Comcast Finds Its Voice

To the outside world, Comcast Corp. has signaled that its telephony business is on hold, but behind the scenes the connection is far from dead.

The top U.S. MSO has quietly re-engineered its existing switched telephony business and is moving ahead with plans for its first voice-over-Internet protocol market test in 2004.

Industry observers seem to agree that this strategy is more prudent than a pure voice land grab, but now the question is this: Can Comcast can light up the phone lines soon enough to keep ahead of its Bell competitors?

Making the switch

If anything, 2003 has been the year Comcast brought phone home, working to revamp the circuit-switched telephony business it largely inherited from AT&T Broadband.

Of Comcast's 18 voice markets, 16 are former Broadband operations.

In addition to transitioning Broadband customers to the Comcast brand and service, the MSO also has taken an editing axe to the dozens of call packages AT&T offered.

Comcast boiled three international calling plans down to one, while paring domestic plans down to two basic packages, "just trying to simplify and create higher-value packages that include multiple long-distance features," said vice president of marketing for telephony Tom White.

The two domestic plans, first rolled out in Minnesota in May, include Connections, which offers select call features, local calling and options for per-minute, 180-minute or 300-minute long-distance calling; and Complete, which offers a full set of call features plus local calling and the same three long-distance options.

"It's something that allows our sales and care people to rally around a structure that is very easy to sell, and it maps well to what consumers are looking for in the marketplace, and much of that is due to what the [regional Bell operating companies] are doing with their package selling," White said.

Most of the 18 markets have now converted to this package structure for new customers, which offers a discount of 10% to 15% compared to comparable packages from other carriers.

"We're seeing a really good take rate early on in the markets where we have put those in place," White added. "We are also testing an unlimited type offer similar to what we are seeing out there from the RBOCs, and AT&T [Corp.] and MCI [Inc.] and others. We've got that in a few markets right now."

Wanted: profit

Meanwhile, the larger focus with switched telephony is on making it more profitable. That's a big step, considering the operation inherited from AT&T Broadband.

While Comcast is not certain how much the AT&T Corp. subsidiary lost from its telephony business in 2002, the best estimate puts the figure between $100 to $200 million.

AT&T Broadband "grew so quickly, because that was their business plan, that very basic things like credit screening and collections processing didn't even exist," said Rian Wren, senior vice president and general manager of Comcast Digital Phone. "We started the year with a run rate of about 15% bad debt and by the end of the first quarter we were able to get that down to 8%, and now we are somewhere around 7.5% to the low 7%, just by focusing on credit screening and collections processing, fraud management — these are all things that were not in place that we immediately put into place."

Comcast also picked apart its operating-cost line items, looking for savings.

It renegotiated the leasing deals with AT&T Corp. for Class-5 circuit-switch capacity, a move that on its own cut telephony operating costs by 25%. In addition, the MSO centralized its customer-care operations and is in the process of outsourcing its directory-assistance service to another vendor to gain better rates.

Comcast also revamped network-management and operations systems, eliminating a lot of the outsourcing to big-house vendors and in the process cutting those costs in half. With all of these changes, Comcast now expects to generate about 25% operating profit in 2004.

"That's a huge swing, and some of the ways we did this was based on the strategy," Wren said. "We decided proactively to slow the business down, to re-engineer the processes."

Subs are down

Nevertheless, the telephony subscriber numbers have dropped in 2003 — down from the first quarter's 1.41 million to a little less than 1.37 million in the second quarter, while revenue per subscriber dropped $8 from the second quarter of 2002 to the same quarter in 2003.

The loss in subscribers was prompted in part by the fact Comcast backed off its marketing efforts and by the fact that some customers have opted to drop a second phone line in favor of Comcast High-Speed Internet service.

On the revenue side, losses are attributed to paring bad subscriber debt and therefore reducing outstanding balances traditionally counted as potential revenue; the loss of some AT&T Broadband long-distance revenue during the transition; differences in the way Comcast accounted for federal and state surcharges, compared to AT&T; and the fact Comcast didn't market higher voice tiers as agressively during the period.

With the business reformed, Comcast now expects those numbers will rebound.

"Now what we are seeing is that the packages are in the marketplace and our people are incentived to sell the packages, we're seeing our take rate go up," Wren said. "We're actually seeing on the intake on our connections RPU is going up, because we are getting a little bit more robust of a customer. So I think it was a transition issue."

Comcast's approach to telephony has been more cautious, but it also appears financially sensible, according to Aryeh Bourkoff, managing director of equity and fixed-income cable and satellite at UBS Investment Research. For the first time, he noted, Comcast recorded a profit on the business in the second quarter.

"So far they've had the right strategy, to take a approach that focused on profitability first and cleaning up the legacy properties first — changing the strategy of the company, boosting the margins and I would say gaining financial flexibility, which is a necessary precursor to competing with the Bells and the satellite operators," Bourkoff said.

While it did slow voice marketing elsewhere, Comcast maintained a strong telephony push in Boston and Minneapolis-St. Paul because the business was good and the sales team was strong. With the overall effort now refocused and rebranded, "there are a couple other markets, potentially, that could be a little more aggressive in marketing telephone," Wren said. "So what I think you will see next year is maybe a handful now — we will be going from two to maybe five where it makes sense for us."

But while there may be selective footprint expansions for the circuit-switched telephony markets, no one should expect Comcast to light a fire under that product.

"Overall, I would couch that by saying that in the long run, we really are putting our investments on the IP side, and the main reason is it creates a better business model for us," Wren said. "We will fully own the assets, and therefore be able to get the full return on our capital investments. In the AT&T model you don't fully own the assets, so the return isn't as promising."

Picking up the IP phone

Indeed, work is now accelerating on the IP telephony front. Comcast already has a technical trial involving a group of its employees in Coatesville, Pa., powered by a Syndeo Corp. softswitch. Going forward, the MSO is going to expand that trial, adding softswitches from Cisco Systems Inc. and Cedar Point Communications Inc.

The key issue is system reliability and quality, with systems that recover quickly after a failure and offer the features phone customers have come to expect as standard.

"It's a whole bunch of nit-gnat things that we are feeding back to them, but in general we are pretty pleased with what we've seen so far," Wren said.

With much of the technical underpinnings coming together, the MSO is planning trials in two markets during the first half of 2004 — one in a long-time Comcast system and the other a former AT&T Broadband system.

With paying customers, these two trials are aimed at testing out marketing, provisioning, billing and customer-care elements. They'll also allow field-operations staff from the two selected markets to get into the act.

Other elements to be tested will include provisioning options, powering scenarios, wiring within the home and possibly a single bill for voice, video and data service.

For Comcast, the key to the market trial expansion is to nail down all of these variables before going ahead with any IP-telephony rollout.

"So the major salient differences between what we have done is we are moving more towards trying to commercialize the product. However, it will be in a controlled introduction phase, so we're not saying we are rolling out everywhere," Wren said. "What we are saying is taking the next step, to really say: Can we provide the quality and customer service and the offer set to the marketplace in a way that we are comfortable with? And then also get another set of feedback on that."

Capital construction also comes into play for the IP telephony trials, and the strategy there is to build the necessary fiber optic routes to make it easy to expand.

"That means that certain things like fiber rings — we will probably build them in a way that covers a lot more footprint than the first phase of the trial, because it makes sense to do it all at once," Wren said. "So we will phase our footprint build such that we build just enough to accomplish our trial objectives but do the smart things so when we decide to roll it out in a greater way, we don't have to go back and re-invent the wheel."

That said, the trials will focus on performance and quality rather than trying to rack up customer volume — participating subscribers will number in the few thousand, rather than hundreds of thousands.

There's potential for the IP and circuit-switched efforts to intersect. The MSO already is running a hybrid system in Detroit, where IP connections power the loop from the customer to the network, and a gateway translates the IP into circuit signals before sending it on to a traditional Class-5 switch that Comcast actually owns.

"We are next year going to be testing the reverse, where we actually can see if in fact we can maintain the constant bitrate loops and replace the switch with an IP switch," Wren added. If it works, that scheme could allow Comcast down the road to replace its leased switches with its own IP soft switches.

"I wouldn't tell you it is baked just yet, but I think we are very interested in doing it," Wren said. "We have some time — we have another two-year commitment with AT&T through the end of '05, and we also have extension provisions. So it's not like we have a time rush on us, but we are looking at how to test that next year in one of our markets."

Moving fast enough?

Although Comcast's careful steps into the voice world seem sensible, the warming competitive climate does bring concerns that the MSO may not be moving fast enough.

"I think the question eally becomes not waiting too long to get into telephony more aggressively and missing that window of opportunity," Bourkoff noted. "And as long as the company focuses on telephony some time in 2004, I think that Comcast will be very well positioned to grow the model and maintain the financial flexibility that is needed to compete with the Bells. If it goes beyond 2004 into the 2005 time frame, I think that Comcast begins to lose some of that competitive edge given the robustness of the plant."