Editor's Note: To celebrate Inc.'s 35th anniversary, Inc.com is showcasing highlights of our coverage of incredible innovators, risk takers, company builders, and thought leaders since 1979. Here, an article from our archives.

Nobody expected much from Gary Aronson, and he didn't surprise anyone. After dropping out of college before the second semester of his freshman year, he knocked around a while, eventually winding up as manager of a Kentucky Fried Chicken franchise. In 1983, he switched jobs again, this time moving to Au Bon Pain Co., a Boston-based fast-food chain specializing in fancy offees, croissants, and spotty customer service. That fact notwithstanding, Aronson viewed it as a different kind of company, or so he told himself, perhaps as a way of justifying his cut in pay from $410 to 280 a week. Different Au Bon Pain may have been, but three years later, Aronson was still making a meager $26,000 a year, with which he supported his wife and two children.

At 30, Aronson was frustrated, bored, and wondering why he persisted in the food-service business. Fast-food companies don't want managers, he thought; they want trained dogs who will go through their routines and keep their mouths shut. Aronson himself had a hard time keeping his shut. He was known in the company for being "difficult" and "opinionated" -- a whiner and a complainer. He didn't care. "My heart wasn't in it anymore," he recalls. "I had the feeling of being in a dead-end job. I thought if I was lucky, I might earn $3,000 more in five years." This was assuming, of course, that he didn't get fired, a distinct possibility. So he put in his 45 hours a week and tried to figure out what he was going to do next.

Today, Gary Aronson is still in the food-service business, and still with Au Bon Pain, but he no longer works 45 hours a week. Three mornings a week, you can find him in his store at 3:15, and on other days he's there from 6:00 a.m. until the store closes at 7:00 p.m. He works weekends, too, putting in a total of 65 to 80 hours each week, making sure that the food is fresh, the place is clean, and the customers are satisfied. His face looks weary these days, but he is not complaining. He's too busy thinking up ways to bring in more customers, sell more food, and make the whole operation run more smoothly. "This is the first time in my life I've been treated like a professional," he says.

It's also the first time he's been paid like a professional. If he continues at his present rate, he will make at least $80,000 this year. Suddenly he and his wife, Donna, also an employee, are the talk of the company. The betting is that they will be the first pair of Au Bon Pain shopkeepers to arrive at a managers' meeting in a chauffeur-driven limousine.

Companies have traditionally viewed compensation in a fairly narrow context -- as just one of the many levers available to influence the direction of a business. Of all the factors that affect a company's performance, compensation is seldom listed among the most important. If the company fails (or succeeds), the owners will usually blame (or credit) the product or the strategy, the financing or the timing. Seldom will they say that the crucial difference between success and failure is the way they structure the system for paying their employees.

And, in some cases, that may be true. After all, many companies succeed with a compensation system that's little different from their competitors', while those that fail usually have a multitude of other problems. And yet it is a fact that a company with an extraordinary record of performance almost always has an extraordinary compensation plan as well. In most cases, it has been installed by the founder, who had a vision of the kind of company he or she wanted, and an acute understanding of the kind of reward system that would inspire employees to create it. Then there are the handful of companies such as Au Bon Pain, which grope their way through a maze of obstacles before finally hitting on a compensation structure that makes most of the other problems go away.

The truth is that Ron Shaich (pronounced shake) had not given much thought to the issue before he became president of Au Bon Pain in 1982. He was only 28 years old at the time, with limited experience is business. As a student at Clark University, in Worcester, Mass., he had founded and managed a nonprofit compus convenience store in competition with a local Store 24. He had been so successful, and had had so much fun, that he decided to get his M.B.A., graduating from Harvard Business School in 1978. Thereafter, he worked briefly for a national chain of cookie stores, did some grass-roots political organizing, and dabbled in the world of campaign consulting. But his idealism soon led him back to business. "In politics, you build organizations and then tear them down," he says. Hoping to build something more permanent, he moved to Boston in 1981 and opened a cookie store on a busy downtown street.

The cookie store brought him into contact with Louis Kane, a 50-year-old Boston businessman who had acquired Au Bon Pain in 1978 from a French oven manufacturer. Shaich was interested in buying croissants to sell in his store, but Kane had other things on his mind. His company was in serious trouble, and he did not know how to save it. His expertise lay in real estate, not food service. Impressed with Shaich, Kane suggested a deal: the two would merge their companies, with Shaich becoming a partner, president, and chief of internal operations. Kane would focus on external issues, selecting expansion sites and arranging financing. Shaich agreed.

At the time, Au Bon Pain consisted of three bakery-cafes located on prime Boston real estate and staffed by its own French bakers. It was an expensive operation, and it was losing money at a rapid clip. Beyond that, the company lacked any sense of purpose or direction. Customers were treated carelessly, as if they were intruders, and employee turnover was high, even for a fast-food operation. The situation called for dramatic action. In short order, the new management team got rid of the instore bakers, eliminated the wholesale side of the business, and brought in Shaich's father, a New Jersey accountant, to design some financial controls. Then they turned to the stores themselves, replacing the old managers with new ones, whom they paid the going rate -- about $18,000 per year.

But Shaich was not interested in building just another fast-food business. "I wanted to create a truly better food-service company," he says. Good food -- "food you wanted to eat" -- was a given, as was making money. He dreamed of a company built around a general, and passionate, concern for its customers. That, he realized, demanded a certain type of employee, "people who did things not because the boss was looking but because they really cared." In order to attract those people, he knew he had to create a different type of environment. "We didn't want to accept the low standards of the rest of the food industry," he says. "We wanted to show the big guys -- Pepsi and McDonald's and Sara Lee -- that the conventional ways of treating people were not the only ways. We felt we could to better . . . . I wanted an organization where I'd want to work."

With that goal in mind, Shaich began to tinker with the compensation system, setting up a program in which managers could earn monthly bonuses for generating sales above a budgeted level, provided the store stayed within bounds on its food and labor costs. It was an idea he borrowed from the famous businessschool case study of Lincoln Electric Co., and it seemed like a surefire method of pointing managers in the right direction, thereby reducing the pressures on himself and the rest of the management system as the company grew.

And grow the company did between 1982 and 1984. As stores increased their volume, Au Bon Pain began adding units. Most weeks, Shaich worked 90 hours, spending the bulk of it in the stores, devising systems to handle the growth. Everywhere he went, he carried a message to employees -- that growth, if properly managed, would create opportunities for those who took care of customers. "I did everything I could to make people feel that they wanted to be here," he says.

But growth also put strains on the company, strains that promised to get worse with time. For one thing, the Massachusetts labor market was getting tighter and tighter, making it more difficult to find new managers and crew. That situation created opportunities for employees, but dangers for the company. "We were promoting people left and right," says Shaich, "sometimes before they were ready."

By the beginning of 1984, the company had 14 stores, generating annual revenues of more than $6 million, but the company's management resources were stretched perilously thin. The game plan, moreover, called for opening 10 to 15 new units in the next year and, at the same time, moving into the lunch market with a new line of soups and sandwiches. Shaich himself found that he no longer had time to give store managers the support they expected. So in April he brought in a regional manager from McDonald's Corp. as the vice-president of operations. "We wanted to give the stores the best leadership we could find," Shaich says.

It soon became apparent, however, that the addition of another top manager was not going to solve all the problems, many of which seemed to be related to the compensation plan Shaich had installed so optimistically in 1982. It wasn't working. In the atmosphere of constant change and growth, the company could not come up with meaningful budget targets for managers. Beyond that, the systems for recording operating results were overloaded, and people were constantly being moved before their actual numbers came in. As a result, the compensation plan had lost its integrity. Managers realized that their bonuses really depended not on their performance, but on Shaich's perception of it. Not that he was stingy. In the absence of clear guidelines, he tended to give something to everybody, but on such a discretionary basis that the system became known as "pennies from heaven."

To make matters worse, the new vice-president was busily destroying whatever lingering credibility the compensation system had. To fill the slots in the new stores, he hired new managers, many of them from McDonald's, at salaries $6,000 or $7,000 above those of the old managers. The latter were understandably furious, and they told Shaich so; a few even left. But he didn't intervene. "I felt I needed to give the guy the freedom to do his job." Unfortunately, it soon became clear that the guy wasn't doing his job very well, at least when it came to providing support for store managers, whose morale continued to plummet. "He managed downward," says Shaich. "He expected their loyalty but didn't feel he had to earn it. And he showed no interest in taking care of them as people."

By the end of 1984, Shaich began to have the feeling that the company was coming apart at the seams. Customer complaints were increasing, and the turnover problem was growing. Hard as it was to recruit new employees, the average stay had dwindled from one year to a mere seven months. The company also lacked adequate operating standards -- governing, say, where to keep the lettuce for sandwiches. Per-unit operating profits, meanwhile, were deteriorating badly, even as sales continued to rise, and some of the wrost performers were the new managers brought in by the vice-president of operations.

"Everywhere I looked," Shaich says, "there was another mess to clean up." He was frustrated, but no more so than his managers. They told him bluntly that they didn't trust the company anymore. Finally, in June 1985, Shaich did what he had to do, firing the VP of operations, putting the brake on expansion, and calling his father back to help rebuild the company. Once again, he took charge of operations -- and tried to figure out where he had gone wrong.

Things were worse than they had ever been," Shaich says, and he had a point, although on paper the company looked just fine. With 31 units from New Hampshire to Texas, it had annual revenues of $15 million and was still highly profitable. There was plenty of cash available, thanks to the recent sale of a franchise and Kane's success in raising $11.7 million from private sources. Perhaps most important, Au Bon Pain had established a clear identity for itself in the market. With a sandwich menu featuring tarragon chicken and ham with Brie, it could never be mistaken for another burger chain.

But all that was in jeopardy, Shaich realized, thanks to rising turnover, sinking morale, and operational chaos. He moved quickly to reverse the trend. Hoping to boost the managers' spirits, he raised the salaries of the capable ones and returned to the old policy of promoting from within. To aid the recruitment of crew members, he hiked starting wages 50? above those of competitors and began giving free televisions to employees who brought in new recruits. Those new people who stayed got college-scholarship assistance. And in case any managers failed to get the message, Shaich sent them all a memo telling them to holler whenever they had to work more than 55 hours a week. More than once, they did, and he dropped everything to go lend a hand.

But important as these measures were, they did not address the company's underlying problems, as Shaich was well aware. Somehow he had to regain the confidence of the people who judged the business every day: its customers. Again, he turned his attention to developing a compensation system that would keep managers focused on the all-important goal of satisfying the customer, but he found that he scarcely knew where to begin. The bonus system had been a dismal failure. What else was there? Looking for ideas, Shaich called a professor he knew at Harvard Business School, who put him in touch with a young colleague by the name of Len Schlesinger.

Schlesinger was a budding expert in the field of organizational behavior. He and Shaich met once, and again, and Schlesinger spent a few days at the company, talking with employees. Shaich liked him. "Len was somebody I could talk to about the business," he says, "and he really seemed to care." So Schlesinger was invited to join the company as a partner and executive vice-president.

For Schlesinger, accepting the offer meant giving up the likelihood of tenure at Harvard, not to mention a lucrative consulting business, but Shaich was persuasive. It was an opportunity, he said, "to build a company -- to create a system you care about." Schlesinger says that, it the end, his decision came down to one question: "Was I willing to believe my own bullshit?" The answer was yes.

His first job was to help the company come up with a new compensation system. After the "pennies from heaven" fiasco, Shaich wanted a program that would be simple to explain, easy to sell, attractive vis-a-vis competitors, and equitable within the company. It also had to encourage managers to focus on customer satisfaction.

Schlesinger began by assembling a compensation committee from among the company's managers. Together they explored the options. "People were tired of inside deals," he recalls. "So we wanted something that was very mechanistic, something we could defend." In the end, they came up with a simple system under which managers would be paid according to their level of responsibility and the sales activity of their stores.

Under the plan, every store's general manager would earn a base salary of $375 a week. Salaries would then rise as weekly volumes increased, up to $633.75 a week at the highest-volume store. "We were willing to pay more for the high-volume store," says Shaich, "because it was worth more to the company."

Managers responded enthusiastically to the new system, but -- unfortunately -- it did not accomplish what it was intended to do. Very quickly, managers figured out the fastest way to make more money was to be assigned to a higher-volume store. "The guy we wanted to be focused and caring was spending a lot of his time lobbying for a transfer," says Shaich. "What's more, we needed to move them through the system, so they usually got their way." As a result, the new system had minimal impact on the actual performance of the stores.

The situation was further aggravated by continued turnover among crew members, which was running 40% to 45% in the summer and fall of 1985, despite the fact that the company paid hourly workers a premium wage. Nothing they did succeeded in stemming the tide. "We'd run big help-wanted ads," says Shaich, "and we'd get maybe two or three replies for an opening." Often the entire corporate staff -- some 50 strong -- had to help make sandwiches and serve customers at lunchtime. And there was no end in sight.

"The pressure was really on," Shaich recalls. "I remember thinking, 'Why aren't we located in the Southwest? Why is all this happening to us?"

In October 1985, Shaich and Schlesinger took a break from the crisis to fly down to Orlando for the annual meeting of the Multi Unit Food Service Operators. Both of them felt battered and weary. They could take some solace in spending a few days with people who were struggling with similar problems, but that didn't help them forget their own. During one of the afternoon sessions, Shaich began doodling on a piece of paper, listing all the company's failings. Soon the page was filled with loops and arrows. The analysis went something like this:

It was hard for managers to find quality recruits, and when they did, there was no time to train them. The result was substandard work. Managers operated on the assumption that recruits wouldn't stay long, and they usually were right. When employees left, managers filled in at the counters, thereby ceasing to be managers. Eventually, they burned out. But though the system kept most managers from succeeding, they were promoted to other stores anyway -- where the cycle repeated itself.

And that's exactly what it was, Shaich and Schlesinger agreed: a cycle of failure. Suddenly, 1,000 miles from home, the nature of their troubles seemed crystal clear.

It was a revelation, albeit a rather depressing one. That evening, however, they had a revelation of a different sort. They were having dinner with Ed Eynon, then vice-president for human resources for a company called Golden Corral Corp., headquartered in Raleigh, N.C. A chain of 435 steak houses, Golden Corral had a reputation as an unusual food-service operation. Listening to Eynon, Shaich and Schlesinger quickly found out why.

Several years back, Eynon said, the company had come up with its own solution to the challenge of running a restaurant chain: it had given unit managers a piece of the action. In addition to a modest base salary of $16,000, a typical manager owned 20% to 30% of his restaurant. Some wound up earning $100,000 a year or more; none required much supervision; and turnover among hourly workers had been cut in half. Shaich and Schlesinger could hardly believe their ears. Says Eynon, "I felt like I was bringing water to thirsty men in the desert."

Quenched though they were, they doubted such a system could work in Boston. Nevertheless, the conversation did offer hope that there might be a solution to their problems, and they returned home determined to break the cycle of failure once and for all.

During the next few months, the two inundated the company with resolutions, exhortations, and memos. Schlesinger advocated a coordinated program to attract high-quality people, with clear incentives, better training, and more timely performance reviews. Shaich followed with proposals for a $1 increase in hourly pay; time off to participate in comprehensive training; and additional bonuses to crew members who stayed on board.

As time went along, however, they began to develop nagging doubts that more bells and whistles would do the trick. Burger King, they heard, had tried similar ploys, to no avail. The cycle seemed to have a life of its own, impervious to adjustments and modifications. Au Bon Pain's own director of planning had to confess at one point that he would have difficulty recommending the place to his brother. On reflection, Shaich himself realized that he could not conceive of working as a store manager at Au Bon Pain -- not under the current system.

But the gravity of the situation did not really come home to him until January 1986. One day, he was visiting a Bostonarea store, and -- before he had a chance to identify himself -- a customer came in and ordered a turkey-and-Brie sandwich on a croissant. The employee behind the counter rang up the order, then realized that the store was out of croissants. Fine, the customer said, make it with another kind of bread. That wasn't possible, the employee said; the other bread could only be sold in whole loaves. The manager appeared and asked what the problem was. The employee explained. Well, said the manager to the customer, we can give you your money back. A disappointed customer left the store empty-handed.

There you had it in a nutshell, Shaich thought. The company had gotten so far away from its purpose that customers were being turned away to avoid slicing up a loaf of bread. All Shaich's efforts to develop operating standards had succeeded only in rendering store managers unable to think for themselves. They followed policies and procedures instead of common sense.

New policies and procedures were not the answer. (Who cared if the lettuce was on the left side of the counter instead of the right?) Nor were more stringent appraisal systems or better policing, piecemeal incentives or bigger bonuses. He could provide his staff with all those things, and the customer would still leave the store disappointed, without his turkey-and-Brie sandwich.

At last he saw the problem clearly. It was a waste of time to try to tinker with the system. A year from now, he and his district managers would still be baby-sitting store managers, solving the problems that they seemed incapable of solving themselves. Somehow, the system had to be turned around 180 degrees. Store managers had to want to solve their own problems. They had to be able to decide for themselves that a satisfied customer was worth a couple of slices of bread.

"We had to turn assembly-line foremen into shopkeepers," says Shaich.

Says Schlesinger, "We had to bust the system wide open."

The revolution began in early 1986. Schlesinger and Shaich read everything they could find about the innovative management techniques of successful food-service companies -- Chick-fil-A Inc., in Atlanta, for example, and Luby's Cafeterias Inc., in San Antonio. They also hired a Lincoln, Nebr., consulting firm, Selection Research Inc., to help in their search. Then, in April, they flew to California to visit a company called Harman Management Corp., based in Los Altos, that had the distinction of being the first Kentucky Fried Chicken franchisee in the United States.

Executives of both Golden Corral and Selection Research had mentioned Harman as a model of effective management, but Shaich and Schlesinger were skeptical as they pulled into the company's parking lot. They already had doubts about applying to Au Bon Pain the techniques developed by a company operating in such places as Utah and Colorado. When they saw Harman's cinder-block office, their hearts sank. It looked more like a chiropractor's office than the headquarters of their industry's most innovative leader. Inside, their skepticism mounted. There, in the lobby, was a copy of the company's yearbook. "It was incredibly hokey," recalls Schlesinger -- filled with pictures of smiling Harmon's employees and their families at company outings.

Then they sat down with Jackie Trujillo, Harman's vice-president of operations, and their skepticism began turning to wonder.

She told them that the company's 200-odd stores did, on average, 20% more volume than company-owned Kentucky Fried Chicken units. Turnover, she said, was not a problem, nor was supervision. Store managers were responsible for watching their own costs, recruiting their own crews, and making their units successful. How was this possible? Well, the managers -- like those at Golden Corral -- were owners, getting a salary of $18,000 to $20,000, with opportunities to earn a share of profits on top of that. What's more, they had the option to buy 30% to 40% of the stock in their stores. The company's role was to cheer them on with banquets, rankings, and awards.

Schlesinger was furiously taking notes. He and Shaich could not help but contrast the situation to their own. Harman's stock-purchase agreements with managers were 3 or 4 pages long; Au Bon Pain's directives to managers often ran to 20 pages. Au Bon Pain was spending $250,000 a year on help-wanted ads; Harman spent almost nothing. But perhaps the greatest contrast was between themselves and founder Leon W. "Pete" Harman, whom they met later that day. A homespun fellow in his sixties, he seemed to have all the time in the world. They spent an hour chatting leisurely, without a crisis or an interruption. He said lots of people came to see him, searching for solutions to the same kinds of problems. They listened, but they seldom followed his advice, which he found a little baffling. "You know, at a private company, you don't have to be greedy," he said. "You can share it with your good people, and it all comes back to you."

Schlesinger, the former Harvard professor, was impressed. "Here was a guy with maybe a high-school education, and he seemed to have it all figured out. It was the most humbling experience of my life."

That evening, they visited two of Harman's Kentucky Fried Chicken stores and several other fast-food outlets in the same neighborhoods. Again, they were struck by the contrast. At Harman's, if not elsewhere, the bathrooms sparkled, and the crew members were well dressed and upbeat. Indeed, one of the assistant managers tried to recruit them, saying how great it was to work there, and pointing out that the manager was making $100,000 a year. "You could feel it," Shaich says. "Everything we had been talking about was real."

Two weeks later, Shaich and Schlesinger were back on the road, this time heading to Raleigh, N.C., to visit Golden Corral. It was the same story all over again. There, managers worked as hard as their counterparts at Harman's, and the best of them had incomes to match. "The big difference between them and us," says Shaich, "was that they had adjusted the reward level to the point where it was worth all the pain."

By the time the pair returned from North Carolina, they had seen all the proof they needed. There were but two remaining questions: Would it work in Boston? And if so, how? Schlesinger spent two full weeks tinkering with profit-and-loss statements from individual stores. He tried to apply the profit-sharing formulas from Harman's but found that many of Au Bon Pain's stores had too wide a range of volume to produce meaningful incentives. Then it hit him: why not tie the store manager's incentive to "controllable" profits (that is, profit less rent and depreciation), instead of store profits? A manager, after all, could do nothing about his rent. But he could adjust his use of labor, shrinkage, and controllable expenses depending on the level of business he did. If the manager took care of his crew, used them when he needed them, and watched his other costs, the stores could become more efficient and profitable than ever. "Our basic premise," says Schlesinger, "was that it was a lot more important to control outputs than to control inputs." It all made sense, and -- unlike Harman's system -- it seemed equally applicable to low-and high-volume stores.

So they decided to put their plan to the test. The six-month trial period began on July 15, 1986, at two Au Bon Pain stores, selected because their managers were, at best, average performers. One store, managed by Brian McEvoy, was located in a Hartford office building. The other, in a shopping mall in Burlington, Mass., was managed by Gary Aronson.

Schlesinger, who supervised the test, explained the rules to each of the managers. The company was, in effect, leasing the stores to them. It gave them goals for labor and food costs, but agreed to split the controllable profits on a 50-50 basis. They understood and went to work.

Almost at once, the stores began to change. Aronson got rid of one assistant manager to save on overhead, and hired his wife as a crew member. He and his remaining assistant began working longer hours, as many as 80 hors a week, and looking for ways to control costs and boost volume. They reorganized the store to increase its seating capacity; they developed wholesale and catering accounts; and they raised employees' wages. McEvoy took a different approach, holding his own hours to 55 hours a week, but increasing staffing during peak periods to assure prompt service. He also introduced a telephone express ordering service.

Schlesinger was amazed at the speed of the change. Overnight, managers began solving the problems they had previously dumped on the company. "We finally had a system that didn't accept excuses," Schlesinger says. "My role was helping them to build sales." Shaich was equally impressed when he paid a surprise visit to the Hartford store. The place was spotless. "I know when a store is running well," he says. "I could feel the difference. It was everything I had hoped it would be."

The numbers were just as exciting. Crew turnover at Aronson's store, for example, fell to almost nothing. Meanwhile, both stores were beating their targets by substantial margins. By the end of the six months, McEvoy had exceeded his sales goal by $74,000, and his controllable-profit goal by about $27,000. At that rate, he could expect to earn at least $55,000 a year. Aronson did even better: he was ahead by $54,000 on sales and around $45,000 on controllable profit -- meaning he was earning close to $75,000 a year. "We were convinced," says Shaich.

The next step was easy. In January 1987, they began rolling the plan out to the rest of the company.

It is a Friday afternoon in April 1987, and Ron Shaich is sitting with Len Schlesinger in an office at the company's headquarters. A store manager named Jim Morgan walks in. Morgan, an effervescent fellow of 26, has been with the company for four years, and right now he is in a hurry. The Boston Marathon is being run the next Monday. He just wants Shaich and Schlesinger to know that he'll be there at the finish line, selling croissants and beverages for his Au Bon Pain pushcart. Does he have the necessary vending permits? Oh, yeah, he took care of that himself. Will he need any help? No, no, he's convinced some friends to give him a hand. The bosses can relax. The situation is wired.

Perhaps nothing better illustrates how profoundly Au Bon Pain has changed in the past year. Time was when Shaich would have had to plead on bended knee to get a manager to take a pushcart to the Boston Marathon. Indeed, he had often pleaded with managers to take pushcarts to the Bayside Exposition Center south of Boston, but his entreaties went unheeded. These days, he doesn't give it a thought. Gary Aronson, who now manages a store in downtown Boston, has decided to cover Bayside. He expects to be there 60 to 80 days a year.

The new compensation system is working. It is already installed at 10 of the 40 company-owned stores, and those operated by manager/partners are outperforming the others by a wide margin. During the first three months under the new system, partner stores as a group ran 40% ahead of their profit goals, while the nonpartner stores were pretty much on target. So the company is going full speed ahead with plans to convert the remaining stores by year end. Meanwhile, experienced managers from other food-service companies have begun applying for jobs, and Schlesinger and Shaich are extending the concept to district managers and others.

To be sure, the system entails certain risks, especially for store managers. Each of them gets a base salary of $25,000 and a chance to win or lose. Even if the manager wins, and earns the monthly bonus, half of it (up to $7,500) goes into a reserve fund that is not paid out until his or her contract expires, thereby locking successful managers into their stores for the duration. "They're in the same position as company owners," says Shaich. "They can't just walk away." And like company owners, they have to solve their own problems, hire and fire their own people, set their own wage scale, cut their own deals. What they can't do is to compromise on food quality and customer service, which the company regularly monitors through in-store audits and visits by unidentified "mystery" shoppers. Aside from that, they're on their own.

Some managers will no doubt fail and have to be replaced. After all, the system is not for everyone. But for Au Bon Pain it is working so well that even Shaich finds the results hard to believe. The company projects sales of $35 million for 1987, and the stores have never run better, or with less support from headquarters. "We're out of the picture," he says. "It's a closed loop." The loop is so closed that he and his partners feel confident about letting the company grow at the rate of 12 new stores per year for the next few years.

Not that Shaich feels they have solved all the problems. The journey has been too long, with too many valleys and swamps, for him to believe it is finally over. But now, at last, Au Bon Pain meets his own test for a business. "This," he says, "is the kind of company I'd like to work for."