(FORTUNE Magazine) – Short of an earthquake in Scotts Valley, Seagate Technology's headquarters town in California, near Silicon Valley, it would have been hard to imagine anything else going wrong when the computer hard-drive maker wound up its June 30, 1998, fiscal year. Sales were down almost 25% from the prior year. Operating income, previously more than $1 billion, had evaporated. CEO Al Shugart, a founder of the company and one of the inventors of the hard-disk drive, was at odds with the board and would, in fact, be fired within weeks.

Almost nothing was going right. After years in which Seagate had enjoyed comfortable technical leadership in drives for high-performance computers, competitors including IBM and Quantum were catching up. The volume part of the business, hard disks for personal computers, was overloaded with production capacity and glutted with inventory. The largest customers, the computer makers, were demanding not only more storage per dollar--nothing new about that--but also more service, including caches of completed drives near their assembly operations. But Seagate seemed unable to respond expeditiously.

Something had to change--fast. Even before Shugart left, new managers had arrived from outside the Seagate walls. Five of today's top nine executives have joined the company since 1996. In early 1998, when Ernst & Young consultants recommended changes in Seagate's product-development process, it was the newcomers who pushed through their adoption. Not long afterward, the new managers nodded enthusiastically when Craig Nichols, an E&Y partner, suggested it was time to go further and "transform the whole operations side of their business."

Still, Nichols' idea of how to shake things up seemed a bit radical, especially for a hierarchically minded company like Seagate: Avoid project teams and months-long investigations, and don't bother with reports, committees, or levels of approval. Instead, select people from all levels of the company, add some consultants and even customers, put them all together, stir well, and come up with a completely new concept of how to organize and run the business--all in just three intensive days.

What Nichols was proposing was a DesignShop, a concept (and trademarked name) developed and worked out by Matt Taylor, an architect, and his wife, Gail, an educator. Their idea is to push--some would say shock--a group to leave behind the conventional and instead collaborate creatively on a new product design, a new process, the solution to a business problem, or even a completely new business strategy. In Seagate's case, the three-day session laid the groundwork for a return to financial health. Though further restructuring is needed and many bumps lie ahead, the company is running with dramatically improved efficiency. Elsewhere, companies as diverse as Compaq, Johnson & Johnson, and Kellogg, as well as a giant U.S. Air Force development center, have used the DesignShop process to map out changes in a hurry.

Essential to the process is the environment in which the exchange of ideas happens. Participants arrive to find neither conference tables, auditorium-style seating, nor--possibly most disturbing--packets of slide duplicates and fixed agendas. Instead, DesignShops are equipped with Taylor-designed furniture that can be shifted about and movable walls that can be rolled around to accommodate what will prove to be a frequent forming and reforming of teams. Most of the walls are huge whiteboards or places to post clippings and illustrations.

Participants are strongly encouraged to store away anything they've brought along, particularly cell phones, beepers, and Palms. To stimulate thinking, the centers have collections of books on a sweeping range of topics. There's also an eclectic assemblage of puzzles, and even a stuffed Kermit the Frog and other plush toys. Stuffed animals, as well as soft footballs, are for the tactile types who think best while using their hands; fiddling with them takes the place of paper-clip bending or pencil tapping--to the relief of other participants.

The Taylors began running DesignShops in the '80s, mostly at client companies or at temporary locations. Since then they have persuaded some clients, including Hewlett-Packard, Johnson & Johnson, and Detroit Edison, to set up their own permanent facilities. Among those with such venues is the Arnold Engineering Development Center (AEDC), the Air Force's advanced equipment-testing complex in Tullahoma, Tenn., which in the early '90s escaped downsizing and closure by using the DesignShop approach. In a series of sessions, AEDC reinvented itself as a facility for other military services and, most profitably, for civilian aerospace projects.

Taylor's company, the M.G. Taylor Corp., also conducts DesignShop sessions at two facilities it owns called KnOwhere Stores. One is in Hilton Head, S.C., in a former drugstore, and the other is in a former department store in Palo Alto. The Taylor group hopes to open another ten KnOwhere locations over the next five years. However, the biggest physical expansion has come about through a licensing deal with Ernst & Young's consulting wing, which has taken the concept worldwide under the name Accelerated Solutions Environment, or ASE. Now part of Cap Gemini Ernst & Young (CGEY), the consulting firm has built 12 ASE centers on the Taylor model, five abroad and seven in the U.S.

Among the newest believers in the process is Fleming Cos. of Lewisville, Texas, which generates annual revenues of some $15 billion buying and distributing merchandise to mass merchandisers such as Kmart and Target, independent retail grocers, and small supermarket chains. Fleming also runs its own supermarkets in eight states. In April 1999 the company went into its first ASE session, looking for ways to cut operating expenses. Three days later it left with an action plan. Bill Marquard, a Fleming executive vice president, says that the result was a savings of $50 million over the remaining eight months of 1999, and triple that in 2000.

The biggest contributors to the sessions' success, according to Marquard and others, are trained facilitators who know when to push to keep things moving and when there's enough energy in the room to let the group build its own momentum. CEGY now has two dozen full-time facilitators, including Bill Rutley, who ran the Arnold Center while an Air Force colonel. CEGY has also lost some consultants, including Tracy Tsuetaki, now a senior vice president at Kaiser Permanente. Tsuetaki recently convened a cross section of employees at a DesignShop session in Palo Alto to work out a core strategy for the big health-care organization.

No session shows the DesignShop process better than the one run for Seagate in July 1998. Though organized by Ernst & Young, it took place at Taylor's premises in Palo Alto. And while there were 20 E&Y consultants on hand, the principal facilitator was no less than the man himself, Matt Taylor, whose buzz phrase is "putting group genius to work." The goal was to convert Seagate from an aloof designer-manufacturer into one that takes its cues from what customers say they want.

Responding quickly to customers was precisely what Seagate was unable to do at the time. Built in large part by acquisitions that had been only partly digested, the company was divided into vertical "silos," or baronies, overseen by executive vice presidents who viewed one another more as arm's-length customers or vendors than as members of the same enterprise.

Up to mid-1998, Seagate had bucked the infotech industry trend to hive off and outsource everything but core operations. It prided itself on vertical integration, which, it kept saying, was a key competitive advantage even when the company was obviously growing less competitive. At factories spread around the world, Seagate built all its drives' major components: the disks on which data are magnetically stored, the motors that spin them at thousands of revolutions per minute, the head and tracking mechanisms that read and write information, even the printed circuitboards.

It didn't help that one part was made in Minnesota while others were produced in California and Ireland, that subassemblies were put together in Malaysia, and that final assembly was done in Singapore and China. Each entity kept its own lead times and its own buffer stocks of parts and inventories of finished goods. Manufacturing went on producing, relying on the salespeople to push out the product. In a business where change is everything, recalls William Watkins, now chief operating officer, "manufacturing defined the supply chain. They told the sales guys, 'Here's what I can build for you, so here's what you will sell.'"

To get ready for the DesignShop session, E&Y and Seagate people gathered material on what other companies were doing, how they were doing it, and how they compared with Seagate. Benchmarked companies were not limited to those in high tech. For COO Watkins, one great model is the Seiko watch factory in Japan. But even Wal-Mart and P&G were included, because, says Nichols of CGEY, "the high-tech industry is becoming more like them." As he further explains, with an observation that a lot of infotech companies ought to post on the wall, "It's all about product availability and turnaround and low inventory. You've got to forget you're a tech company. You've got to start thinking differently."

Some 80 Seagate people, carefully chosen from different sections of the company and different levels on the organization chart, were brought into the session. Says Nichols: "We spent days choosing the right people--progressive thinkers, people who are looked up to in the organization." Also asked to attend, he recalls, were "people we thought might be barriers. Sometimes your best approach is to target them and include them in the solution." Getting underlings to speak up in the presence of bosses was, he recalls, not a problem. "They were so frustrated, they knew if they didn't fix the company they'd be looking for another job anyway."

Progress over the three-day events can be roughly divided into "scan, focus, and act." "Scan" in the first hours of the Seagate session involved looking at what other companies were doing, as a way of getting the group to realize that Seagate's operations weren't necessarily world-class and that major change was possible. Before the day was out, the participants divided into four teams. Three were asked to act as disk-drive manufacturers: The first was to assume that it was Seagate; the second, IBM; a third was told to think like a hypothetical company named Swarm, with unlimited capital backing from a Japanese firm and no technical legacy or restraints on where or how to build product.

The fourth group played the role of the customer, a computer company operating in 2003. This group, which included representatives of a real Seagate customer, worked out a request for a quote on disk drives that was handed to the three other groups for bidding. Nichols says that setting the problem five years in the future helped everybody to start "thinking out of the box." In the end, recalls Taylor, it was Swarm, with no organizational constraints and no fixed way of doing anything, that won the theoretical contract. As day one came to a close, just about everybody was ready to concede that Swarm deserved it.

The goal for day two was to take the Swarm group's new ways of thinking in the proposal and create a new Seagate operating model. Participants were advised, in Nichols' words, to "ignore functions, ignore who currently does this or that at the company, and just figure how you want to do it." Subteams looked at different aspects, reported back, argued with other teams, and then, after arriving at some sort of consensus, went away to rework their ideas. By the end of the day the whiteboards were filling up, the pieces were starting to lock together, and a broad concept began to emerge of what Seagate should look like and how it should act.

Day three was devoted to laying out a road map. Here's where we want to get; how do we do that? By when? At the end of the session, Seagate had a two-year timetable with projects and goals broken into seven sectors, each under the wing of a senior executive: management, product development, supply, manufacturing, logistics, planning, and information technology. Improvements and changes were targeted in each area, with an overall goal of saving as much as $500 million a year by the start of fiscal 2001.

Watkins claims that the results have been impressive. He says, "We were doing 6.5 million drives a quarter with $800 million in inventory. Now we can do almost double that with $400 million." Like all the contenders in the disk-drive business, Seagate is also de- livering more storage for the dollar: The average capacity of its desktop computer drives has about quintupled. However, com- petition hasn't subsided, and there's still too much capacity. As is ever true in the computer business, prices keep declining as the capability of the equipment increases.

Seagate is still far from back on track. In fiscal 2000 it shipped 39% more drives than in 1998, but total revenues were down about 5%, to $6.5 billion, lower than they've been since 1994. Even when restructuring and other unusual expenses are excluded, there was an operating loss of $4 million. Only nonoperating activity, including sales of investments, kept the company in the black. Nevertheless, some light started to break in Seagate's first 2001 quarter. While prices continued to erode, revenues crept up a bit on record volume. Net income jumped to $75 million, from $2 million a year earlier.

Moreover, the ground has been cleared. Seagate is reorganized. Manufacturing is being leaned down on the way to quintupling output without expanding space. New Oracle-based computer-planning systems for supplies, demand, and manufacturing are in place or about to get up and running. The number of suppliers has been cut sharply to what Dave Wickersham, executive vice president for global disk-storage operations, calls "partners that really have the technology to stay up with Seagate and where we're going." Even drive motors, a once jealously guarded key component, are being partly outsourced. And the first "factory of the future," in which almost all production is integrated into one line that can make any drive, is about to start up in Singapore.

Futurists see a day when the complicated mechanics of the disk drive could be replaced by semiconductor circuitry--it's available now as expensive, low-capacity "flash" memory. But Seagate says that there's lots of improvement of the old technology still to come. Watkins envisions disk drives the size of a wristwatch, with 300 gigabytes of storage. He also forecasts an explosion in the use of hard disks in consumer products of all types. Seagate has already shipped a couple of hundred thousand drives for use in personal video recorders, known as PVRs, that store movies on hard drives instead of on tape and that search out and store programs about your sports, hobby, or favorite actor.

Wall Street remains skeptical, but management believes there are still profits to be made in the hard-drive business. Last November, in a "going private" deal with more angles than a billiard tournament, an investment group acquired the operating assets of Seagate for about $2 billion in cash and shifted them into a new private company that took the Seagate name.

Senior managers stayed onboard, swapping stock and options for a piece of the new private operating company. Now they can continue to restructure and tidy up without kibitzing from Wall Street or the need to file those annoyingly revealing SEC reports. It's no bet at all that Seagate's executives dream of someday taking the company public once more. But how they cope in the disk-drive market in the meantime and what Seagate looks like when that day comes will certainly owe a lot to three days in July 1998, when they worked among those movable walls, books, and toys.

FEEDBACK: psiekman@fortunemail.com