AT&T READY TO RUN, NOWHERE TO HIDE

SITTING IN HIS OFFICE on a frozen morning at the end of a very long winter, AT&T CEO Robert Allen bears the oppressed demeanor of a character in a Kafka story--perhaps because the press and the body politic have lately been treating him as if he'd turned into a giant insect. The worst blow in the recent public pasting he has endured landed when Newsweek branded him a "corporate killer" for cutting 40,000 jobs while collecting a huge bonus. Now the company's new annual report reveals just how much he lost on the takeover and coming disgorgement of computer maker NCR, which, despite AT&T's efforts to digest it, will have passed through Ma Bell like a large stone through a goose. In response, he is struggling to mount a PR counteroffensive, with full-page ads appealing to companies to hire the well-trained people he just laid off. "I've been better," Allen says in a voice even softer than his usual pianissimo.

The pressure comes at a pivotal moment for both Allen and the 120-year-old corporation that he runs, as AT&T prepares to split itself in three. The biggest piece--the one that will continue to wear the AT&T nameplate--is mainly a long-distance company today. And the job it faces is daunting: nothing less than redefining what a communications company does. As president and chief operating officer Alex Mandl says, with unusual understatement: "Therein lies the excitement of the new AT&T."

Therein also lies the risk. While the company's ambitious plan is simple enough in concept, it will be devilishly hard to pull off. After the trivestiture is completed early next year, the long-distance business will still be AT&T's crown jewel, and the company will seek to keep its setting polished and sound. But lest its diamond turn into a cubic zirconium, AT&T must quickly nurture a host of new, fast-growing businesses to help it meet three critical goals: generating cash, locking in long-distance customers, and restoring a direct link into people's houses--a link that was lost when the Baby Bells became the owner of the telephone line into the home in 1984.

So starting this year, AT&T will reenter the local phone business in all 50 states. It will push to finish building its wireless network, the most extensive in the U.S. It will hook up computer users to the Internet. And it will market digital satellite TV. It will sell these services in bundles, giving a greater discount the more a customer buys, and will rely on the power of its brand name to lock in long-distance users before newcomers poach them.

The extent to which AT&T is counting on these new services is startling. Mandl says that a decade from now, businesses other than long distance will account for over 50% of total revenues. That means services from which AT&T earned not a cent 18 months ago would account for over 80% of future revenue growth. Mandl also wants AT&T to achieve a double-digit growth rate over this ten-year period, a far faster pace than the old AT&T has managed. In all, sales for the new AT&T would swell from $51 billion last year to around $130 billion in 2005. Telecom analyst William Deatherage of Bear Stearns says, "That's clearly what they have to do, and it is attainable. It's all in the execution."

The metamorphosis of AT&T became urgent with the passage of the Telecommunications Act of 1996, which will break down the barriers between telecom's separate fiefdoms. When this happens, the industry will unbundle in the same way the computer business did when IBM's hegemony was broken in the early 1980s, and unleash, among other things, lower margins, higher volumes, and stiffer competition. Compared with what's ahead, today's long-distance wars will seem like a game of croquet on a manicured lawn.

To thrive, the new AT&T will have to perform a lot better than the old AT&T. The company's latest annual report offers, for the first time, separate results for what will soon be three new companies--NCR, the computer company; Lucent Technologies, the telephone equipment manufacturer; and the new AT&T, which will include the cellular-phone and credit-card businesses in addition to long distance. That breakout underlines the extent of the company's dependence on long-distance income. Over the past three years, NCR toted up losses of $2.8 billion; the equipment company delivered only $630 million in operating income, including $2.8 billion in restructuring and other charges last year. By comparison, the communications company racked up $19.7 billion in operating profits.

Think of it this way: The new AT&T is putting itself on the line. It will have neither the security of being leader of a cozy long-distance oligopoly nor the excuse of those weaker hardware units to blame for slow growth or a slumping share price. As general counsel John Zeglis puts it, "There's no place to hide, no cash cow to cover current mistakes and wait for a long-term payoff. We have to do it right and do it fast. It's a jolt, the back against the wall."

Zeglis is one of the four men of a stripped-down chairman's office that will steer the new AT&T. He and Allen are old Bellheads; the other two members are Mandl, the suave and forceful chief operating officer and a former head of shipping giant Sea-Land Service, and chief financial officer Richard Miller, a veteran of corporate restructurings at Penn Central, RCA, and Wang. This quadrumvirate supersedes a 15-person executive committee that often took weeks or months to make decisions that could have been settled in hours.

That structure clearly had to change. Indeed, if the company's historical edge was size--the unmatched scope and scale of its long-distance network--speed may hold the key to its future. The new AT&T can't afford prolonged deliberations as it tries to respond quickly to varied demands from more knowledgeable customers, to trends in technology, to curve balls from regulators, and above all, to the gambits of its many new competitors.

Chief among them are the Baby Bells, who have one huge edge: They own the wires that snake into every household, making it technically easier for them to offer a wider range of services more quickly. AT&T will try to keep them out of its water for as long as possible by drawing them into the still vast Sargasso Sea of regulatory Washington. The new telecom act, which lays out a checklist that the Bells must satisfy before they can enter the long-distance business, is a monument to ambiguity. Expect to find AT&T's legal department splitting more hairs than Sweeney Todd.

Zeglis will wield the legal scalpel. Boyish and earnest, Zeglis composes palindromes--phrases that read the same backwards or forwards--in his spare time ("Sit on a potato pan, Otis," was one he shared in a recent interview. "Star comedy by Democrats," he added, as if to explain). While he freshens the air around AT&T, his effect on the Baby Bells may be more purgative than tonic. Zeglis will likely challenge them every time they file for permission to carry long-distance service in a new state. He will also defend AT&T against Baby Bell counterattacks. In the latest of many lawsuits, Bell Atlantic, his most persistent gadfly, is seeking $3.5 billion in damages, alleging that AT&T refused to make phone switches that can plug-and-play with ones made by other manufacturers. Palindromes fail Zeglis when he is asked about Bell Atlantic. He says: "Almost no words of profanity work both backwards and forwards."

His animus will likely swell. The fate of the new AT&T hinges on how successfully it reenters the $90-billion-a-year local phone business. Bob Allen's audacious aim is to claim a third of that market in five to ten years--thus accounting for $30 billion of the $80 billion in new revenue the company wants by 2005. While the Bells maintain that state regulators keep prices so low that they make no profit on basic local service, they aren't about to give up the business without a fight. By owning the local customer, the Baby Bells and other local providers like GTE tap into two lucrative money streams: toll charges for middle-distance calls, as between a city and its suburbs, which bring in around $15 billion a year; and access fees paid by AT&T, MCI, and Sprint for delivering long-distance calls to local customers, which raises another $32 billion.

The appeal to AT&T is obvious. Signing up local customers means avoiding access fees, which siphon off as much as 45% of the money it collects for a given long-distance call. What is more, the health of AT&T's biggest business may depend on the local foray. Says Joseph Nacchio, who runs the residential long-distance business and will carry AT&T's flag into the local market: "Two-thirds of the people we survey say they will buy local and long-distance service from the same provider." In other words, AT&T feels that to keep its long-distance customers, it has to sign them up for local service too.

AT&T's biggest problem is that it can't afford to replicate the Baby Bells' plant--phone lines, switching centers, and the like--in most local markets it wants to enter. No one can. So it has to rent space on somebody else's network. But which somebody? Cable TV systems won't be able to carry lots of phone calls for a while. Bypass companies like Teleport Communications Group and MFS Telecom are beefing up local fiber-optic lines but for the most part don't stretch their tendrils into residential neighborhoods. That leaves the Baby Bells as wholesaler of first resort.

And that is a headache. The Bells are in no hurry to offer AT&T the wholesale discounts it is seeking--25% or more. A potential deal to rent hundreds of thousands of local lines from Ameritech, the Baby Bell headquartered in Chicago, has meandered at a stately pace since the middle of last year. AT&T accuses Ameritech of foot-dragging. Ameritech fires back that AT&T is to blame. In a recent interview, Ameritech CEO Richard Notebaert's face grows ruddy with chagrin. "Nothing will happen until AT&T gets its act together," he says. "They must be in disarray. We're dealing with their third negotiating team." He throws his hands up in disgust. "I'm sure they'll straighten it out, but they've got a problem."

"When you do something the first time, you don't do it perfectly," Mandl admits. "It may not have been the effort we would have liked. But I would say to Mr. Notebaert that Ameritech wasn't perfect either."

AT&T can only hope that the company's expansion of its wireless business goes more smoothly. Mandl says the company looks to build revenues from $3 billion last year to $20 billion by 2005.

This is one arena in which AT&T's mass should still confer an advantage. The number of cellular providers could triple in the next few years, with the newcomers competing madly to woo customers from the incumbents. As a result, churn, or the rate that customers switch providers, will soar. Profit margins will head in the opposite direction, since cellular companies bribe new customers by giving away free phones, and since new customers tend to spend less than old ones, according to the Yankee Group, a Boston research firm. If AT&T can entice enough customers to buy both cellular and long-distance service, it could tolerate a lower margin than most of its wireless rivals.

Freedom from the other arms of AT&T will at least make the wireless unit more nimble. Steve Hooper, AT&T's wireless chief, is nearly as happy about the spinoff of Lucent as the people who work at the equipment maker (see box). Shortly after AT&T assumed ownership of McCaw in 1994, the wireless unit decided against using AT&T gear to build a cellular system in India. Hooper says: "We had to spend a lot of time with Alex. I'm sure he had to take the heat. It put everyone in an uncomfortable environment back in Basking Ridge." And wasted a lot of time.

The experience of integrating McCaw with AT&T may have taught Mandl and others a thing or two about the need to act with alacrity. Two recent moves into new businesses--the company's latest Internet offering and its purchase of a slice of DirecTv, a satellite TV service--caught the industry by surprise.

AT&T's Internet play comes in the wake of some false starts with online services that disappeared in about the time it takes to boot up a computer. The new service, though, is a deft bid that may transform the online industry while helping to protect AT&T's long-distance revenue. AT&T will give its long-distance customers five hours of free Internet access every month for a year, or $5 off the monthly rate for unlimited use. Daniel Briere, president of TeleChoice, a Verona, New Jersey, consulting firm, believes AT&T will turn people into net surfers by the millions. "AT&T will suck them all in," he says. "It's going to do what no one else in the industry could." But the transformation won't happen overnight; swamped by demand following its February 27 announcement, the company had to remind customers that it was going to phase in the service over several months.

AT&T also plans to use the Internet to offer new services to its business customers, especially those with 800 numbers, and make it easier for them to keep shop on the World Wide Web. AT&T will help virtual shopkeepers process orders and will guarantee the safety of purchases made by users of its Universal Card. Internet commerce is tiny, for now: Last year, according to Jupiter Communications, a New York research firm, sales on the Web totaled only $132 million. But AT&T likes the idea of getting a piece of every transaction on a more consumer-friendly Web. As Mandl sees it, online services will bring in 10% of AT&T's revenues by 2005, or about $13 billion a year.

If electronic commerce really takes off, it could help AT&T achieve its most ambitious goal: transcending the traditional telephone business. Says John Petrillo, the head of corporate strategy: "People who see this as a win-lose game between us and the Baby Bells are missing the point." Petrillo sees the telecommunications pie growing bigger and bigger, at the expense of other industries, like, say, transportation. He believes this shift portends a day when goods such as books are shipped around the country electronically and published in many places, rather than printed centrally, shipped, and warehoused. "We will rip the economic surplus out of transport," he says. That's a pretty lofty, not to mention long-term, mission.

AT&T's DirecTv deal, by contrast, smacks much more of the here and now. It rounds out AT&T's portfolio with a touch of Hollywood. The highly popular service, owned by GM's Hughes Electronics division, beams cable TV and sports programming to owners of 18-inch dishes. AT&T bought 2.5% of the company for $138 million and has an option to buy more later. It will market the service through its long-distance division and offer discounts to subscribers who take both.

AT&T executives took only three days to approve the deal. CFO Miller says, "In the old structure it would have taken weeks. We would have had meetings with people from every division asking what it meant for their business."

Despite its newfound speediness, AT&T still has to prove it can cultivate new businesses. Today Alex Mandl may swear that the company's latest Internet vision isn't just another mirage, but a year ago he was lauding AT&T Interchange, a proprietary online service that's now history. So is Network Notes, a work-sharing software service developed in partnership with Lotus. A host of other investments in small outfits like EO and General Magic never really panned out either. All in all, a spotty record for a company setting out to find $65 billion in revenues from businesses it is just entering.

Bell Atlantic Chairman Ray Smith, a veteran of the original breakup, thinks his old employer and new competitor is still at sea. "People inside AT&T are really confused now," he says. "We were not confused in 1983 and 1984. We weren't right. But we were not confused."

To put confusion behind and matters of substance ahead, CFO Miller intends to be done with the restructuring as soon as possible. Miller took Penn Central through bankruptcy, reorganized Wang, and merged GE's consumer electronics business with RCA's. "I've been involved with some restructurings that have gone on for a good deal of time," he says. "Employee morale gets worse the longer they take. Customers want to see finished plans."

But will AT&T's newest split-up really be different from the last one, which dragged on for years? When the recent layoffs were announced, much was made of the huge size of AT&T's central staff, which included between 800 and 900 public relations people. So how many will be left in the three new companies after the cuts? Between 700 and 800.

The list of unanswered questions extends to AT&T's executive wing. Who will be running this new show? After Allen announced the trivestiture last September, he said he would spend the next year managing the split while Alex Mandl ran the future AT&T. Many onlookers assumed that after guiding the company into the promised land of independence, Allen would retire to the links, leaving AT&T to Mandl, the heir apparent ever since he took the reins of the communications division in 1993.

Mandl, 52, is a man of supreme self-confidence and the sort of flashy, urbane European charm that would make for a good James Bond villain. As it happens, he bears a passing resemblance to the actor Klaus Maria Brandauer, who did play a Bond villain and who, like Mandl, was born in Austria during World War II. One imagines that Mandl must be champing at the bit.

But his ascendance won't happen quite as soon as people imagined. Allen now asserts that he will retire only when he turns 65 in the year 2000. "I have no plans to do anything else," he says. "We're in such a critical stage. I can be in a good position to make key strategic decisions."

That would seem to mean less authority for Mandl. As COO of the new AT&T, he will be responsible for pretty much the same range of businesses that he is now. Yet the man he is reporting to would suddenly have a narrower focus--the same span of control as Mandl, rather than the overarching purview of the soon-to-be-sundered empire.

In fact, the arrangement may work out just fine. In his operating capacity, Mandl at least sounds as if he's running the company. He has his own ten-year master plan, which formed the basis for much of his discussion with FORTUNE. (The plan is also codified in a document someone leaked to New York magazine for use in a story Mandl calls "an absolute piece of shit.") And he sounds as if he appreciates Allen's presence in Basking Ridge. "I've got my plate full," Mandl says. "I welcome all the help I can get."

Allen's laid-back manner and lack of hubris could make him amenable to Mandl's activist role. Their division of labor stands out in sharp relief in the way they talk about AT&T's future, with Mandl describing the fine brushwork of strategy, Allen speaking in broad strokes. In a way, Mandl will be prime minister to Allen's monarch. Just like in Europe.

AS HEAD OF STATE, Allen is the one who bears the brunt of the invective aimed at AT&T. The day after the "corporate killer" article appeared on the cover of Newsweek, he was visibly distraught. And he has recently been lambasted, in the pages of this magazine and the Wall Street Journal, for the dismal failure of NCR. Yet his decision to stick around till retirement fits with his temperament. His ties to the company are long and deep. His one display of alarm in an interview is over an erroneous report that he had sold most of his AT&T shares. He did sell some recently, but only enough to pay taxes on stock options he had to exercise before they expired. "Other than that, the only AT&T shares I have ever sold were to buy my first house in 1961," he says. "I sold 20 shares. I was so poor that my former girlfriend's brother sold them at no commission. He knew I didn't have a pot to piss in."

He falters over the penultimate word of the last sentence. It's the sort of remark you'd expect to hear from Mandl. Perhaps his younger colleague has helped Allen develop an edge. Perhaps Allen will temper Mandl's brash nature. If the two find the right chemistry, they may help AT&T weather the years ahead better than it has the past few.