Back in 1974, Clifford Elling discovered his company had sprung a leak. Tiny cost overruns on a $2.3-million contract were beginning to drain money out of Elling Brothers Mechanical Contractors at a rate that threatened to put the 53-year-old Somerville, N.J., firm out of business.

As Elling watched in dismay, a carefully budgeted 15-month job to install industrial piping turned into a financial nightmare. Day by day, design specifications changed, the price of materials shot up, and unexpected delays increased payroll costs.By 1975, when Elling Bros. closed its books on the contract, 30 months had elapsed, and the trickle of tiny overruns had become a flood. Elling's company was out $250,000 on the job -- half of the privately owned company's total net worth of $500,000.

Cliff Elling decided it was time to plug the leaks on costs. He'd always taken estimates and budgets seriously; every bid was based on a meticulously prepared estimate of the materials and manpower each job would require.But budgets alone weren't enough to keep costs under control. Like any business, Elling Bros. functioned in a world of constant change. If budgets weren't updated promptly to reflect these changes, even tiny errors could multiply until they wiped out the company's profits altogether.

The problem was particularly acute in the construction industry, in which even the strongest firms typically operated with narrow profit margins and very little net worth to cushion the business against mistakes. One thing Elling did know was that there weren't any easy solutions to cost control problems, no magic wand he or a consultant could wave.

Elling Bros. wasn't prone to sloppy bookkeeping or risky ventures. The firm had been launched in 1921 by Elling's father and uncle, hardworking sons of a German immigrant farmer, who began by using the family barn as a small plumbing shop. Cliff joined the company in 1950, at age 26, after a stint in the Army and an M.B.A. degree from Wharton. For his first three years, he worked in the field, doing heating and plumbing work as an apprentice to the journeymen.

That fieldwork proved invaluable in giving Elling a sense of how to estimate costs and figure schedules. "Back in the '50s," he says, "there was a great lack of professionalism and training in the management aspect of construction. There was no way to become an estimator except to get a feel for the practice in the field. There was no school you could go to, no course you could take, no book you could read, to find out how long it takes to install a toilet, for instance."

As Elling worked his way up through the ranks of the family company (he became president in 1969), he learned that cost control was a subject most contractors paid lip service to, but felt very little urgency about. The 1960s in particular were "lush, plush days" for contractors, with plenty of highly profitable jobs for everyone. "If you have a budget of $850,000 for a job that pays a million," he says, "well, it's very difficult to lose money. You say, 'Why worry about all this detail? So what if I go over budget a little? I know I'm going to make money anyhow.' You worry about budgeting only when things are tight and competitive."

By the mid-1970s, Elling realized that conditions in the construction industry were changing dramatically. Competition was increasing and margins were becoming tighter. Elling's own company, moreover, found itself involved in far more complex projects. Instead of standardized piping for schools and government buildings, the company was now bidding on industrial jobs that often required innovative piping systems for moving liquids and gases. "If you make 10,000 Chevies you know how much the next one will cost and how long it will take," Elling explains. "It will be pretty much like the 10,000 before it. But in these kinds of construction projects you don't have that same degree of repetition. Each job is unique."

That lesson was hammered home by the disastrous cost overruns Elling Bros. experienced on its 1973-75 contract. The budget for the job predicted a modest but acceptable gross profit for Elling Bros.But very little went according to plan. The Nixon price control program first created critical shortages of materials, then caused prices to soar after controls were lifted. The work force, supervised by unfamiliar foremen in a relatively distant part of New Jersey, was performing at only about two-thirds of expected efficiency. And dozens of changes ordered by the customer often went unbilled and unrecorded. "We weren't yet sophisticated enough to keep detailed account of them all," Elling concedes. "We didn't have the mechanism to price and record them, so we didn't get the extra money in time."

Just to survive, Elling Bros. had to cut costs quickly, and scrounge for every penny. Purchasing was rationed; only urgently needed deliveries were accepted, and only when there was money to pay for them. The top six company executives, including Elling, all took salary cuts. Bonuses and profit sharing disappeared for everyone in the office and warehouse. Though the company cut its volume of work and trimmed its fixed overhead, Elling and other stockholders had to pledge their homes and personal assets to the bank to raise operating capital. "I've always prayed a few times a day," says Elling, "but I had never prayed for the company.For the first time I started to."

Elling knew he had to go beyond emergency survival measures if he wanted to stay in business over the long haul. His financially shaky firm would have to do something more drastic than slashing overhead. Elling would have to develop a budgeting and cost-monitoring system that would blow the whistle on costs early enough to do something about them before they escalated. Most of the problems of the 1975 fiasco, he realized, could have been minimized if he'd had accurate, up-to-date information -- instead of reports that were months old.

Elling huddled with his accountants and pored over books by cost-control experts. Just getting more information, he quickly realized, wasn't the answer. Elling Bros. had three purchasing agents who handled hundreds of daily transactions and a labor force of 75 that might be employed in 15 different locations at one time. Elling couldn't hope to keep track of the details of all this activity without drastically expanding his office staff and overhead.

Elling found the nucleus of an answer in a manual written by Jack Baker and a committee of the national Mechanical Contractors Association of America (MCAA). Baker, a contractor himself, pointed out that large, complex budgets can't be properly monitored by checking individual line items or by comparing current expenditures against the overall progress of the job. Instead, Baker suggested that budgets can be monitored by subdividing them into small, easily evaluated units and by focusing attention on those items on which costs are most likely to get out of hand. (See "Jack Baker on Cost Controls," Page 50.)

The new cost-monitoring system Elling established for the company drew heavily on Baker's principles. Each week, a foreman and a site manager report to Elling about the progress made on well-defined stages of work at each site. Progress is measured by man-hours expended and the amount of work completed. (Most units involve fewer than 100 hours of labor; none exceed 1,000.) Elling has his foremen and site managers send him separate opinions on each stage so he can compare reports: "You can measure a roof in square feet or a road in cubic yards of concrete. But I defy anyone to walk through a system of pipe or electrical conduits and say the job is half done or three-quarters done."

At the same time, Elling gets reports on the most cost-volatile materials that go into the job, chiefly pipe valves and fittings, small $5 to $20 items that add up to a major proportion of job costs. Overruns are more likely to occur here, he points out, and will have the greatest long-term effect, even though individual dollar amounts are small. "You concentrate your energy where the danger is," says Elling, "as any good general will tell you."

The information from these reports goes into a small Nixdorf computer the company acquired in 1974. The computer then prints out an analysis of where Elling Bros. stands in its projections for each job -- man-hours budgeted against man-hours actually spent, work accomplished, man-hours spent as a percentage of both the whole job and work accomplished, and finally, deviation from expectations, with a figure for man-hours over or under budget.

With the new monitoring system in place, Elling finally felt that he had the company's costs under control. Yet the system did not work any immediate profit magic: Elling Bros. continued to lose money. In fiscal 1971, the company had shown a respectable pretax profit of 5.5% on $4.8 million in revenues. By 1974, in the middle of the contract morass that taught Elling that he needed to monitor costs more carefully, pretax profit was down to 2.2% and, by 1975, to.9%. In 1977 Elling Bros. actually lost money. "It was," Elling says, "a rather bleak picture of marginal earnings."

Yet small gains slowly began to offset the losses, and none of the losses developed into catastrophes -- because they were caught early. In November 1979, for example, Elling Bros. started a job for a chemical manufacturer. By January Elling's reports showed that he was losing $1,000 a month. The reports also pinpointed the source of the losses: major overruns in labor at several job sites. Elling hopped in his car with a foreman and drove to one job site. Here he found union jurisdiction disputes over certain tasks, combined with late deliveries of critical materials. The problems were relatively easy to solve -- and by the time the job was finished, in August, Elling had come out well below budget, despite his early losses.

In dozens of jobs, the cost-monitoring system brought problems to Elling's attention promptly, when dollar amounts were still small and dramatic solutions unnecessary. Says Elling, "It's really a story of what didn't happen rather than what did happen. The triumph is that we survived. It may sound ho-hum, but not if you compare it to years of losses, frustration, and cliff-hanging."

Eventually, as economic conditions in the contracting industry brightened and the company recovered its lost volume of work, Elling didn't have to be satisfied with just breaking even. In fiscal 1979 Elling Bros. turned a 2.5% pretax profit on $6.6 million in revenues. In 1980, revenues jumped to $10.3 million and pretax profits were 5.1%. And last year the company hit $12.2 million in revenues, with a 6.0% pretax profit -- at a time when the industry average for mechanical contractors was only 1.5% profitability on revenues of $10.6 million.

"We're making more money than ever before," says Elling, "not because of fatter markups or less competition, but because we know in a timely way about any problems. The essence of good budgeting is having timely accounting and monitoring of changes. We have found it is critical to know where we stand, and the only way to do this is by having reasonable control over the budget and schedule.

"What you have done is constant," he adds, "but what you have to do is constantly changing. If you can imagine a football game where the goal line is changing as often as the line of scrimmage, then you can understand the kind of game we're playing."