Most of us who have shepherded an information system project plan to life rarely came to the job with project management skills developed through specialized training and honed from project to project. Rather, we arrived at these opportunities armed with no more than program, generalist or information system management backgrounds. We were lucky if we also brought to the table a reputation for getting things done. Unfortunately, most of us realized that "an ability to get things done" and having the "right stuff" to oversee large-scale system development efforts to successful implementation are not the same skill sets.

Some of us enjoyed success, because the right elements were in place on the right project at the right time, albeit in the absence of a proven project management methodology. Some of us enjoyed success in spite of government bureaucracy, and others in spite of limited or no technical expertise. Christine Rozek, project director for an information system for the License, Inspection and Environmental Protection Office in St. Paul, Minn., observed, "It can be an overwhelming responsibility. In the absence of technical knowledge and prior project management experience, there are no tools to support qualitative judgment and decision making as one deals with the vendor community."

SETTING A STANDARD

In August 1995, Government Technology magazine reported on the remarkable success of the Oklahoma State Department of Human Services' SACWIS project. It enjoyed unparalleled success in meeting federal timelines for the statewide implementation of the child welfare information management system known as KIDS. Taken at a high level, the project's overall management approach could create a template for other government projects. The key factors to KIDS' success were identified and reported on after the project was completed.

There were several critical factors in KIDS' success that could have led to failure -- issues that have been blamed for contributing to the failure of other projects:

* The project director's position in the reporting hierarchy;

* The lack of prior experience in managing large IS projects;

* Externally enforced time frames for completion;

* A functional rather than conceptual transfer of the base system;

* Use of new technologies;

* Use of new hardware/software architectures; and

* Significant change management issues.

These were significant negative issues to overcome. They had little impact on the KIDS project, however, because there were other elements in place of such importance that their influence overshadowed the negatives.

CAN THIS BE REPLICATED?

Of course! Other government agencies can point to success stories for many of the reasons identified in the Oklahoma project. In the same environment, however, other projects might have failed or been marginally successful. This means that, without a formal methodology to guide the project management process, the probability of success is still happenstance. It means there is a great risk that success is dependent upon coincidental factors coming together at the right time and for the right reasons. It might also mean that success is dependent upon a very capricious factor: the personality of the individuals assigned to the job at the time.

Larry Harmon, the Oklahoma Department of Human Services' associate director for administration, was a member of the KIDS Steering Committee. When asked to describe the characteristics of a good project manager, he stated, "Most leaders don't make good project managers, but a good project manager has to be a strong leader ... . A good project manager doesn't mind doing anything to get the job done."

Tom Llewellyn, the former project director for a public assistance system in West Virginia, said, "When we had a problem, it didn't matter who or what caused it. We set about determining our options for fixing the problem and establishing a process to assure that we wouldn't repeat it."