On the face of it, there's little to distinguish $30-million Offtech Inc. from any other independent office-equipment dealer. A regional representative of Japan's Ricoh Corp. products, the six-year-old New England company sells and leases office copy machines, with smaller side-lines in facsimile machines, paper shredders, and other instruments of white-collar productivity.

One difference, however, is that after only a few months in business, founder Stephen Albano decided to bring in a top-shelf manager of service. Not to minimize the company's financial exposure to a necessary evil, though, as fellow dealers viewed the service end of their operations. Rather, Albano wanted to test his notion that service -- however labor dependent, parts hungry, and at the mercy of random cycles it may be -- could become a significant profit center.

"People will hire the best sales manager they can get their hands on," Albano observes of the competition, "yet they refuse to pay up for a service manager. Not only do they not look at service as an opportunity, in most cases they take a loss on it." Albano, on the other hand, paid up plenty in 1983 to win over a former leading branch service manager with Xerox Corp., his own corporate mentor. Sure enough, the new man, John C. Milioto, developed Offtech's service turf to the point where contributions from service alone now account for about half the company's gross income and revenue. It's safe to say that a fraction that large is rare, if not unique, among the country's 5,000 or so office suppliers -- not to mention automobile dealerships, consumer-electronics retailers, and anyone else who is obliged to fix things that break.

But before you go around hiring the country's best service managers, Albano cautions, locked-in revenue from full-service maintenance agreements is a priority. That way you can tell exactly what you can afford. Other dealers, Albano has found, place full-service contracts with only 30% or so of their customers. Virtually every Offtech customer is sold a full-service contract, and with those contracts on the books, says Albano, "we knew we could bring in a vice-president of service operations, and that we could automate our dispatching, and that we could hire a full-time service trainer and afford the facility in which to do service training, and we knew how much parts inventory to stock." End of Lesson One.

Lesson Two: margin. To maximize margins, Offtech sets up its data on a per-copy, or "meter-click," basis -- an important calculation, says Albano, who even shaves revenue into revenue per copy. Milioto rides herd on per-copy service costs, breaking down charges into fragile portions literally as thin as a mil. Such precision yields an unassailable portrait of the cost of supporting a customer, against which a pricing schedule can reliably be drawn. While a difference of a tenth of a cent may not seem like much in the fiscal scheme of things, from one customer processing 2 million photocopies a month (as some do) it adds up to $24,000 a year. A reasonable piece of change from the investment of a little care, yet, Albano marvels, "most independents can't be bothered with detail."

Thus does Milioto carefully formulate service costs into so many mils for labor, so many mils for parts, and so on. To achieve profitability, per-copy labor and parts cannot exceed those two numbers, which together comprise about two-thirds of service costs (travel, training, and fixed overhead take up the remainder). Overcontrolling parts affects the reliability of the machines, however, and stinting on labor plays havoc with response time to service calls. "It's a delicate balance," Milioto admits, and one which never quite settles down. "But in this business, if you control expenses haphazardly, you sacrifice customer satisfaction."

Still, if you create a sufficiently sensitive measuring system, it can be achieved. Milioto separates out elusive line items that other service operations simply sweep into overhead. Apportioning a cost for just about everything in sight -- administrative support, management support, use of company computers, even a share of commissions paid on service contracts placed by the sales department -- results in "a true cost, as best as we can determine."

But how to generate appropriate income based on that cost is left to corporate and sales. Letting others decide the charges for providing service, Milioto explains, "allows the people who work for me to focus on one thing: how do we lower the cost of doing business? They don't have to go to the other side and ask, how do we increase revenue? As long as the other side is setting the right prices and I'm controlling the costs, we'll always have good margins."

The other side seems to be setting the right prices, all right: its service charges are at the high end of the industry, yet nobody's complaining. But on the subject of how good the margins are, Offtech is mum. Still, they can't be too shabby. When Milioto came to Offtech in 1983, its patched-together service operation employed only 15 technicians, plus a "lead" technician who ran the department when not out repairing machines. Now there are some 140 technicians working under 10 service managers working under one Milioto -- a synchronized force that accounts for a hefty 60% of the company's labor.

Which brings us to Albano's next lesson. By themselves, selling full-service contracts and controlling labor and parts do not a profit center make. Vigorous recruiting, he insists, is equally essential. "Our peer group around the country doesn't focus on recruiting," he again marvels in disbelief. "We do it better than almost anybody." Indeed, Offtech rarely hires experienced repair personnel, preferring to snatch graduates out of two-year technical schools, give them their first career job, and train them from scratch. To that end, Milioto haunts the halls of electronics academia, delivering fervent pitches to promising students like a football scout for Alabama.

After having his or her (well above an industry standard of about none, one-fourth of the company's technicians are women) appetite whetted by Milioto's slick show, it then proves about as hard to get into Offtech as into MIT. To gain entrance merely for a follow-up job interview, the applicant must first take a comprehensive test designed to reveal desirable service traits: reasoning ability, loyalty, and the willingness to work hard. If that is passed, the rest goes swiftly. "You sit there and you get one hour with me," Milioto sums it up, "and you're either hired or you're not."

Expanding manpower at the same rate as growth is also critical to service profits, and there's little room for error. If you underhire, the quality of service will suffer, high-end prices can't be justified, and customers will complain -- or worse. The timing is made all the more difficult by the fact that it takes Offtech about three months to hire and train a recruit. For manpower to keep up with sales requires Milioto to predict needs at least a quarter in advance. "Don't get ahead of me, let me know so I can prepare with you," he pleads with sales.

Should sales get ahead, however, Offtech builds in a certain percentage of salary for overtime. "It's always helpful to plan overtime as part of your manpower requirements," says Milioto. But because too much overtime indicates that something may be wrong with the estimates, Offtech keeps monthly watch on the empirically determined figure. "We've found 3% to be a good number over the years," he says.

For objective guidance about overtime and recruiting, though, Milioto relies on a detailed computerized business model that predicts manpower based on the number of machines and customer volume. The projections are updated frequently with sales estimates and the period's empirical data -- how many parts each technician was shipped, calls per day, how long the call took, what percentage resulted in call-backs, and so on. Factor those by Ricoh's own predictions of how many copies a new model is expected to produce between breakdowns, and the future pretty much takes care of itself.

But not quite. Just as there's a danger of losing customers by stinting on labor, there's a danger in too much labor as well: expenses without revenue. "Anybody could hire 300 people to cut down the response time to 15 minutes," Milioto says, "only you'd go out of business." Thus he aims to keep manpower stretched that without breaking. While current data might show the machines are reproducing 15,000 copies between service calls, Milioto will put them down for 16,000, just to make sure the gang has something to do. "If I planned for 15,000," he contradicts the computer, "I'd be building in slop."

To whip up worker enthusiasm and beat down labor costs, some dealers stage contests to see who can fit in the most service calls per day -- 5, 6, maybe 7. Of course, that drives labor costs down, but at the expense of not satisfying clients. "We don't do that," Milioto boasts. "We plug in performance right here." The computer insists 1.29 hours is what the average call takes, given that the technician replaces parts when necessary and doesn't hurriedly patch together the machine's innards with baling wire. Take on half an hour for travel, and it's gospel that a lunch-taking Offtech serviceperson will execute 3.5 missions per day. "What we've tried to do is balance our cost of doing business with our ability to respond to the customer. If you manage to fit in 7 calls a day," Milioto reasons, "the equipment you're working on will deteriorate to the point that the first time a competitor's sales rep comes in and offers something better, you're going to lose the account."