A "price to win" initiative at competitor TruServ used BI to measure the impact of lowering prices in certain segments. "We gave away gross margin hoping to increase sales," says Hastie. "It didn’t increase velocity." So in a matter of three to four weeks, TruServ ended the sale. Hastie says that without BI, the company would have analyzed the data at the end of the quarter, reanalyzed it after the next quarter and continued giving away margin for six months instead of one.

At Fairchild Semiconductor, pricing its 50,000 products manually was such a chore that the company did it once a quarter. "It wasn’t necessarily consistent. And it certainly wasn’t mathematical," says Bill Hall, vice president of interface and logic. "A lot of pricing in the past had been done with gut feeling." Using a BI-powered dynamic pricing engine, prices are easily adjusted weekly, factoring in current worldwide market conditions, manufacturing cost, and capacity and current inventory. If Fairchild can’t produce enough of a hot-selling item, BI helps analyze the merits of raising the price, pinpoint where the item could be sold for the most short-term profit and identify which customers clamoring for the product are likely to generate the most long-term profit. CIO John Watkins says Fairchild spent "a couple million" on BI, but the investment has already paid for itself through better pricing alone.

Whatever uses CIOs foresee for BI, Nucleus’s Wettemann advises going for the easy payoff first. (See "Seven Rules to Rolling Out BI," Page 85.) Quaker Chemical’s Tyler concurs. "Build a simple infrastructure," he says. "You can always do more, but you can never do less. There’s a tendency to do something fancy." It’s better to build something simple, get it into users’ hands, and let them tell you what more they need. "Don’t deliver reports," he says. "Deliver tools."