Wednesday, July 2, 2008

We told you so: FairPoint’s phone-line takeover is as bad as regulators feared.

Published in the Portland Phoenix

We knew it would be bad. Heck, beyond all the ink in all the other newspapers, we at the Portland Phoenix printed 4500 words over the course of six months explaining what was wrong with the Verizon-FairPoint merger, in which a North Carolina-based little-phone-company-that-could spent $2.3 billion of mostly borrowed money to take over the northern New England operations of one of the world’s largest telecommunications companies (see “A Bad Idea Triumphs,” by Jeff Inglis, February 29).

But it is with a distinct feeling of dismay (though perhaps just a touch of schadenfreude) that we report that the change-over has been more disastrous than even we thought: FairPoint is performing terribly now, and all signs point to the situation getting far worse, and probably never getting better.

Let’s move past the MISSING ONLINE BILLING SYSTEM that has customers in Maine, New Hampshire, and Vermont upset at having to buy stamps to mail in their payments for phone service. That’s still not resolved, but it’s relatively minor — and the company says it’ll be fixed by late fall. Liberty Consulting Group, the Pennsylvania-based company monitoring FairPoint’s transition for regulators in all three states, says it shouldn’t be a big problem. (You’ll see shortly that neither FairPoint nor Liberty is establishing a very good track record for this sort of promise, but there are much bigger fish to fry than complaints about adding 42 cents to everyone’s phone bill.)

We can less easily dismiss the fact that more workers have left the company than FairPoint predicted, leaving the new outfit SHORT OF EXPERIENCED WORKERS at a time when customers need reassurance — which usually comes in the form of speedy, competent service. That goes for both in-person physical repair work and over-the-phone support.

FairPoint had said that, upon closing the deal, it would hire an additional 675 employees in northern New England. But as of March 31, according to a report from Liberty, the company had 10 percent fewer employees than Verizon had had 10 months earlier, meaning it needs to hire replacements for roughly 270 people before economic-development number-crunchers can even begin to count any “new” workers. FairPoint corporate communications manager Jill Healey Wurm says the company needs to hire a total of roughly 900 people, but wouldn’t give a reason for the increased number. The company had hired 260 people as of the end of April, the most recent numbers Liberty or FairPoint have disclosed.

Liberty, though, says not all 900 positions need to be filled, and has claimed that only “key” positions do — while simultaneously recognizing that defining the word “key” is ... well, key, and refusing to define it.

But honestly, those issues too are small potatoes compared to three other problems lurking just under the radar, hidden in plain sight, in Liberty’s reports — all of which are available on the Web sites of the Maine and New Hampshire Public Utilities Commissions (the NHPUC one is much easier to find and use, atwww.puc.state.nh.us/Telecom/FairPoint.htm).

Moving at a snail’s pace
The first major problem is the speed of FairPoint’s takeover from Verizon, which is ALREADY FOUR MONTHS BEHIND SCHEDULE, a delay that means customers will not see the lower phone rates promised by FairPoint until December at the earliest, rather than August, as regulators had hoped. (In December, FairPoint customers in Maine will get a credit retroactive to August, totaling around $20 per phone line.) The original transition plan gave FairPoint four months after the date the deal actually closed to prepare to take over all phone-system operations from Verizon (an event called the “cutover”).

But months before closing, FairPoint was saying it would need more time to get ready, in January estimating it would need five months post-closing. Liberty’s January 14 report, its second monthly summary of FairPoint’s preparedness, called that deadline “very aggressive” and expressed “doubts that FairPoint can meet” it. In its February 11 report, Liberty was even more worried, calling the five-month schedule “extremely aggressive.”

By Liberty’s March 7 report, the closing was slated for March 31, and FairPoint was saying it needed six months post-close, delaying the cutover to “late September” at the earliest. While Liberty called the delay “helpful,” the consulting firm wrote that it was “too soon to assess the likelihood that FairPoint will be able (to) meet a September cutover date.”

Its April 10 report also saw Liberty saying it was “too early to judge” FairPoint’s ability to take over in September, but by May 9, Liberty was calling it “unlikely” that FairPoint would be ready in time.

And a month later, on June 6, Liberty’s report just plain said it: FairPoint’s four-months-plus-two-extra target of a September cutover was “unrealistic” because testing was nowhere near complete, neither of new software created by FairPoint to handle the former Verizon systems, nor of connections with other telephone companies. Liberty then recommended the cutover be delayed another two months, and happen sometime in November, saying it did “not anticipate any substantial roadblocks to FairPoint’s meeting that date.”

On June 17, FairPoint acquiesced to Liberty’s judgment. But before November, there are many important milestones for FairPoint to hit, including making sure it can import Verizon’s customer and wiring data into its computers without problems, finishing developing its software, hiring staff, and then training them to use the software that’s not yet developed.

Maine officials, and FairPoint, remain convinced the cutover will happen. “Liberty believes FairPoint is up to the task,” but just needs more time, says Fred Bever, spokesman for the Maine Public Utilities Commission. FairPoint’s Wurm takes a less-reassuring tone: “The cutover is going to happen in a very reasonable and hopefully seamless way.”

Emergency calls
There is one more project FairPoint needs to complete — or actually start — before it’s ready for cutover: Maine’s emergency calling system, E-911. It’s a core issue, and one we talked about though nobody else did: FairPoint is the company handling emergency calls from Mainers (and New Hampshirites and Vermonters) in dire, life-threatening situations. Which is why it gives us no pleasure to report that FairPoint is UNPREPARED TO HANDLE 911 CALLS.

To date, broken telephone systems have temporarily blocked 911 callers from reaching three dispatch centers in Maine, and have caused problems in New Hampshire as well.

First struck was the Cumberland County center in Windham, which serves nearly 70,000 people in 17 communities, and lost its ability to accept 911 calls five times on April 17 and 18. The phone company is supposed to maintain backup systems to recognize such failures and immediately re-route 911 calls to other offices, such as the state’s main dispatch center in Gray. But on one of those five occasions, the re-routing took more than 30 minutes.

When the dispatch center lost its connection again on May 16, the switchover took more than an hour, which cost the company $25,000 of the $6.4 million the state pays every year for handling 911 calls.

And another failure May 17 took 15 minutes to re-route calls. That’s not how long it took FairPoint to fix the broken system — that’s how long it took them to arrange for any 911 calls coming in to be answered by a human being who could help them, rather than hearing a busy signal or endless ringing.

The fix was actually relatively simple — there is now a physical transfer switch (much like a light switch) in the county dispatch center, and when the system goes down again, rather than relying on the questionable automatic system, a dispatcher will just reach over, flip the switch, and 911 calls will go to the state police in Gray.

On May 27, though, that office lost its connection for 10 minutes, and on May 28, a second malfunction caused a technician to shut down the system for seven hours. Another dispatch center was able to pick up the slack, but still, two people who called 911 were disconnected mid-call. That led state officials to demand physical transfer switches in six more dispatch offices.

And then between June 13 and 15, the Penobscot County dispatch center in Bangor, serving roughly 150,000 Mainers, lost service twice, once for six hours and the second time for more than 30 hours. In both cases, calls were routed to another dispatch center, but the first one was not detected by any automated system or even any dispatch staff — a citizen called 911, got a busy signal, and called her local police department’s non-emergency number to seek help.

On June 28, dispatchers at a 911 call center in Concord, New Hampshire, had trouble reaching police and fire departments in southern New Hampshire because of telephone problems lasting roughly 90 minutes, according to Foster’s Daily Democrat.

FairPoint says it has found and fixed the problems, but they’re not drawing anyone’s attention to the fact that soon we’ll be far worse off. Company officials and regulators agree that the systems that have been breaking down, while FairPoint’s responsibility, are actually the old systems created by Verizon. The problems likely “would have happened under Verizon,” says Wayne Jortner, senior counsel at Maine’s Office of the Public Advocate, which represents the public in utilities-regulation cases and has been a major player in the FairPoint dealings.

But soon, those old computer systems will be replaced with new ones created by FairPoint. Except Liberty’s reports say FairPoint’s SOFTWARE FOR HANDLING 911 CALLS IN MAINE IS NOT EVEN READY FOR TESTING! Liberty’s May 9 report says “a working version ... is not planned to exist until August.”

We can take some comfort in the fact that FairPoint officials and Maine regulators say they won't allow the cutover to happen until the state's 911 system is fully operational. But even if testing starts on schedule next month, any major problems could cause yet another delay. And while Verizon’s gear is old and breaking, FairPoint has nothing — nothing — to replace it.

Parting with cash
FairPoint will need to pay millions of dollars for all these new systems and testing and staff and training. Many of those costs were predicted before the deal was approved, though they included a few surprising financial assumptions by FairPoint (see “No Raises for Seven Years,” November 16, 2007; and “No Raises — It Gets Better,” November 20, 2007, both by Jeff Inglis).

But the four months of delays in the cutover will cost FairPoint $66 MILLION IT WASN’T PLANNING TO SPEND. And every month of delay beyond November will cost another $16.5 million. That money is paid to Verizon by FairPoint as, effectively, a lease of Verizon’s staff, software, and other behind-the-scenes systems.

And FairPoint has just issued its first post-purchase dividend, unloading $23 million in cash to its shareholders, which is money it can no longer spend fixing problems, or making service better. The company says it will have enough money to do what needs doing.

But to a pessimist’s mind, FairPoint is cleverly positioning itself to cry “poor” to state regulators if it runs into unforeseen expenses at some point in the future. Without those millions — and any other millions it may hand out to shareholders down the road — the company will actually be poor, and will be telling the truth if it asks for emergency rate increases or extensions on other commitments. (Maine, for example, has “required” FairPoint to expand the proportion of phone lines that can handle high-speed Internet service from 70 percent to 90 percent over the next five years, but then said that if the company hasn’t done so in time, it can have an extra year with no penalty.)

Jortner says concerns about FairPoint’s financial model failing are “absolutely valid,” though he takes pains to say “we’re not predicting that at this point,” and to note that the regulatory approval was structured so that if FairPoint is running low on cash, “it’s the dividend that gives,” not cash to run the phone system.

Even to an optimist, FairPoint is putting itself in a position with relatively little wiggle room. The company just spent $15 million on new trucks, none of which run on biodiesel or ethanol, Wurm says, though the company told regulators its financial model didn’t include any allowance for gas prices to increase. On top of that, with transition delays, fewer workers (none of them fully trained on FairPoint’s systems), and major software elements not even ready for testing, the company’s time is running out.

More alternatives
And the pressure is really on. Nationally, millions of landline customers are canceling their service — on average, 350 customers in northern New England do so every day. (Verizon numbers indicate as many as eight percent of customers disconnect in any given year.) They’re moving to using just cell phones, or pairing cell phones with Internet-based telephone service, such as TimeWarner Cable’s Digital Phone service, which allows TimeWarner to deliver a customer what is called a “triple play” — cable television, high-speed Internet, and telephone service — over one wire and paid for on one bill.

FairPoint’s business plan depends on the company retaining more of those customers than Verizon did, and having fewer of them seek communications services — including high-speed Internet access — from other companies. That will take some doing.

A key element of customer retention will be FairPoint’s own “triple play” service. Company spokeswoman Wurm observes that because of the cutover delay, FairPoint has partnered with DirecTV to create something like a “triple play,” with DirecTV providing satellite television and FairPoint delivering telephone and Internet service. But FairPoint is using regular telephone wires, which in many cases are decades old and may need replacement to carry data as well as voice traffic. And even when equipped with top-notch technology (which costs millions), the copper telephone wires FairPoint is depending on transmit data more slowly than fiber-optic connections, which are the real future (see “Internet Disconnect,” by Jeff Inglis, August 24, 2007).

In many parts of Maine that have telephone service from FairPoint, RESIDENTS AND BUSINESSES ARE UNWILLING TO WAIT for high-speed Internet access; wireless-Internet providers have set up shop and are expanding rapidly to meet demand, in communities from Presque Isle to Bar Harbor.

Even Democratic Governor John Baldacci has seen the future. On June 10, he attended the ceremony opening a fiber-optic line offering businesses in Bangor access to an all-fiber network reaching to Augusta, Portland, Portsmouth, and Boston.

Who owns that network? Not FairPoint, still struggling with software development and testing, but Lewiston-based Oxford Networks, which has more than 600 miles of fiber strung throughout Maine already. The future is here. Maybe one day FairPoint will arrive, too.