In AT&T's (T) - Get Report cable television business, the once-vaunted prospect of phone calls over cable lines continues to take a back seat to ... cable television.

For AT&T Holders, an AWE-Inspiring Offer

As Ma Bell's Family Keeps Splitting, Should You Stick Around?

Gut Check on AT&T: Breaking Up Is Hard to Do Right

AT&T Sees Wireless Unit as Undervalued

Ready to Leave Home, AT&T Wireless Grows Revenue and Subscribers

Three years ago, when AT&T launched its expansion into cable TV by buying industry giant


, Ma Bell made it clear that a motivating factor was to provide multiple communications services through the cable lines: high-quality digital video, high-speed data connections and, perhaps most important, local telephone service.

"Today we are beginning to answer a big part of the question about how we will provide local service to U.S. consumers," CEO C. Michael Armstrong said in a statement then. The prospect of marketing local phone service under the AT&T brand name was "stunning," a TCI executive told

The Wall Street Journal.

But while phone service over cable lines has hardly been a failure for AT&T, it has been slower to develop and less profitable for the industry than other new services offered via cable television, say analysts.

Penetrating Analysis

At the end of the March 31 quarter, the

AT&T Broadband

cable unit had 700,000 local phone customers, amounting to 4.4% penetration of its base of 15.9 million cable subscribers at the time. By comparison, 3 million subscribers, or 18.9% of AT&T Broadband's customers, were subscribing to digital video services -- essentially, an expanded menu of cable channels. In the first quarter, the unit added 153,000 telephone subscribers, 204,000 high-speed data customers (AT&T controls the


(ATHM) - Get Report

high-speed cable modem service) and 340,000 digital video customers.

In another comparison, AT&T's rollout isn't significantly ahead of that of another major cable operator. As of March 31, cable operator

Cox Communications


said 134,000 of its 6.1 million customers, or 2.2%, were taking phone service from the company, ordering an average of nearly 1.5 phone lines per household.

At AT&T and other cable operators, what's going on is a response to competition, says Joe Galzerano, high-yield research analyst at

CIBC World Markets

. Though in 1998, the cable industry's competition appeared to be local telephone companies that were expanding into video, the more fearsome competitor has turned out to be direct broadcast satellite companies, such as

Hughes Electronics


, which market home satellite dishes to current and potential cable TV customers. Cable operators "went for the products that were easy to do, that they knew they had a competitive advantage in," says Galzerano. "And at the end of the day, I think that was the smart thing to do."

Many Happy Returns?

Salomon Smith Barney

cable analyst Niraj Gupta says that the telephone-over-cable product that operators are offering now has a lower return than do high-speed data services or premium digital video offerings. (He declined to talk specifically about AT&T, with which his firm has an investment banking relationship.) "Telephony at its core has really benefited companies like Cox Communications in terms of significantly improving customer satisfaction and reducing churn," or the rate at which customers drop cable service, says Gupta. Improved customer relations has been the single biggest benefit to cable operators so far, he says.

But another analyst says the return on cable telephony will probably turn out to be better than others expect.

Yankee Group

analyst Bob Lane says that, indeed, it's costlier for cable companies to upgrade their systems for telephone service than it is to upgrade for digital video or high-speed data. But, he says, the key assumption in calculating a return on this investment is estimating penetration -- what percentage of cable customers offered a service will take the company up on the deal. When consumers are asked what types of telecommunications service they'd like to be offered in a package, 80% ask for local and long-distance phone service together, he says, and 63% want cable, too -- about the same percentage of Americans who subscribe to cable. In addition, the Yankee Group's research indicates that one in five local phone subscribers would switch to another provider for the same price, and two out of three would switch for a 15% discount.

In fact, Lane points out that AT&T has gotten higher penetration for local phone service in its systems in which its available than it has for high-speed data in the areas where that service is available. In the first quarter of 2001, AT&T discloses, telephone service had 11% penetration in marketable homes and high-speed data had 8.3% penetration; both trailed digital video, which had 19.7% penetration.

"One of the things that makes phone so attractive is that it's already a big market," says Lane. "So if you're just entering the market, it's a huge growth opportunity."


How big an opportunity? Back in 1998 -- after the TCI deal was announced but before the company's subsequent deal to acquire


-- AT&T executives said they could reach $700 million in annual local telephone revenue by 2002, according to one analyst's report. Given the rough estimate that consumers spend $35 a month on local phone service, AT&T's current 700,000 local phone subscribers would generate nearly $300 million in annual revenue. At AT&T's current growth rates for its local phone business, that $700 million target is "definitely within reach," says Lane.

Of course, if AT&T didn't reach its goal, that wouldn't be surprising, either: Technology inevitably moves more slowly than expected. For example, the type of cable telephony described by Armstrong in June 1998 -- phone calls using Internet standards over cable lines -- has yet to be rolled out in force by AT&T, Cox or anyone else. It's arguable, say analysts, that this Internet-based telephone service will reduce the cost of providing cable telephony. But, says AT&T Broadband spokesman Steve Lang, there's no gear yet available for this type of telephony that's robust enough and has enough features to replace phone service from traditional local phone companies. "Maybe 18 months from now the product will be ready," he says, "in which case we'll be ready to deploy it."