When John Williams bought the Spence Corp. -- a $250,000 industrial fittings distributorship based in Milwaukee -- he found his salesmen were taking phone orders on scraps of paper. "Our paperwork systems looked like something out of a sixth grade art class," he says. "We had cards and forms and records all over the place, and I still had trouble keeping track of everything that was going on."

The mess forced Williams to take a hard look at the information systems he had inherited from the previous owner. He soon realized that paperwork was a significant barrier to the kind of growth he planned for the company. Yet simply cutting out paperwork wasn't the answer. "We were producing a lot of records because we needed to keep track of important things like inventory and orders," Williams says. "Believe me, nobody wanted to do any paperwork they thought was unnecessary."

Slowly, Williams began to streamline Spence Corp.'s office systems to make them more effective and less time-consuming. The process was successful: Over the last eight years, Williams has boosted sales by 900%, yet has added only a handful of new people to keep track of the much higher volume of business. "Our productivity is one of the highest per employee in the industry," Williams boasts, "and we do it without a computer."

Williams has developed eight rules for streamlining office systems that he says apply to just about any small company:

Rule 1: Always question traditional assumptions. "Just because everyone else always does something," says Williams, "doesn't mean it's necessary for your company to do the same thing." Sales at Spence Corp. took a sharp turn upward in 1974, he recalls, and he found his bookkeeper was struggling to keep up.

"Doris was typing about 50 statements at the end of each month for invoices that hadn't been paid, and one day she mentioned that they didn't seem to be very effective," says Williams. "We weren't getting paid any faster." The company's system was generating lots of work for no good reason, he realized. So the company eliminated statements.

Statements are an old fashioned convention that came out of stores, says Williams, because people liked to see where their account stood at the end of each month. But Spence was dealing with large industrial customers who had orderly accounts receivable and didn't need to be reminded of their balance each month. Although it eliminated statements, Spence still had all of the necessary account information on its internal ledger sheets. It just wasn't sharing it with the customer every month.

"Aging reports" for accounts receivable is another convention that Williams found impractical. "People will track their receivables over 60, 90, and 120 days," he says. "My feeling is if it's over 30 days, what difference does it make? It's past due. We hop on it and get it collected."

Williams has his bookkeeper go through the ledger once a month to make a list of past due accounts. Then he sets aside time to call the dozen or so offenders personally, whether the bill is $35 or $3,500.

Rule 2: Define the problem before you pursue a solution. There's a tendency to go off half-cocked designing systems and forms without understanding what the real problem is, says Williams. Even after dropping the statements, Spence's business was growing so fast that there was a 10-day lag in getting out invoices.

"The question was, Do we hire another person to type invoices, or do we change the system?" he recalls. "At first, adding another person seemed like a logical solution, but that would have just allowed the problem to continue because we hand't defined it properly. As we looked at the situation, it became clear that we were typing a lot of things twice -- that was the real bottleneck." As a result, Williams designed a combined order-entry/invoicing form so that all necessary ordering and billing information could be entered at one time. "That one form cut paperwork in the office in half," he says.

Rule 3: Ask the right questions. Williams knew that the effectiveness of Spence's paperwork system was a function of the answers it was providing. But, to be producing useful answers, the system had to be created by asking the right questions.

"What is the fundamental goal of all our forms?" That's the first question Williams asked. "The answer," he says, "is to allow for the smoother flow of products in and out of the company when the customer needs them. Likewise, we try to ask questions like, What information is relevant to our profit objectives? and, Do we really need this?"

Williams began asking these questions about his inventory cards, and ended up making another radical change in the company's systems. "Keeping track of every shipment received from suppliers and every customer's order for every part was a tedious, time-consuming task," he says. "By 1975, it was getting hopeless." Williams hired a summer intern to work on the inventory cards in 1976. "She did a beautiful job," he says, "but, again, we hadn't really defined the problem. A salesman took over the job when the intern left, and almost immediately, he was spending more time on the cards than on sales."

Then Williams started asking the right questions. What information did the inventory cards contain that the company really needed to know? Were the cards the only place to get it? The historical sales data on the cards was interesting but of little importance to Spence Corp., which was projecting sales on the basis of expectations, not history. In addition, the inventory cards had no bearing on inventory control levels because Williams's philosophy was always to carry enough inventory to meet the customer's needs. "This means if we stock an item, we stock it so the bin it's in is full," he says.

Williams realized he didn't need to keep a record of running inventory or of prices paid and sold at. The price paid already appeared in the accounts payable ledger, and the price an item sold at was recorded in accounts receivable. The inventory cards were a duplication of effort, and since Williams couldn't establish a clear link between the cards and the company's profitability, they were eliminated.

Rule 4: Keep it simple. This is the heart of Williams's philosophy on office information systems.

"The simpler the system, the easier it is to train a replacement," he says. Williams has had little turnover, but keeping the paperwork systems simple makes it easy for Spence employees to pinch hit for one another. "We always know where everything is," says Marianne Wulf, who is in charge of pricing and billings. "If the bookkeeper goes on vacation, I can easily fill in for her."

Keeping systems simple also helps gain the commitment of employees to their use. The call reports salesmen filled out when Williams first bought the company required the standard explanation of what happened on each call. Salesmen accepted them with the usual resistance, and because their use wasn't enforced, the call reports were discontinued. When the recession hit Spence in 1980 and knocked sales for a loop, Williams reinstated call reports as a way of instilling discipline in account coverage and as a tool he could use to give feedback to his salesmen.

"I always hatted doing those reports myself," he says, "and we wanted something that would only take a moment to fill out. Those old reports could be a pretty lengthy deal."

Williams found a solution by developing a much simpler form that gave salesmen a list of boxes to check indicating the type of call that had been made. "The salesmen just zip through these reports," says Williams," and it does the trick."

Rule 5: People are the key to good systems. "Information systems are only effective if employees understand the reason for them," adds Williams, "but you've got to ask for and encourage their input."

When he decided to redesign the order-entry/invoicing form, for example, he asked those who would be using it what information they needed. Then he had them proofread the printer's sample to make any last-minute adjustments. "We feel strongly that the people who are going to be using a form had better have a hand in developing it," he says.

Rule 6: The system must be enforced. "Making a decision doesn't make it a success. You've got to enforce it. That's a large part of management's job," he says, pointing to the call reports as an example. Because the reports are now easy to complete, Williams has no compunction about enforcing their use. He's even willing to withhold paychecks until reports are turned in on Friday afternoons.

Rule 7: Never stop looking for ways to improve your information systems. "We're always asking, How can we do it better?" Williams says. "This is particularly true whenever we have an angry customer. The question that comes up is, What could we do differently to avoid such a thing happening again?"

Spence now has weekly operations meetings to discuss the state of the business. "We used to hold them only when problems arose, but now we use them to nip communications problems in the bud," he says. During a recent meeting, one employee asked if the company couldn't make the order-entry forms carbonless to save time and avoid dirty hands. "Some of these things seem nitpicking, but if you multiply them over the years, they add up," says Williams.

Rule 8: Keep your systems in perspective. Spence is in a business where dependability is a highly valued commodity, and the company's streamlined paperwork systems have helped establish its credibility in the marketplace. Williams knows that consistently giving the customer fast, accurate answers goes a long way in building sales. But he is also careful not to let the importance of his systems get out of perspective. "Employees must always have the latitude to step outside the system, if that's what it takes to get the job done," he says. "After all, that's what our systems are all about -- getting the job done."