Editor’s note: This article was originally published April 8, 2001.

The dilemma

I hadn’t talked to Matt since I helped him create an estimate for converting all our company’s databases to a new version. When we caught up recently, he was both happy and concerned.

“The estimate we prepared was just what my manager was looking for,” Matt said. “Fortunately, she has the budget approval to begin the work, but she said that the budget is very tight and that we would have to complete the work within our original estimate.”

I asked when he would begin planning for the project.

“We are just starting the planning now,” Matt replied. “That’s why I came to see you. I know we put together a nice estimate based on a model, but it was still just a model. We haven’t done any conversions yet. If the actual effort is more than we estimated, we may be substantially over budget. That could put the entire project in jeopardy.’

The solution

Matt was beginning to see how he should proceed. “I think the first thing we need to do is validate the estimates,” he said. “What’s the quickest way to do that?”

I noted that the quickest way isn’t always the best way. I recommended that Matt conduct a short pilot project. “First, identify a representative sample of databases — large, medium, and small,” I told him. “Migrate them to the new release and keep track of the effort and duration for each one. Then you’ll have some factual information to apply to the estimates — and you’ll know if you should break out the champagne or the aspirin.”

Mentor advice

One of the common aspects of software development is that the solutions tend to be unique. However, you’ll sometimes find that the work efforts involved are similar or repetitive. For instance, you may have to implement that unique solution repetitively in different offices, different departments, different Web sites, and so on. With proper planning, using a pilot project in such instances can give you the chance to:

  • Document common procedures that can be utilized on the other similar efforts.
  • Focus and be successful on one project first, rather than having to start off with multiple efforts.
  • Reduce the risk of project difficulties since you can look at what works well and what doesn’t before tackling the remaining work.
  • Get a handle on the time and cost required to complete the entire effort.

In Matt’s case, the pilot project makes great sense because he has a similar process that needs to be applied to hundreds of databases. By properly planning the project, he can select a representative sample of databases, transfer this subset of databases to the new release, document common procedures, and find the average effort and cost required.

When the pilot is complete, he can compare the actual averages with the estimated averages. If the numbers are not close, Matt will need to notify his manager so that appropriate actions can be taken before the project gets too far along. If the actual migrations are close to the estimates, he can be more confident that the entire project can be completed within the budget.


Get weekly PM tips in your inbox
TechRepublic’s IT Project Management newsletter, delivered on Wednesday, offers tips to help keep project managers and their teams on track. Automatically sign up today!