After years in I.T., Niel Nickolaisen believes IT projects fail in such high percentages because both IT and business executives spend too much time and energy debating systems that ultimately don’t generate a single dollar of ROI or the slightest increase in market share. And then they devote too many resources to developing technology to support those projects.

Nickolaisen joined Deseret Book, a book publisher, distributor and retailer based in Salt Lake City, as its first ever CIO in March 2003. He found an IT department in chaos. "They had a highly customized homegrown system. The IT development staff spent its time responding to every imaginable request," Nickolaisen says. "They didn’t have a CIO, IT reported to finance, and there were business processes embedded in their legacy systems that actually got in the way of people getting things done." So Nickolaisen, who was then a business turnaround specialist at a venture capital firm, was brought in to overhaul Deseret Book’s 20-year-old technology infrastructure and bring order to a random IT project prioritization process.

The goals of his approach are to help Deseret Book make more effective IT project decisions and to bring a sharper focus and rigid analysis to IT project decision making. This method helps Nickolaisen and his executive colleagues identify business priorities around each decision and allocate the appropriate amount of resources?staffing, capital investment and time?to each project. By prioritizing and classifying these decisions, the CIO and business unit leaders can examine them more objectively. They avoid squandering intellectual and financial capital on processes that, while important to the business, would not help the company generate additional revenue or secure greater market share.

Using this mode of classifying projects has yielded tangible benefits, according to Nickolaisen and has greatly simplified the company’s project strategy. He estimates that using this decision model has reduced IT project time lines, costs and resource requirements by 40 percent to 70 percent.

Rewriting the Rules

One of the first major processes that Nickolaisen classified with his model was a project overhauling the Deseret Book system for shipping customer orders. The rules the company had encoded into its legacy systems for shipping products exemplifies the ad hoc manner in which the company had historically made IT and process decisions. Besides its 36 retail stores in the western United States, Deseret Book sells directly to customers through three channels: catalog orders, Internet orders and its book club. Since those three channels evolved at different times, each channel used different shipping rules and fees.