I've been told that while we can use PERT to estimate the duration of an uncertain activity, the reason for the uncertainty doesn't go away. Can you please explain?
Uncertain in Jersey City
PERT, or Program Evaluation Review Technique, is a nice way to take into account the uncertainties in cost and/or schedule of activities and derive a figure that takes those uncertainties into account. By using the below formula, you end up with a duration (or cost) that is pretty realistic. For this example I'll use duration though the same applies to cost.
Duration = (Optimistic + Pessimistic + 4 x Most Likely)/6
So, take for example a situation on a critical project that will last only a few days where one of the activities is to drive a package to Kennedy airport from your location in Jersey city in the middle of February next year. How long will that take?
Optimistically, 40 minutes. Pessimistically, with snow on the ground at rush hour on a work day, 4 hours. Most likely, 90 minutes. Using the formula, we get:
Duration = (40 + 240 + 4 x 90)/6 = 107 minutes
So, by giving yourself 107 minutes for this activity you are likely to complete it on time. But you have not removed the underlying uncertainty behind the activity. It may still snow, there may be car accidents or rush hour, all those bad things might still occur.
The only thing you have done is taken statistics into account and provided a duration that should allow you to finish your project on schedule. Not every possible bad thing will happen so, even if it does snow for this activity, some other activity will hit the optimistic duration and you will be back on track.