May 7, 1999

AT&T Conjures Up Its Vision for Cable, but Can It Deliver?

By SETH SCHIESEL

n a series of deals over the last year that culminated in its agreement this week to acquire Mediaone Group Inc., AT&T Corp. has committed itself to spending more than $90 billion on a technological vision that is largely untested and that may not exist anywhere on but AT&T's drawing boards.

The promise is great: a system that could allow AT&T to provide competition against Baby Bells in local phone markets through cable connections -- even while delivering interactive television, which could allow viewers to select their own camera angles or replays, and lightning-quick Internet access to households more accustomed to molasses-like speeds.

But the technical and organizational challenges standing between that vision and AT&T's present are so great that one AT&T executive likened AT&T's ambitions to launching a space station. It is one of the most grand technical voyages in communications since the old Ma Bell starting stringing telephone wires across the nation.

"We need to figure out how to build it, how to deploy it, how to support it, how to maintain it," C. Michael Armstrong, AT&T's chairman, said Thursday in an interview. Referring to new customers, he added: "The issue is not doing this for a few hundred people a month. It's for tens of thousands of people a month."

On Thursday, AT&T enlisted Microsoft Corp., the leviathan of software, to help in that effort. As expected, Microsoft agreed to invest $5 billion in AT&T, while AT&T agreed to use Microsoft's software in some of the advanced television set-top boxes that AT&T plans to use to bring its vision into living rooms.

But the $90 billion that AT&T has already committed to its vision is just to acquire Tele-Communications Inc. in a deal that was announced last summer, and MediaOne, in an agreement formalized Thursday. Now billions more must be spent turning relatively raw assets into systems that can deliver the dream.

It will probably take years for most people, even those in AT&T's cable markets, to see tangible results of the company's ambitions. And it is not clear if customers will end up footing the bill, though AT&T is well aware that its pricing will have to undercut local telephone incumbents if it is to have much chance of success.

Put simply, the main technical challenge for AT&T is to get cable television networks to do things that they were never meant to.

Traditional cable networks are well suited for their original job: transmitting television images. Cable networks have very high capacities for information; a cable subscriber's home is generally receiving every channel at once while the set-top box serves only as a filter to display one at a time.

But traditional cable networks are built to transmit information in only one direction: from the cable company toward the user, and not the reverse.

That poses a problem for telephones, which provide a two-way service. Traditional telephone networks are in some way the obverse of cable systems: They easily transmit communications in two directions, but they do not have to carry much information at the same time.

"Think of the copper telephone wire as a very thin but very intelligent pipe," said Sender Cohen, a data communications analyst at Lehman Brothers, "and the cable wire as very fat but very dumb."

So AT&T's first challenge is to make all of the cable systems it has agreed to acquire in some ways more like two-way telephone systems. That project, which requires the deployment of new equipment into cable hubs across the country, has already cost the cable industry billions of dollars, and in Mediaone, AT&T is set to acquire a cable operator with one of the most advanced networks in the industry, but one that still requires significant upgrades. AT&T has also struck partnerships with the Comcast Corp. and Time Warner Inc., two big cable operators, to offer telephone service using those companies' systems.

But even once a cable system has been adapted to send and receive data, voice and television signals, it is still not ready for the digital future. To offer high-speed Internet service, huge investments must be made in high-speed Internet switches that can route millions, even billions of bits of digital information every second. Even more daunting is the prospect of offering telephone service.

Every house that intends to switch from conventional to cable-based phone service must be visited by a trained technician to install an electronic box outside the house to connect the home's inside telephone wiring to the external cable wire. Big telephone switches the size of a van must be purchased and configured, almost by hand, to link with the cable network.

Then there are a myriad of seemingly mundane concerns that could be vital to consumers. For instance, allowing people to keep the same phone number when they switch to cable-based phone service is quite complicated.

Another challenge is that traditional cable systems generally use public power. When there is an electrical blackout, the cable television also fails, but the phone almost always works. That is because telephone wires carry power as well as communications, allowing most phones to work without being plugged in. To match the reliability of standard phone service using cable systems requires big investments of money and ingenuity in complicated power generation systems.

All of these technical challenges are just to offer phone service using conventional telephone technology, known as circuit switching, albeit over cable wires. To offer phone service over cable lines using Internet technology is even more complicated. In fact, the technology to do so reliably does not even exist yet. Armstrong said that he does not anticipate using Internet phone systems until 2001.

The upshot for AT&T is that all of the billions of dollars and millions of hours it will spend on the obvious technical challenges may pale beside the time, effort and money it will spend revamping its organization and deploying the anonymous "back office" computer systems that are the backbone of any modern business.

Thousands of trucks may have to be purchased, thousands of technicians hired and trained. Customer service representatives who now work for cable companies will have to be trained to answer questions about telephone and Internet service. One challenge in finding qualified dispatchers, supervisors and other sorts of support personnel is that almost no company in the world has tried to sell integrated bundles of telephone, television and Internet services to residential consumers who might not know a modem from a remote control. (RCN Corp. is one company that has tried.)

Moreover, internal billing and customer information systems for AT&T will have to be added and revamped to support a broad range of services for millions of customers who expect to be billed promptly and accurately.

"These companies will win or lose on the billing battlefield," said Howard Anderson, managing director of the Yankee Group, a technology consulting firm in Boston. "If your bill is wrong or it doesn't go through, you're going to have a fit. Billing is the Achilles heel of this whole thing."

Referring to AT&T's main business of selling long-distance service, a relatively simple product, he added: "It's one thing to send you a bill for long-distance minutes. It's just the time of day, the price and whatever stupid plan you bought. But when you start talking about different television packages and different speeds of data, you're talking about many more variables. And soon those bills will have to be delivered electronically."

AT&T is well aware of the incredible complexity of its ambitions. It has enlisted Lockheed Martin, the big military contractor with experience in managing complex projects, to help keep track of all of the undertaking's moving parts.

"The biggest issue where management attention has to be applied is to integrate all these things and cause them to come together in a way that's coherent," said John C. Petrillo, AT&T's executive vice president for strategy.

The stakes are especially high for AT&T because its reputation with consumers will be on the line. When a telephone cuts off for no apparent reason, when an Internet connection drops in the middle of a big download, when the television shuts off during the big game, customers rarely forget.