(FORTUNE Magazine) – The most happy fella in U.S. business has to be a German-bred accountant who is doing as much as anyone to redefine the future of the global computer industry. Eckhard Pfeiffer, president of Compaq Computer, is just coming off the kind of record year other CEOs would kill for. Since 1992, his first full year as chief executive, sales have more than doubled, to $10.9 billion, and profits more than quadrupled, to $867 million. Nobody complained when the board recently raised his salary, bonus, and other annual compensation from $2.5 million to $5.05 million.

Pfeiffer's Compaq has become a classic case of corporate comeback. And the company's next moves may well shake up--and shake out--the entire industry. Ahead lie still deeper price cuts, lower profit margins, and many more new products.

What has unfolded so far is a drama in three acts:

Act I, October 1991: Compaq, which had surged as a producer of high-quality, high-priced personal computers for business users, hits its first crisis. Pressured to cut costs, large companies that are Compaq's customers no longer will pay plump premiums for its machines; instead they turn to cheaper, almost-as-good clones. Sales and profits tank, and the price of the stock drops by two-thirds. Out goes CEO Rod Canion, and in comes Pfeiffer, the driven, disciplined, and hugely successful chief of Compaq's international business. He cuts the worldwide work force from about 12,000 to 10,000, expands dealers from 3,500 to 38,000, and presses suppliers for better terms. He expands into the PC markets for home, education, and small-business uses, puts in 24-hour mail-order and customer-service phone lines, and generally brings focus and fresh efficiencies. Critics laugh and call him arrogant when he sets a stretch goal: Compaq will become the biggest producer of PCs--in 1997.

Act II, January 1995: In front of 16,000 foot-stomping employees and spouses who fill the Summit arena in Houston, Pfeiffer-wearing lederhosen and sounding like a cross between Elmer Gantry and Henry Kissinger--jokes, exhorts, and then trumpets, "We are No. 1! We made it in 1994! We've replaced IBM as the world's top PC vendor." Indeed, Compaq has shipped 4.8 million PCs in the year, far ahead of IBM's and Apple's four million each. In just two years, Compaq's share of the worldwide market rose from 4.8% to 10%, while IBM's fell from 10% to 8.3% and Apple's slipped from 8.5% to 8.3%. Now Pfeiffer sets another stretch goal: Triple the company's sales, to $30 billion, by 2000. The crowd goes bananas as the usually buttoned-up boss, calling on everyone to scale new heights, snaps himself onto a nylon cable and is lifted off the stage, straight up 150 feet, disappearing into the ceiling.

Act III, March 1995: Appearing before a couple hundred analysts and journalists in Manhattan, Pfeiffer--whose company used to introduce perhaps eight new models a year--unfurls 118 new PCs. Most are based on Intel's high-performance Pentium microprocessor. He also announces price cuts of 5% to 23%. Analysts call the moves aggressive. They warn that profit margins may shrink in the first half as Compaq clears out its inventory of older, 486-chip machines, but may well rise in the second half. Not surprisingly, Apple, Hewlett-Packard, and IBM announce price cuts of approximately the same size. All this is part of what Pfeiffer calls building creative destruction into your business. As he writes in this month's Euromoney magazine: "Nothing is harder than casting aside the thinking, strategies, and biases that propelled a business to its current success. Companies need to learn how to unlearn, to slough off yesterday's wisdom."

It is a risky strategy, but Pfeiffer, 53, has profited from taking big risks before. In late 1992 he saw opportunity in competitors' weaknesses. "IBM's PC division was in total disarray," he told me, "and Apple seemed to drift off to the Newton just before John Sculley left." The Newton, an electronic notepadlike device to keep track of appointments and phone numbers, among other things, bombed in stores. With rivals distracted, Compaq rolled out the Presario line of computers, which prom-ised the company's well-regarded performance for only $1,400. Says Pfeiffer: "Customers just went wild."

Late last year the company took another big risk by building an inventory of $2.3 billion worth of PCs, two or more times as many as any competitor. When Christmas sales took off as Compaq had expected, other companies were caught short.

Pfeiffer's aim now is to persuade business users to replace huge mainframes and medium-size minicomputers with much less expensive networks of Compaq PCs. Outfits as diverse as Brigham and Women's Hospital in Boston and China's Ministry of Railways are already doing just that. One Pentium PC, Pfeiffer notes, has 1,000 times the number-crunching power per dollar of a mid-1980s mainframe. And a $50,000 network of four PCs now can accomplish just about everything a $250,000 minicomputer can.

For all the potential of the business market, Pfeiffer thinks the consumer segment will grow twice as fast. He looks at the hole in the doughnut and aims to fill it: In the U.S. alone, 60 million families don't have PCs. He vows to capture more of them as prices fall and the PC beats out the TV as the all-in-one device for work, education, and entertainment. "Today's multiple-phone households," he says, "will become tomorrow's multiple-PC households." Compaq will concentrate on the market's "sweet spot" of PCs selling for $1,500 to $2,500. (The upgraded Presario now retails for about $1,700.)

Pfeiffer believes in manufacturing close to his market. All Compaq's computers and related gear are designed in its Houston labs, and 90% of components for the products it sells in the U.S. are made here. Lately, Compaq has opened plants in Brazil to serve Latin America, and in Shenzhen, China, to serve much of Asia. "After the year 2000," he says, "China may grow into the biggest PC market in the world, potentially outpacing the U.S."

At all of its plants, Compaq plans to replace the old production line system with one based on four-person "cells." Already in place in Houston and Scotland, it works like this: One person gathers and lays out components, two assemble them, and the fourth checks the finished machine. Cells enable the company to put together each device according to a customer's specifications and to build them on the basis of actual orders instead of forecasts. This should hold down inventory and speed delivery. The goal is to ship your machine within two days of your phone order, or five days if you order through a dealer. Accomplishing that should help Compaq achieve another Pfeiffer goal: "To be recognized as clearly No. 1 in customer satisfaction--a combination of delivery, service, support, product, pricing--you name it." Surveys now put Compaq only among the top firms in customer satisfaction.

You can sum up Compaq's strategy this way: By cutting costs, you can cut prices. By reducing prices, you can build market share. As share rises, you can negotiate lower prices from suppliers--and cut your own prices some more. It's a virtuous circle. As Pfeiffer told me, "Never underestimate a hungry, purposeful, committed company."