Tell us the bicycle story," a voice called out to the man in the pin-striped suit standing at the front of the room.

"No-o, not the bicycle story," another voice moaned.

It was early on the first Friday evening of September 1984, and 90-some employees of Sequent Computer Systems Inc. and their spouses and friends, in ties and jackets and silk dresses, were sitting on the carpet like kindergartners.

"You really want to hear it?" the man in the suit asked incredulously. He had sandy hair, a beard, thick glasses, and two buttons on his lapel; one said "Sequent=Easy to Do Business With," the other, "Casey Powell." He was Sequent's president. That afternoon, the Portland, Ore., company had introduced its first product, a new type of supermicrocomputer, and Powell had just shown an abbreviated version of the slide show he had been taking around to editors and industry gurus all week. The box itself, about the size of a dorm-room refrigerator, blinked red lights from a table behind him. There was wine to drink, stuffed mushrooms to eat, and five kinds of cheesecake in the next room, yet the people in front of him were asking for what sounded like a bedtime story.

"No," yelled a chorus of voices.

Powell cleared his throat, and several people reached for their wine.

"I'll tell the short version," he said, hooking his thumbs behind his lapels.

"Boy," said Powell. "Bicycle. One hundred dollars." People laughed.

The long version began "Once upon a time," and most people at Sequent knew it by heart. It was a hokey story Powell had made up about a little boy who wanted a bicycle more than anything else in the world -- and not just any old two-wheeler, but a "very flashy bike" that cost $100. Being "a real process person," this boy set himself "measurable goals" to make sure that at the end of the year he would have enough money to buy his bicycle. He figured that, giving himself two weeks off for vacation, he needed to save $2 a week, or 40? a day, or 5? an hour to earn the money in a year.The moral, of course, was the importance of setting goals and working hard for what you want.

Powell never told the end of the story -- even the beginning had become a joke -- but it was still a company ritual. Whenever Sequent reached a goal, a nickel would be dropped ceremoniously into a special jar.

For this particular occasion, Sequent's engineers had planned to buy a bicycle, which they would pass on to the marketing group to symbolize the transfer of responsibility for the company's progress. But money was tight, so Roger Swanson, the director of software, had borrowed a Schwinn Pixie from his daughter Kris. Swanson had attached plastic bugs to the handlebars, representing the problems the software people still had to work out. When Powell was finished with the story, Swanson wheeled out the bicycle and handed it over to Barbara Slaughter, the director of marketing.

People clapped, then stood up and wandered back to the wine and the stuffed mushrooms. "No, I'm not going to ride it," Slaughter said. "I'm wearing a skirt."

For an anthropologist trying to define the "corporate culture" at Sequent Computer Systems, the bicycle story would be a good place to start. Like much of the rest of Sequent's culture, Powell's fable was manipulative, self-conscious, even silly -- and carefully thought out to focus the attention of every person in the room where management wanted it, on the achievement of yet another milestone on the company's planned path to success.

An industrial Margaret Mead wouldn't have to stop with nickels and bicycles, however. There would be all kinds of strange artifacts and rituals to examine.

Take the red light on the manufacturing floor, which gets turned on whenever anyone finds a quality problem on a production line. The entire line then shuts down, with the red light burning day and night until the problem is solved.

Or take the objects displayed in the glass case next to the computer room: the Dom Perignon bottle that was emptied after the company was named for the second time; the photo of all the dogs, kids, spouses, and employees at the company's first annual camping trip; the silver baseball bat used by Gary Fielland, Sequent's computer architect, to tap the people he needed for the second product.

Sequent even held "culture classes." Three days before the company was to ship its first product, 22 employees spent an entire working day in a local motel talking about shaping their company's culture. "No" was not an acceptable R.S.V.P., even if the software wasn't debugged yet.

But most of all, a visiting anthropologist with his or her eyes even slightly open would soon see that at Sequent -- beyond any single artifact or ritual -- the culture is the company. As Barbara Slaughter puts it, "We really believe it's how we work together, not our technology, that's going to make to break us."

At the time Sequent was launched, in the winter of 1983, few presidents of companies were making up stories about little boys and flashy bicycles, but a growing number had begun to believe that a company's culture -- its values, its environment, and the way its people interacted -- was a major factor in its success or failure. That idea was not new. Early in this century and even before, people like Tom Watson, the founder of IBM Corp., and William Cooper Procter, one of the early leaders of Procter & Gamble Co., had talked about the importance of making a company a good place to work. Until recently, however, many of their pronouncements were viewed as quirks of autocratic or paternalistic personalities rather than as useful managerial tools.

But in the 1970s, attitudes began to change. Up the Organization, Theory Z, and other best-sellers focused attention on the importance of delegating responsibility and keeping people working toward common goals. In 1982, Terrence E. Deal, then a Harvard professor, and Allan A. Kennedy, a management consultant, came out with Corporate Cultures, a book that offers a conceptual framework for looking at the web of relationships and values that influence how people work. In brief anecdotes, the book describes how certain elements of a company's environment -- most notably the values a company espouses and its corporate rites and rituals, by they formal meetings or Friday afternoon beer busts -- interact in a variety of American corporate settings. A corporate leader, Kennedy and Deal said, could manage a company's values and rituals just as he managed a marketing program to help achieve corporate objectives. They dubbed the executives who did this "symbolic managers."

Corporate leaders across the country picked up this message. People telling stories about IBM, Tandem, or People Express started using terms like "values," "heroes," and "rituals." Managers made cult heroes of people like Mary Kay Ash, the cosmetics queen who uses diamond bumblebee pins, pink Cadillacs, and motivational stories to inspire housewives to sell millions of dollars worth of cosmetics. Digital Equipment Corp. even hired an anthropologist to help design a team-management system in one of its plants.

Casey Powell has never read Corporate Cultures. (He classifies it as a how-to book, a genre he avoids, believing it is a mistake to overanalyze something one does well intuitively.) When asked, he didn't know what a symbolic manager was. But he and the rest of the people at Sequent are refining the management of corporate culture into a high art. They believe that the environment a company provides for its employees is the key factor in how well they work together to achieve company goals. From the beginning of Sequent's two-plus years of existence, Powell and his three vice-presidents, Scott Gibson, Dave Rodgers, and Larry Wade -- along with Barbara Gaffney, the company's human resources director -- made culture a deliberate priority, and they have nurtured Sequent's values, heroes, and rituals as carefully as a parent nurtures a child.

It makes you wonder where the dedication came from.

On a Tuesday morning in the fall of 1982, Casey Powell flew down from Oregon to the Santa Clara, Calif., headquarters of Intel Corp., one of the world's leading semiconductor companies, where he was then employed. A boyish-looking 37-year-old general manager who had joined Intel seven years earlier as a salesman, Powell was four months into what was perhaps the biggest challenge of his career -- the turnaround of a $100-million microprocessor operation.

The turnaround was crucial to Intel. The microprocessor operation represented 10% of the company's total revenues, and that summer, IBM had announced it was dropping Intel's latest microprocessor product, the 8086, to use a competitive chip in its Displaywriter. Intel management was nervous. The semiconductor industry was in a slump, and for the second year in a row, profit margins were a fraction of what the company liked to consider the norm. Powell had 30 minutes at an executive staff meeting to explain how he intended to proceed.

Powell knew that several of the executives present thought he could do his job better if he moved to California. Although the operational part of the job was based in Oregon, the marketing program, known as Operation Checkmate, was corporate, and the executives had wanted him to lead it from headquarters. Powell had told them that he would not move. He had moved for Intel twice in the past three years; he had been in Oregon only 16 months, and his daughters were just beginning to learn their friends' last names. Although he thought Intel was one of the best-managed companies in the world, and wanted to continue his rise within it, he and his wife liked the more traditional values of the Pacific Northwest and did not want to raise their children in the Bay Area.

Powell is a natural performer, with a salesman's love of telling stories and a stage actor's feel for an audience, and he generally enjoys presentations, even the tough, confrontational kind for which Intel had earned some notoriety. He appreciated Intel's emphasis on achievement and the formal management-by-objective system that allowed ambitious managers like himself to set high goals for themselves, then earn the rewards of reaching them. Although his turnaround was taking longer than expected, Powell was sure that Operation Checkmate would work. His task that day was to convince the rest of management that he was right.

At Intel, such grilling of managers is officially called "constructive confrontation"; privately, it is called "destructive confrontations," "guerrilla warfare," and "table pounding." Although the intensity of the inquisitions is sometimes hard on newcomers and the thin-skinned, most Intel-ites are proud of their willingness to point out weaknesses in one another's thinking, and they enjoy working for an organization that demands high performance. So it didn't surprise Powell when one of the executives interrupted his presentation to ask questions. What was surprising was the direction the questioning soon took.

As the Intel inner circle cringed, a second, very senior executive started asking a series of questions that were, according to one of the men present, "depersonalizing." The tone was that of an officer berating a cadet who hadn't done his homework, and the message was tied in their minds to Powell's unwillingness to move to the Bay Area. They say they will never forget the faces.

Here was Powell -- a loyal, hardworking employee who had done well on the Intel grading system ever since he joined the company -- standing in front of them saying "Yes, sir" and "No, sir" as if he were back in military school. And here was a senior executive lighting into him in a way that the most battle-hardened Intel veterans found too painful to watch. Finally, two of them told the executive that if he wished to continue this line of questioning, he should do it somewhere else. The executive called a break, and everyone left the room. In the hall he apologized, but Powell, according to friends, was so upset that he turned around without saying anything and walked off.

It was one of the few times Intel veterans can recall a "constructive confrontation" turning into a personal attack. The outburst would serve as a reminder of what could happen when the pressure got so intense that people forgot a basic tenet of a management-by-objective system, which is the faith that talented people, given an objective and the freedom to do things the way they think best, will produce results. It changed Casey Powell's life -- and, as it turned out, the lives of a lot of other people as well.

Powell refuses to comment on this incident for publication, but shortly afterward, he told a friend that he was going to execute the marketing plan in the manner outlined in the meeting, and then quit. The senior executive later apologized again, and praised the marketing effort as one of the best ever executed at Intel. But Powell never looked back.

He realized that he was tired of solving problems that were the result of comeone else not doing something right. So when a venture capitalist asked if he would like to try turning around a troubled company, Powell said he would rather deal with his own mistakes. He decided to start a company that would: a) provide him with a job in Oregon; b) enable him to work quickly and efficiently, which was one of the things he liked about Intel; c) give him the opportunity to satisfy professional goals without sacrificing his family; and d) create a working environment that would motivate people as well as Intel did without beating them up.

It was an ambitious agenda, but Powell set about making it happen in a very systematic way. Over the next several months, while still working at Intel, he approached a number of the people he wanted to help him. He was looking not only for outstanding track records and talents, but for people who shared his values.

The first person he talked to was Scott Gibson, a 30-year-old buy wonder from the Midwest who years earlier had become Intel's youngest general manager. Gibson was extremely bright and possessed of an enormous capacity for detail that complemented Powell's intuitive management style; he was also interested in building a company that valued both people and achievement. Systems architect Gary Fielland had a streak of pragmatism that let him focus his inventiveness on the kind of product an embryonic company needed. Larry Wade, another general manager who had spent much of his professional life at DEC, was looking for a way to make money and have fun -- with fun defined as working with people he enjoyed and producing something beneficial to society.

After agreeing to work together, Powell, Gibson, Fielland, and Wade quickly and discreetly chose 13 other Intel employees and one engineer who was working as a consultant in England.

On January 17, 1983, all 17 Intel-ites resigned.

That night, they celebrated at the Upper Level Pub, a bar at a nearby shopping mall. They sat at a long table in the back drinking beer and talking about the shock waves at the company, the attempts to get them to stay, and the local TV coverage describing them as "the cream of the crop" at Intel. One of the first toasts of the evening was about allowing the simultaneous accomplishment of company goals and personal objectives. The last toast was made as the sun came up.

There were 15 men and three women, ranging in age from 23 to 39, wearing everything from pinstripes and shined loafers to T-shirts and running shoes. Some were close friends, some knew each other only by reputation. All had agreed to buy stock, to work for as long as six months without salary while the company lined up financing, and to do whatever was necessary to make the company succeed. For months, this included taking home the company garbage in large plastic bags; there was no dumpster on the premises.

Within days, they had arranged a lease on a "deathlike space" in Portland recently vacated by a company on the road to bankruptcy. They negotiated what they came to call "very special terms" -- six months use for free -- persuading the owners that they would benefit later when Sequent got its financing.

That first Saturday they held a "Name the Company" party to which they invited family, friends, business associates, and The New York Times. More than 100 people ate, drank, and wrote suggestions on large paper banners hung on the walls. They vetoed such names as Trillium, Topaz, and Osprey, settling on Sequel, the name of a local rock band.

After the party, they gathered to write down their corporate objectives. Powell was the Thomas Jefferson, holding the pen longer than anyone else. Ten drafts later, they came up with six objectives, involving profitability, customer satisfaction, market domination, "a culture that rewards our employees for their contributions," "an organization that provides individuals with the means to accept the maximum responsibility for the overall success of their company," and acceptance of community responsibility.

During the next two months, they worked on their business plan and negotiated for various necessities of life -- furniture, telephones, a copy machine, legal and accounting advice, and a VAX computer -- all on "very special terms."

In April a group of well-known venture capitalists, led by Reid Dennis of Institutional Venture Partners, agreed to pay an unprecedented $5.2 million for 35% of a company that consisted of 19 people and a business plan the size of a small-town telephone book.

At the end of the month, the founders received their first paychecks, for $1, suitably framed for hanging in cubicles.

By then, they had begun work in earnest on their first product. It was to be a multiprocessor parallel computer, meaning that a number of microprocessors were to be put into one box and linked together with software and hardware. Although other companies had put more than one microprocessor in a box before, each microprocessor had been dedicated to a particular function. Sequent's machine was to be the first in which the microprocessors could work either on the same task at the same time or on different tasks at the same time. The low-end configuration (2 processors) was to be competitive with workstations. The high-end (later determined to be 12 processors) was to be competitive with superminicomputers, and even, it turned out, with the low-end of Cray Research -- type mainframes.

Powell and his crew believed that they could carve out a significant market for their computer, but to do so, they had to get there first, with a quality machine. The pressure to produce was intense enough to strain any culture. Yet unlike so many other start-ups. Sequent never lost sight of the kind of company it had set out to build.

People took great care with recruiting, for example, and refused to change their style just because their need increased. "We saw a few people who fit our technical needs, but didn't really fit in," says Roger Swanson, the software director. "There was one case I remember very well. I was in need of a particular kind of engineer. Our schedules were beginning to slip because we didn't have the people. Larry Wade and I had a long talk about one fellow. I really needed this person. And Larry said, you know, it just doesn't feel right. This fellow was just too laid back. And I said, Larry, I have to agree with you when I think about it. But it was really hard to make those trade-offs."

The concern didn't stop once a person signed on. Periodically, Barbara Gaffney, Sequent's unofficial minister of culture, led workshops in which about 20 employees would spend the day talking about whether the company was living up to its values.

One of those values dictated that there would be no walls between departments. Everyone, not just the managers, was expected to walk around the building to find out what was going on.Marketing people were expected to show their faces in engineering, engineers in manufacturing. People were expected to serve on cross-functional task forces, and got graded on cross-functional interactions during their performance reviews.

Scott Gibson -- vice-president of operations and finance, and inventor of the red-light process -- put his desk on the manufacturing floor.

At a cost of $2,500, Sequent installed a terminal in every employee's house so family members could talk to one another all day -- and so employees could go home for dinner or to tuck in their children even when they had to work until midnight. There were family parties, camping trips, ski trips, and picnics. On weekends it was not unusual to find children crawling around on the carpets.

The company started a yearbook, written mostly by employees' spouses, with sections on people, kids, culture, and the year's significant events. The 1983 edition includes the bicycle story, complete with photo of Sequent's jar of nickels.

In order to foster a sense of shared responsibility, employee involvement in even the smallest decisions was encouraged. Barbara Gaffney once came back from a conference to find her office stacked with ice chests and cans of pop. It had been converted into a soda-tasting stall. It seems that as the company had grown, the percentage of vending-machine slots dedicated to natural drinks had declined, and one day a software engineer had rebelled. To find the best natural sodas to occupy the available slots, Dave Rodgers, the vice-president of engineering, had organized a blind tasting. (The winner: Ol'Bob Miller's Natcherly sodas.)

Powell, meanwhile, played the symbolism of small events instinctively. Even at Intel, he had been the kind of manager who delegated operational and strategic responsibilities to top managers he trusted, leaving himself free to wander around the cubicle maze or the manufacturing floor or to make his own coffee. Now that he had his own company, he was as apt to talk about not leaving coffer rings on the furniture as to comment on the schedule. People could walk into his office any time -- like everyone else, he had a cubicle with no door -- and offer him a cookie, or sell him a mug to raise money for the year-book, or tell him they might not finish a particular task on the appointed day. Powell could bear down on a problem, but he would always offer to help, and then he would do what he said he would do, whether it was bringing in a pizza for the people who were working through dinner or buying an expensive engineering tool. Then he would let the people know that the life or death of the company depended on them getting their piece of the machine done on time.

It all added up to a carefully cultivated message that this company really cared.

What no one knew was, would it make any difference when the going really got tough?

By the fall of 1983, the walls had been painted, the floors carpeted, and the loaner furniture replaced with desks and chairs in the company colors -- gray and cranberry. The head count had doubled, and the company name had been changed from Sequel to Sequent because of a trademark conflict. As people settled into a routine of working 60 to 80 hours a week, start-up life seemed to be living up to its promise -- except that the hardware group was missing some of its deadlines. Not by much -- a day here, a week there, two weeks somewhere else -- the kind of slippages that were said to go with the engineering territory at larger, more established companies.

But Sequent was neither large nor established.

Roger Shelton first heard that the hardware schedule was in serious trouble in early November. Like most of the other hardware engineers, Shelton had spent the previous few months buried in the design of his own piece of the machine, and although he was vaguely aware that "things were not happening as quickly as we had thought," he had not stopped to worry about the bigger picture.

Shelton, like everyone else at Sequent, had a computer terminal on his desk. Important messages were announced by the terminal with a little beep, then green letters would run across the screen in a ribbon like a news flash on television. Although most of the messages had something to do with food -- the morning cookies have just come in, or there is popcorn in the kitchen -- the message that Friday afternoon announced an ad hoc, allemployee meeting in the cafeteria.

There were not enough chairs, so most people stood. Powell was direct: They were behind. They were approaching a big, externally visible milestone -- their first anniversary, a key board of directors meeting, and the deadline for providing the first physical proof that the company could do what it said it would do. If they kept moving at their current rate, the hardware would not be up and running until February 15, which was not good enough. They were starting to go out for their second round of financing, and beginning to talk to customers. They had to have working hardware by January 17.

Powell had talked to Gibson, Wade, and Rodgers, and they all thought it could be done. But it was going to take the help of every person in the company. This was not some big corporation where, if they were a month late, they were a month late. If this hardware wasn't finished on time, their desks and chairs might go away. The bank might come for their terminals and their CAD machines.

To keep attention focused on the problem, Powell had had buttons made up. He and Gibson and Wade and most of the rest of the company would wear green "How Can I Help?" buttons. Rodgers and the engineers on the critical path would get red "Priority" buttons. People with green buttons were to do anything to remove obstacles for the people on the critical path, whether it was getting them coffee, registering their cars, or talking to the vendors that were supposed to be supplying them with parts and services. These were all obstacles, he said, all the same. All obstacles had to be removed.

Powell passed out the buttons, and everyone went back to work. The whole thing took less than 15 minutes. The buttons cost $73.50. And every person in the company now knew what had to be done.

Shelton went back to his desk with a red "Priority" button pinned to his shirt. Although Powell had presented the buttons as a device to make the rest of the company aware of the gravity of the situation, Shelton and the other hardware engineers got another not-so-subtle message. "I think it was a way to make sure we understood how important what we were doing was," explains one engineer.

"We were already working hard," says Shelton, "but now there was this sense of 'Gee, we really need to get done."

It was symbolic management in action. Powell hadn't solved the problem; in fact he hadn't even defined it. (Rodgers had done that in his regular meetings with the director of hardware and his wanderings around the engineering cubicles.) He had merely clarified the company's priorities and driven a stake into the ground.

Powell didn't even go to the second meeting, scheduled for the next day, a Saturday. Neither did Gibson or Wade. This meeting, which lasted three hours, was for the hardware engineers; their director, Walt Mayberry; their vice-president, Dave Rodgers; and two guests, Barbara Gaffney and Roger Swanson. Gaffney was there because she knew the people in the company well enough to suggest who might help. Swanson was offering the services of the software group.

Mayberry, a quiet man with glassess and carefully trimmed hair, started by presenting the historical data -- how they had fallen behind -- and the new schedule, which he described as "reasonably aggressive, but possible." But Shelton and most of the other hardware engineers thought the situation was out of their control. A hardware engineer designing a circuit board first develops a set of schematics that describe the logic of the board, then designs the placement of the components on the board. Then, at a small company like Sequent, he sends this design out to a layout vendor, which lays out the design to scale and makes a "physical interconnect" for the photographic mask from which the physical boards are fabricated. Shelton had sent out a board to a California vendor in September, and although the standard layout time was four to six weeks, the board had still not come back. Another vendor had laid out another engineer's design incorrectly and sent Sequent a chip that didn't work. The company would fix the chip, it said, but it would probably take another eight weeks.The engineers didn't see anything they could do.

What Powell, Gibson, Rodgers, and Wade were telling them, said Mayberry, was that the whole company was going to pitch in. It was OK to take risks, OK to spend money to get things done, OK to insist that the vendors deliver what was promised. Rodgers was working on the design. Gibson had volunteered to talk to vendors. But they needed the engineers to tell them where they could help.

One engineer said that if someone else could design a piece of test software, he could concentrate on a more critical task. Another engineer had once seen boards laid out in a week, but he was pretty sure it had cost a lot of money. They spent an awful lot of time in meetings, another engineer pointed out. As the engineers offered suggestions, Mayberry wrote them down. After the meeting was over, everyone went back to work.

They came in on Sunday, too. People had been working 12 hours a day, six days a week; now they came in seven days, and instead of going home at midnight, they stayed until two in the morning. Gone were the days when they could grab a sandwich and a couple of video games at the Inner Space Deli down the street. They had to look at their watches to figure out what day of the week it was. "Sometimes you'd be in a meeting and you'd find out the guy next to you hadn't been home for 24 hours," Mayberry says. People actually wore their buttons, and if a person with a "Priority" button had a question for a person with a "How Can I Help?" button, the "How Can I Help?" person dropped whatever he was doing.

Gibson flew down to talk to the layout vendor, and decided to start another company working on the same design. He visited other vendors, telling them Sequent's life depended on them. He offered cash bonuses to get them to complete the work in record time -- 2 weeks instead of 4 or 6 weeks, and once even 2 weeks instead of 12 weeks. When a vendor came through on a particularly tough schedule, Gibson sent the company a singing telegram.

Rodgers took off his tie, sat down in front of a terminal, and only got up to go to the bathroom and to find out what was going on. Every afternoon around four o'clock, he and Mayberry would stroll by the cubicle of each person with a red button and ask what that person had done that day. Was it what he had said he would do? Was it the most important thing?Did he have any problems? Could anyone help? It made an engineer think twice about talking for more than 30 seconds over a cup of coffee, or taking off more than 24 hours for Christmas.

The visits helped draw attention to the task that needed to be done -- but they worked because the help offered was not just symbolic. When he had problems, Shelton says, "I told them what they were, and they fixed them. I'd say, 'We really need another logic analyzer, and it costs $18,000.' Or, 'We're standing around waiting to get onto the engineering workstations, and they cost $60,000 or $90,000.' In 18 hours they'd get us a rental, and then if we really needed another one, they'd buy it. The people who control the money at most companies don't even know what the problems are."

Soon it was apparent that the group was catching up.

It was dark inside the cake, and Casey Powell was shivering.

The date was January 17, 1984, Sequent's first anniversary, and Powell had lost a bet.

By the end of December, when the hardware engineers were clearly making up lost time, Barbara Gaffney had decided the time was ripe for a small wager, which became known as "The Challenge." As keeper of the culture, Gaffney was partially responsible for seeing that company milestones were celebrated with all the ceremony the occasions warranted. She talked to Dave Rodgers, who was all in favor of the idea, provided the stakes were more humiliating than the usual dinner or lunch.

If the hardware group met its deadline, Rodgers suggested, they should make their president jump out of a cake with "due pomp, circumstance, and little else." Gaffney had the proposal drawn up with scrolls and flourishes. The document was signed by the hardware engineers and presented to Powell at a company meeting.

The deadline had been met. Now the hardware engineers -- with a crowd of other Sequent employees, about 50 in all -- were waiting in the manufacturing area to see him jump naked out of this cake.

"Wheel me out," Powell yelled. "I am not going to do this."

"I have a small piece of attire for you to use," said Dave Rodgers. The top of the cake was cracked open, and in floated a paper towel. Powell could hear the hardware engineers laughing.

"Will it cover?" Rodgers asked.

"Wheel me out," Powell yelled.

"No," Rodgers said."This is it."

Flashbulbs popped, a videotape rolled, and one of the women looked away as Rodgers lifted the top of the cake, then opened the sides. The president of Sequent Computer Systems stood up wearing the red yarn wig, the bulbous red nose, and the puffy suit of Bozo the Clown.

"The real Casey Powell!" someone called out through the applause.

Lights reflected off Powell's glasses. He pushed up the mask, shook the hands of the hardware group, and said he had never had more pleasure losing a bet in his life. A champagne cork popped.

The hardware challenge was just the first of many. Sequent went on to complete its second round of financing for $7.5 million. There was a software challenge, and it was Roger Swanson's turn to buy the pizzas. In September, the company unveiled its first product, the Balance 8000, and began shipping units out for independent testing. In December, it shipped its first boxes for revenue. One potential investor called the testing companies and told Sequent that the evaluations he got were the best he had ever heard. Last March, Sequent completed its third round of financing for $10 million.

By this time, members the sales force were working on yet another challenge, which would turn them into "bound serfs" at the next anniversary party if they didn't meet their sales goals. And Powell's Bozo wig had long since taken its place in the artifacts case, next to the Dom Perignon bottle.

"Was it really a motivator?" Mayberry reflects. "No. We did wheel in a cake. Casey did jump out. It was something to talk about. Some people thought it was a bit hokey. But it was a focus.

"In retrospect, it was an event in the corporate culture."