Steve Smith describes the company's technology base as a crazy quilt of custom-built software and old computers sewn together by a parade of consultants. "It was like having multiple tailors making you a suit," he says. "One arm looked good, but when we put the suit together, we were just barely clothed. We were strangling ourselves on labor, on duplication of effort--spending all our money having three and four people doing different portions of a process."
Ironically, when it came to manufacturing technology, Tec Labs was a paragon of precision. The company practices "constraint management," an exacting procedure in which employees log how long it takes each bottle of Tecnu to pass through each step in the manufacturing process. The goal is to identify sticking points, measure their cost in production time, and then translate that into labor and revenue costs. Consequently, if a labeling machine begins to stick, Smith knows just when it is economically smart to replace it.
Tec Labs' manufacturing operations supported the company's business goals: to continually improve processes and boost employee productivity. Its information systems didn't provide the same kind of support. Then one day in 1996, "I realized that we could take the same processes we use in manufacturing and apply them to our computers," Smith recalls. That meant compiling extensive metrics on how technology was serving employees and where it was failing them. With the idea of mimicking the manufacturing process, technology director Kathlene Cowan asked staff members to begin logging the time they spent on computer-related tasks and then extrapolated those hours out to a year.
The results were eye-opening. Companywide, employees were spending 105 hours annually running to printers, 52 hours converting files from one operating system to another, and 75 hours in unnecessary document management. By Smith's calculations, the company was wasting $40,000 in labor every year.
With a clear view of Tec Labs' goals, a meticulous accounting of its weaknesses, and an eye toward what was already running at the company, Cowan began evaluating technologies. As she did so, she quantified for Smith the impact of each potential expenditure on Tec Labs' bottom line.
Ultimately, Cowan persuaded the CEO that a Microsoft Windows NT network would solve most of the productivity problems while also creating some new opportunities--allowing salespeople dial-in access from the road, for example. She calculated the cost at $20,000--half of what Tec Labs had been squandering on inefficiencies. "The ROI estimate made it easy to sell the idea to the board," says Smith.
Today the network is in place, and so is an ongoing planning process. Whenever a department or a group wants to add or upgrade something, it submits to management a formal wish list that includes a detailed breakdown of costs and estimated returns. With that information "the board can use money like a scalpel, rather than a bulldozer," says Smith. "They can look at all the multiple, complex projects we're considering--say, a new advertising campaign versus a new piece of machinery--and make good decisions as to which ones should come first."
Unlike Tec Labs, Minneapolis-based Rosemount Office Systems Inc. wasn't about to let the lack of a plan slow its technology spending. Two years ago the 175-employee builder of customized office furniture discovered that its enterprise resource planning (ERP) system--software products that link functions in manufacturers--was vulnerable to the millennium bug. "We were convinced we should fix our Y2K issue as soon as possible and basically thought that it would be easy," says vice-president and chief financial officer Mary Rolf.
Six months, $80,000, and "a lot of blood, sweat, and tears" later, the old system could handle the Year 2000--but not the company's operations. "After the fix was all done, we found out that the system we'd fixed no longer met our business requirements," says Rolf. In Rosemount's case those requirements were chiefly dictated by a new business plan that stressed customer service, especially the speedy turnaround of orders. Unwilling to leap again before taking a serious look, Rolf hired Cassidy of Strategic Computing Directions to help create a technology blueprint that would both mirror and extend the business plan.
Cassidy's recipe is fairly simple but demands unflinching rigor. The first step is to translate every business goal into a corresponding technology goal. For example, Rosemount's mission statement includes "customer delight" among its business objectives. In tech-plan-speak, achieving "customer delight" means "designing systems so that customers can access information easily and directly."
With goals defined, Rolf and Cassidy set about scrutinizing Rosemount's processes in order to assess its information needs. Everything went on the table: financials (monthly sales, order backlog, gross margin, operating margin), human resources (salaries, benefits, turnover rates, sick days), manufacturing (scrap levels, reworked items, material expenditures, total yield), and marketing (pricing, discounts, regional sales, special offers). "The combined data told us how we were doing now, compared to where we wanted to be in providing quicker, cheaper, and higher-quality products and information," explains Rolf. "It also helped us figure out how we stacked up to the rest of the industry. If we compared favorably, we asked ourselves, 'Is this good enough?' And if we didn't, we asked, 'Why can't we be better?"
Until that point, specific technology tools hadn't even entered the conversation. But armed with quantifiable objectives, Rolf and Cassidy asked such questions as "How do we stay in business with our current systems?" and "What do we need to be technologically competitive?"
