Estimating Duration, Slack and Risk
My work experience informs me that the loss of slack is a big risk to projects. Without this wiggle room, we reduce our probability of success. Projects scheduled out to the last available date just do not work. The reality is these are not manufacturing or routine tasks and jobs. Even the rather routine tasks within the project have variation. One of the tools sometimes used for estimating duration of task is the PERT analysis. This technique attempts to force a normal distribution for three estimates for a task. We will use the Pessimistic, Optimistic, and Most Likely estimates to understand something about the range of distribution. Check out that technique. A better technique still can be Wide Band Delphi or the much more succinct Agile variant Planning Poker and proportional estimating.
Crashing, throwing resources at the project seldom result in the desired results. Our costs will go up, and the quality and deliveries still compromised. Hiring on new people or moving people to the project takes time and those already in the project must take time to get the new comers up to speed.
Imposed dates are interesting. There are times when time or duration is hand down from on high for valid reasons. However, just because someone says the need is at that date, does not mean it is possible to achieve that date. Consider this analogy. I walk into work and tell everybody that I want to take a day trip to Mars. I want the trip to take 2 hours to get there and 2 hours back. I will walk around on the surface for 4 hours and be back in time for dinner. There, I have asked for it. How likely will it be there tomorrow when I go into work? It is not likely, not probably, not possible. Here is where communication and expectation management enters the picture. It is important for the project manager to articulate what is at risk, what may be possible, and what is not likely. Mincing words or couching our meaning not wishing to deliver “bad” news is not helpful. In fact, anything we do that delays the conversation has two problems. First, we are losing time on finding a real solution. Secondly, we set the project sponsor expectations that the goal or objective is achievable. I have seen many project managers’ say things similar to “we will do our best” when the entire team knows there was no possibility of success as defined by the variables at hand.