(FORTUNE Magazine) – If Lucent Technologies President Richard McGinn ever forgets how different the future will be for his company, he's apt to get a reminder at home. "My daughter wants to increase her access to the Internet so she can trade Beanie Babies," he says. "She's 7 years old."

Communications is changing, and Lucent's challenge is to figure out how to change with it. Created last year when AT&T, its doddering corporate parent, spun off telecommunications equipment and software operations, Lucent is enjoying the acclaim showered on the unexpectedly successful. New companies eager to get a piece of local and long-distance phone markets in the U.S., and telecom deregulation abroad, have created a hunger for Lucent gear, so the money rolls in and the stock flourishes. It's hard to remember that just a few years ago most of Lucent's businesses were losing money, and some doubted it could thrive away from Ma Bell's embrace.

Parlaying this early success into something permanent is the chief job of McGinn, 50, and his boss, CEO Henry Schacht, 62, whom AT&T Chairman Bob Allen cajoled from retirement to run Lucent. While Lucent's biggest business is the sale of software and equipment for telephone networks, the future lies in supplying networks that can carry data and images as well. Lucent must also cope with internal problems, like figuring out how to bring down costs and squeeze new products more quickly out of Bell Labs, which it inherited from AT&T. And while liberation from AT&T is cause for celebration, the process has not been hangover-free.

AT&T left its offspring with a lot of blubber, despite making deep reductions in Lucent's staff before cutting it loose. The cost of sales and administration in 1996 amounted to more than 25% of Lucent's $24 billion in revenues, way out of line with competitors like Northern Telecom, where the figure was 17%. Nor did AT&T give Lucent total freedom in picking its markets. The separation agreement forbids Lucent from competing in any of AT&T's businesses for five years. And it prevents Lucent from owning more than a small stake in a communications network--wireline, wireless, or data--anywhere in the world. If Lucent violates the agreement, it forfeits any right to use the AT&T brand in its marketing and loses some patents as well. Schacht says he has no desire to compete with the Baby Bells--his biggest customers--by getting into the telephone service business.

Despite its shackles, Lucent resembles an unblemished portrait that was hidden in the attic. The crest in its basic business is largely a happy outcome of government policy. The telecom law passed last year freed communications companies to invade each other's markets, and the Federal Communications Commission auctioned off new licenses for wireless communications networks, of which Lucent is a major supplier. "Business just exploded," says Schacht. "It was a huge, huge favorable event that we did not foresee. Nobody saw it." The frenzy to build up-to-date telephone networks will not abate until the new players stake out their positions and until McGinn and other indulgent parents install second phone lines so their kids can be online all the time.

Although Lucent is poised to meet the every need of people buying next-generation phone gear, it is less well positioned to supply the broadband networks that can carry voice, data, and video at very high speeds. These unified networks will do everything that telephone and computer networks do today, all over the same set of wires. Such a network could grow out of today's telephone systems. But many believe the Internet will evolve into an omnipotent cyber-network, able to carry data and images and voice everywhere. And the early builders of cyber-networks, such as Internet service provider Uunet, disdain Lucent as a maker of "big iron." They buy from outfits like Cisco Systems, which dominates the sale of Internet routers that send digital bits of information from place to place.

This could be bad news for Lucent. Although it is perfectly capable of developing gear for the cyber-networks, the question is whether it can forge relationships with the companies that will build the networks. "your customers pull you along," says James Moore, CEO of GeoPartners Research, a consulting firm in Cambridge, Massachusetts. "The real threat to Lucent is if its big customers, the phone companies, are slowest to adopt cyber-networks."

The desire of corporations to unify their own data and phone networks could also leave Lucent vulnerable. Lucent's traditional contacts--corporate telecom managers--increasingly report to information services chiefs, who are apt to be Cisco-centric. Since Lucent doesn't have much experience selling to that side of the house, it has recruited 200 salespeople from data shops like Cisco, Bay Networks, and Newbridge Networks. And it plans to open new sales channels, because while Lucent sells phone gear directly to customers, most data gear is sold through resellers.

Lucent is looking to its heritage for an edge. Says McGinn: "For us, the idea of trying to out-Cisco Cisco is a fool's errand." Instead he hopes to adapt Lucent's network-management software for use in notoriously crash-prone cyber-networks, allowing them to run all the time, faultlessly, like the phone system usually does. Lucent is trying to develop a new data switch that can rapidly pass along lots of information with high reliability. It is also experimenting with different ways of carrying the data between switches at very high speeds, including fiber-optic networks and systems built of a melange of fiber and coaxial cable. "It's clear the areas we need to be in," says Bill O'Shea, who runs Lucent's business systems division. "What's not always clear is which technologies are the right ones."

In trying to crack new markets, Lucent will rely heavily on a potent weapon: the patents it inherited from AT&T. Besides covering the technology that underlies telephone networks, these include seminal semiconductor patents, held in concert with Intel and Texas Instruments. Indeed, Lucent's chip division, whose revenues rose 23% last year, is its fastest-growing business and acts as a hedge against trouble elsewhere. Even though Lucent's market share in cellular phones is minuscule, its chips are in 60% of the digital cellular phones sold worldwide. (A favorite Lucent factoid: If a cellular phone had to be made with vacuum tubes instead of microchips, it would be the size of the Washington Monument.)

The creative caldron from which the patents bubble is Bell Labs in Murray Hill, New Jersey. Lucent executives also work out of Bell Labs headquarters, but the scientists are reasonably well isolated from the suits. A stranger who leaves the carpeted executive wing at the front of the sprawling building is soon lost in a warren of narrow corridors with open ductwork and linoleum floors. Giant nitrogen canisters line the walls to cool equipment, and yellow handles dangle overhead for emergency showers if chemicals splatter. Jeans-clad men and women of all ages and provenance dart about, giving the place the feel of an international supercampus.

Until the 1984 divestiture, AT&T was not allowed to market a lot of the work of its Nobel laureates in Bell Labs. But even when it could, company marketers often just scratched their heads. Among the inventions that AT&T didn't know how to sell: solid-state computer memory, the Unix operating system, and Windows--yes, it was Bell Labs, not Microsoft, that first figured out how to make two windows operate simultaneously on a computer screen. Bell Labs also invented the cellular phone, only to see the business abandoned by AT&T, which later bought back into it at great expense.

Bell Labs' occasionally languid style hasn't helped either. "They have the top scientists in just about every field you can think of," says Michael Adamian, who recently retired as a manager for the network systems division. "But everything they design and develop is late and way too expensive." One example: chips for the newest generation of computer modems. Lucent's chip shipped in early April. But by then U.S. Robotics was already in the market with a chip of its own design.

Schacht and McGinn are well aware of the lead-footed legacy. They each spend half a day every fortnight in the labs to inspire and to identify technologies that can spawn new ventures. Under their prodding, scientists are filing an average of three new patents a day, up about 50% since the spinoff. "I feel we're more closely coupled to the people who run the company now," says Howard Katz, a Bell Labs materials scientist. "I never had that sense with the old AT&T."

Not long ago a Bell Labs scientist newly arrived from Taiwan was struggling to design a chip with the tiniest features ever. He solved the problem as he stood outside in his first snowfall, watching how the snowflakes layered on his car. Lucent will shine if it can capitalize on such flashes of inspiration.