Johanna Rothman, known as the "Pragmatic Manager," offers frank advice for your challenging problems. She consults with leaders and teams to help them learn about practical and possible options. They can then decide how to adapt their product development. Her most recent book is "Project Lifecycles: How to Reduce Risks, Release Successful Products, and Increase Agility." See www.jrothman.com for all her books.
Project team members produce good estimates and bad estimates, so how do you, the project manager, come up with the most realistic estimate?
Not everyone is a great estimator. Some people are over-optimistic and underestimate by as much as 50 percent for any task. Some people are pessimistic and add buffers for every single task. Some people estimate small tasks up to three or four days well but can't estimate anything that's longer than a week in duration. What's a project manager to do?
First, know your team. Decide how much feedback you want to give each person about estimation. You don't have to solve all these estimation problems the first week of the project.
Next, eliminate any extra time buffers on each task. Ask each person whether he or she has estimated with or without buffers. Explain that you're not trying to reduce the task time but that you want to make sure you have the most accurate estimate.
If you're using a serial or iterative lifecycle, consider using Theory of Constraints (TOC) to deal with the estimates. In TOC, everyone is supposed to provide you with a reasonable estimate for the task. You take the 50 percent mark and make the task that long in the Gantt chart, adding 25 percent of that original estimate into the buffer. When a task in the critical path requires more time,