Information Processing

Communications

AT&T Takes its First Giant Steps into Commercial Computers

Business Week

April 9, 1984

Question: What company invented the transistor, currently produces one of the world's most sophisticated computer memory chips, and employs 10,000 software engineers, yet has never sold even one computer to another company? Answer: American Telephone & Telegraph Co. For years, the communications giant could not market computers outside its own company, thanks to a consent decree that it signed in 1956 with the Justice Dept.

Now all that has changed. By agreeing to spin off its 22 regulated, local telephone operating companies on Jan. 1, AT&T was free to enter the competitive world of data processing. And on Mar. 27, AT&T took advantage of its newfound independence to introduce its first line of commercial computers. "We intend to play in [the computer] market, and play in it well," declares James E. Olson, vice-chairman of AT&T and head of AT&T Technologies, which makes and markets the new machines.

Hardly anyone expects AT&T to rival the annual computer sales of its chief adversary, International Business Machines Corp., anytime soon. "AT&T has a long learning curve" ahead of it, predicts Robert C. Downs, president of Enmasse Computer Corp. Over the long run, however, AT&T will be the only company large enough to mount a serious challenge to IBM's dominance of the market, industry observers say.

'A BIG CLUB'

AT&T already has sufficient financial and technological clout to give major headaches to such leading computer makers as Data General, Digital Equipment, Hewlett-Packard, Tandem, and Wang. "They swing a big club, and anyone in the business not watching what AT&T's doing is fooling themselves," cautions a senior executive at one large computer maker. AT&T's success in the computer business is vital to its long-term future. While it expects that its basic communications businesses will keep growing, AT&T is convinced that the markets for "data processing and the office of the future will grow far faster than our traditional voice-data communications markets," says Olson. "To meet our long-term strategic goals and attain the kind of growth we want, the office information systems industry is crucial to us."

Divestiture left AT&T with two basic businesses: long-distance services run by AT&T Communications and equipment manufacturing run by AT&T Technologies, which encompasses what was Bell Telephone Laboratories Inc. and Western Electric Co. Both new units already are having their share of problems.

Long-distance services, which were forecast to generate nearly $35 billion of AT&T's projected $56 billion in 1984 revenues, was supposed to be the company's cash cow. But unless the Federal Communications Commission speeds up its restructuring of the nation's long-distance rates, Chairman Charles L. Brown warns that AT&T "has no chance whatsoever" of producing the $2.1 billion in net income this year that the company projected in its prospectus last November. As a result, when AT&T in late March announced its first-quarter dividend of 30 cents, the company cautioned that it may not be able to pay the same dividend for the remainder of the year.

AT&T Technologies is not doing much better. Divestiture has left it bereft of the huge captive market of local Bell operating companies for its network switching equipment. At the same time, it is facing severe competition from companies such as ITT, Rolm, Japan's NEC, and Canada's Northern Telecom in its efforts to sell communications equipment to big corporations.

DUAL APPROACH

To enter the office information systems business, AT&T is following a two-pronged strategy. The first drive centers on Unix, a powerful operating system program that was originally developed by Bell Labs engineers. Unlike other operating systems -- the housekeeping programs that control basic computer functions -- Unix makes it possible for the applications programs that perform specific tasks to be transferred easily from one brand of system to another. Unix can also readily handle several tasks at once, an important feature in an office where many people use a computer at the same time.

To create a market for its machines, AT&T has been actively seeking software developers to write applications programs that run on Unix. "It's the most aggressive third-party software [recruiting] movement I've ever seen," say John R. Rowley, president of Digital Research Inc., which is helping AT&T build a Unix software library.

"We didn't want to enter the computer business as a me-too with an IBM-compatible [computer]," asserts Jack M. Scanlon, the vice-president who heads AT&T Technologies' Computer Systems Div. in Lisle, Ill. So the company first "set out to build a new market around [the Unix] operating system," he says. AT&T appears to be succeeding: The market for Unix applications software will explode from $260 million last year to $2 billion by 1987, estimates Jean L. Yates, president of Yates Ventures, a Los Altos (Calif.) market researcher.

The second prong of the AT&T strategy is its new series of computers. One of the broadest product lines on the market, AT&T's computers range from the 3B2 microcomputer, which will be priced for less than $10,000, to the 3B20 superminicomputers, which will sell for $300,000 and up (table). Desktop personal computers were not in the initial product offering, but they are expected shortly. The machines appear to be comparably priced and about as powerful as those of the competition. What AT&T hopes will set its machines apart is their high reliability and that they all employ Unix. "No one has a line from superminis to desktops with one operating system, all compatible," says Scanlon.

WHERE AT&T'S NEW MODELS FIT IN

DATA: AT&T, BW

MAKING MONEY

Unlike most computer makers introducing new products, AT&T is already in full production at its factory in Oklahoma City. Nearly 1,500 of the minicomputers are installed around AT&T and the former Bell telephone operating companies. Scanlon maintains that the company is already making money on its computers, although some industry experts question if the former Western Electric, with its unionized workers, can compete on a cost basis with other computer makers. "We've been serving our internal needs for several years, and we are doing it on a profitable basis," answers Scanlon. "The commercial business is just incremental capacity."

The new AT&T computers are aimed squarely at two of the fastest-growing information processing markets. More than $5 billion worth of superminicomputers were sold last year, and shipments of these machines should nearly double by 1986, predicts International Data Corp., a Massachusetts market researcher. And AT&T wants to use its microcomputers to penetrate the $2.8 billion office automation market, which the researcher estimates will grow to $6.2 billion annually in the next three years.

To move its products quickly into these markets, AT&T is using several distribution channels. The Computer Systems Div. will sell to so-called value-added remarketers, which buy computers and add applications software to customize the machines for specific technical applications. AT&T Information Systems will sell to those value-added remarketers that are business-oriented and to corporations and other end-users. And Italy's Olivetti, 25% of which is owned by AT&T, will distribute some of AT&T's computers in Europe.

Even before it ships its first computer to an outsider later this spring, AT&T has earned a high reputation among potential customers. "If you look at AT&T's technological track record for developing the Unix language and the proven reliability of their communications products, they're starting off with a lot of strengths," says Charles Carroll, vice-president at Days Inns of America Inc.

Even with that high regard, AT&T faces a hard battle. "Not many people are going to buy computers because they have the AT&T name on them," says Robert Capone, vice-president and director for systems and data processing at J. C. Penney Co. David L. Cotterill, executive vice-president for operations at Wachovia Bank & Trust Co. in Winston-Salem, N.C., agrees. "They are the new kid on the block. They have to prove to all of us that they can do it better."

One big reason why AT&T may find it difficult to convince prospective customers is that the industry still considers marketing to be the company's Achilles' heel. Although AT&T has striven for nearly a decade to train its sales force, horror stories about its gaffes still abound. One company, which does not want to be identified, says it tried numerous times in recent months to get AT&T to bid on a PBX for a large building, but no one at AT&T responded. Another customer claims that when the company was asked to bid on an order for several thousand data communications products last year, an AT&T sales representative answered: "We don't know how to price that -- we've never had anybody order that many." Even AT&T's Olson concedes that "we are not as happy with the success of the sales force as we had hoped."

TOUGH TRANSITION

Computer marketing may also be weak at AT&T, say some industry analysts. They claim the company's prices are not aggressive enough to win customers. "Why would a big user take the risk of going with AT&T, a new player, unless he is saving something?" asks Stephen P. Cohen, an analyst with Gartner Group Inc.

AT&T's marketing push must extend all the way to its product designers, who must do a better job of matching new computer products to the needs of prospective customers. "Although telecommunications is now much closer to computer technology, it is not the computer business," maintains Richard M. Moley, vice-president for marketing at Rolm Corp. It's a difficult transition to make." For example, AT&T has been designing and building its products for the 10-to-15-year product life cycles typical of telephone equipment, not the five-year cycles of the computer business.

The Bell Labs engineers who created and refined the Unix operating system also have little experience in office applications at large corporations. Instead, they have concentrated on more technically oriented tasks. "They haven't operated in the commercial setting," says Andrea J. Curtis, product requirements manager at Data General Corp., which offers Unix on its equipment. "They still aren't tuned in to customers." AT&T claims that it is solving most of these problems. One example of its growing market savvy is the kind of communications networks that it rolled out with its new computers. As a late-comer to the commercial computer business, AT&T acknowledges that it must adapt to equipment that is already in place, especially IBM equipment. So the company has designed its new 3B Net, a local-area network for office buildings, which will link multiple AT&T computers to each other as well as to other brands of computers.

An even more important network is PC Interface, which permits IBM Personal Computers (PCs) to be used as terminals for an AT&T system, making it easy for PC owners to retrieve files and run programs on the powerful AT&T systems. "We wanted to be able to go into an installed base and grow gracefully with it," explains Scanlon. Eventually, he says, AT&T will even link its products into IBM's mainframe computers.

USING OUTSIDERS

AT&T also seems to have learned that it cannot do everything itself, a policy that it usually followed in the days of its telephone monopoly. In addition to relying on the value-added remarketers and other outside companies to supply the applications software for its machines, AT&T is turning to outsiders for hardware. The company has already signed a deal with Convergent Technologies Inc., a Santa Clara (Calif.) microcomputer maker, to make desktop terminals for AT&T to sell. And Olivetti is expected to turn out office automation equipment for AT&T to sell in the U.S.

AT&T is working hard to give its sales force more of a computer slant. It has raised the number of people with data processing sales experience to almost 20% of the total force, reports Robert J. Casale, division president for marketing and sales. And he says that training is proceeding apace -- by July, some 2,000 of AT&T's 6,000 salespeople will have been taught about its new computers.

AT&T customers already are beginning to report a new attitude among AT&T salespeople. "They really have changed within the past year," says Cotterill of Wachovia Bank. "They seem to be a lot more professional." Mayford L. Roark, executive director of systems at Ford Motor Co., adds that AT&T "comes across as having a much more aggressive marketing organization than [the company] we used to know before."

One more thing that may help AT&T: A good many people in the industry are actively rooting for the company to make it in computers. "Somebody's got to find a way to compete with IBM," says Randy J. Goldfield, president of the Omni Group Ltd., a New York office automation consulting firm. Adds Neal Nelson, president of Neal Nelson & Associates, a small Chicago computer remarketer: "I'd like to have another big name [besides IBM] to offer a customer." With that kind of reception from the marketplace -- coupled with its $34 billion in assets -- AT&T could very well be the next big name in computers. 

GRAPHIC: Picture 1, THE NEW SERIES OF COMPUTERS IS ALREADY IN FULL PRODUCTION -- AND PROFITABLE, JIM ARGO; Picture 2, VICE-CHAIRMAN OLSON: A VOW TO PLAY IN THIS MARKET, "AND PLAY IN IT WELL," LAWRENCE BARNS; Picture 3, VICE-PRESIDENT SCANLON: NO INTEREST IN "ME-TOO" IBM-COMPATIBLE COMPUTERS, MARK FERRI

Copyright 1984 McGraw-Hill, Inc.