Executive involvement leads to IT success
W. Wayt Gibbs, writing back in 1994 for Scientific American, delivered a landmark article on software development, 'Software's Chronic Crisis.' The article summary stated that 'despite 50 years of progress, the software industry remains years'perhaps decades'short of mature engineering discipline needed to meet the demands of an information age.'
Gibbs provided some alarming statistics: 'For every six new large-scale software systems that are put into operation, two others are canceled. The average software development project overshoots its schedule by half, larger projects do generally worse and some three quarters of all large systems are operating failures that either do not function as intended or not used at all.'
The software industry has made progress since 1994, but government projects are still fraught with anxiety, especially when information technology managers face scrutiny over slipped schedules, blown budgets or flawed results. If anything, the increased risk has made quality management software engineering programs more important.
In Georgia, we have just completed Phase 1 of a large enterprise resources planning (ERP) project'and accomplished it against all odds. We successfully implemented a PeopleSoft Inc. financial system statewide in 16 months.
I am proud of my team and want to brag a little. But mainly I want to share one of the uncommon characteristics of this project that I feel contributed to its success: executive involvement. Georgia had settled on a PeopleSoft client-server configuration running under Unix with an Oracle Corp. database. The system would be carried statewide via a frame relay network. We called it Phoenix. Phoenix was to replace Georgia's existing financial system, which would have been too expensive to make ready for year 2000. The implementation deadline was July 1, which was also the expected failure date of the existing system.
Every two weeks for 16 months, the Phoenix Executive Committee met with the development team and an audience of agency users. The committee's purpose was to review the project plan and to make design and development decisions in real time. Committee members included the state auditor, the governor's budget officer, the head of the Legislative Budget Office, the head of the Department of Administrative Service'which housed the development team'and the head of the state personnel system.
These leaders took this job seriously, learning the language of development sufficiently to answer design questions, establish business rules and generally be there when they were needed. The executive involvement was contagious; everyone knew each step was important and each person would be held accountable for his or her component.
Government executives do not typically establish oversight committees with anywhere near the interest or the authority of this project'and for good reason. Their time is limited. Too many projects are going on simultaneously. It just isn't realistic.
Or is it? With ERP systems being so dependent on well-defined business rules, having top decision-makers at every step reduced uncertainty and streamlined the design process. Our effort proved the maxim that an organization does well those things that the boss personally checks.Mike Hale is chief information officer of Georgia. He previously was executive director of Florida's Information Resource Commission and is a retired Army colonel. His e-mail address is firstname.lastname@example.org.