Sidebar: Riskiest Decisions

Computerworld -
Taking calculated risks is a daily part of the job for IT leaders. Here, four honorees talk about the decisions that put them out on a limb and how they learned from the experience.

• Phiroz P. Darukhanavala, vice president and chief technology officer, BP PLC, Warrenville, Ill.

As CIO of BP Alaska, I undertook an aggressive outsourcing of commodity IT functions to specialist vendors in the very early 1990s, pioneering the concept in Alaska. This entailed significant political risk, as the Alaskan government was highly focused on state employment levels and hadn't yet encountered this concept. BP's operations, on the other hand, were highly prominent and sensitive to state relationships. Commercial risk was also unusually high, as this was the largest IT operation in the state and relatively remote, and outsourcing was novel to both vendors and the BP organization. Lastly, this decision involved consideration of morale, organizational realignment and a variety of technical risks.
While not without some mishaps and project ups and downs, this program dramatically contributed to the business -- both politically and commercially. By working closely with vendors to create new supply capability in Alaska, we were able to not only maintain jobs in the state, but also initiate an IT services industry not previously present in Alaska. Moreover, outsourcing largely accounted for a significant reduction in IT costs over three years, with maintained or improved service levels.

• Sherry A. Aaholm, senior vice president of express and freight solutions, FedEx Services, Memphis

At another company, earlier in my career, I made the decision to build logistics applications in client/server technology instead of traditional mainframe. That decision required us to work through network performance, database performance and customer-integration issues. The less risky path would have been to develop a mainframe-based solution. However, at that time, to compete in the market, a key selling feature was the user interface. Traditional mainframe interfaces were blocky and clumsy, while client/server interfaces were much more user-friendly.
We learned a great deal from that decision, and it was a positive selling point in attracting logistics customers. But it took a different set of skills to manage the successful development of this client/server technology after working with the traditional mainframe approach.

• Jerry Bartlett, vice president of application development and quality assurance, Ameritrade Holding Corp., Columbia, Md.

By far, my riskiest technology decisions have been centered on the Ameritrade/Datek merger. Instead of selecting a single company's suite of integrated systems, we decided that a hybrid approach would result in a shorter timeline to perform the client account