How to Make Your Estimates 99.73% Accurate. We have seen many times Project Managers blindly trying to estimate Costs and Time of a project which has been poorly defined, randomly establishing numbers that probably never will be true.
But if these Project Managers give those estimates, they will have to compromise for bigger problems soon afterwards.
They must realize that the Project can be badly defined, but this is no reason to give poorly defined estimates. If there is little information about the requirements, the Project Manager must define the assumptions and document them. Then, estimate costs and time for those assumptions. The documented estimates must always be referring to those newly defined assumptions. Now the compromise is only for that specific Scope. If the Scope changes, obviously, the estimates do so as well. So, in this way, the Project Manager will always have the Customer as his friend, having fixed the Scope by the compromised Cost/Time estimates.
To address the uncertainties and risks when estimating (based on the latter assumptions), Project Managers shouldn’t invent either of the contingencies (as many do by simply adding a percentage), and they must refer to the Standard Deviations. For this, use the Three-Point Estimates.
The Three-Point Estimates is easy to use and can be applied either for cost or for time. Ask, for each project activity, three estimates: the most likely (tm), the optimistic (to), and the pessimistic (tp). Then apply PERT to calculate the expected value: te = (to + 4tm + tp)/6. The Standard Deviation will be ts = (tp – to)/6.
If you calculate your contingencies by adding +/-3 ts to your te, you will be statistically 99.73% accurate.
It seems like magic doesn’t it? (We started off with a poorly documented project and we are now 99.73% accurate). No, this is simple Project Management methodology, first documenting assumptions and then applying Standard Deviation.
Have you ever had to estimate a poorly defined project? What did you do?