new, but pretty darn good
Essentially, they just repeated some of the existing good news to make sure Wall Street gets it. The speed increases and open setup described above are still unannounced unless they pull a surprise in the oral presentation or Q & A. Here's their data points:

• They confirmed their public plan for 18 million in 2010, a reduction from their internal goal. Still pretty darn good.
• Announced 100,000 TV customers, and projected 4 million in 2010, well above most outside projections. Expect 6 to 7 million data customers in 2010, reducing customer loss.
• Confirmed reduced cost of homes passed, improved maintenance results with a big payoff, and low churn. Churn of 1.5 percent is not far above the natural rate of Americans who move.
• Lots of nice touches in their software, which they moved away from Microsoft per WSJ.
• Decent financial return over 4 to 5 years, although they fudged by using operating and not full accounting numbers.
• Hooking up each customer and running video around the house is taking longer and costing more than hoped.
• Made clear they do not need "franchise reform" or the national telecom bill until well into 2007 if then. Noted they are only able to serve 1.8 million of 3 million homes already approved at yearend 2006.
• Works much than anything likely at AT&T Lightspeed. But costs much more to build.
• They will do apartment buildings (MDUs), but very few until next year.

For the first Fiber News, I'll have the more details from France, experiences from Calix's 100,000 lines of GPON in smaller telcos, GPON interoperability testing, Tellabs on DMA, Qwest's forgotten TV in Phoenix, and Japan's world lead.

Copyright 2006 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.