If a team member can’t seem to give you a good estimate of the time it will take to complete a task, they might not understand the work well enough. Call upon other experts (internally) to get their opinion. And research (OPA’s and externally) the blind spots and/or areas your stakeholders might be turning a blind eye to.
Here is my takeaway #25 from one of the most useful practical books in my field, “Project Management for the Unofficial Project Manager”:
25) “…find the sweet spot when estimating a duration of a task. Learn how to be accurate with duration. There is no magic bullet for this, but here are a few suggestions: 1) Draw on your own experience. 2) Ask a reference. 3) Get advice from an outside expert. 4) Use the PERT formula to figure out how long each task will take…Expected time = (Optimistic + 4 x Most Likely + Pessimistic) / 6″
Note: This comes from Chapter 4 “Planning the Project: Milestone or Mirage?” (pages 104-105)
You need to realize that projects have something unique about the work that has never been done before or else they wouldn’t be projects. No one will know exactly how long it’s going to take until they do the actual work. Measure it. Record it. Use it for the next estimation of a similar project. That’s valuable information!
______________
STORY TIME:
On database projects I worked on for a few different companies, we didn’t have baselines at the outset for how long it took to catalog data. The problem was depending on the data being cataloged the tasks took a different amount of time. The data management could be put into a few categories: original cataloging (from scratch), cataloging (from new source information and new standards for records that needed to be uploaded into a new format) and copy cataloging (from surface updates needed to a prior record). Besides that, the content of the data varied greatly between types of records for different or uniqueness between products as well as types of classification needed (cosmetic, minimal, or major data cleanups).
In each instance, I used the PERT formula to calculate each of the overall length of the project, work activities, and tasks. My guesses usually underestimated the work and durations like everyone else. Most of us tend to be overly optimistic about seemingly easy, but time-consuming work. With more research I actually arrived at fairly accurate estimates of how long it would take to perform these labors. I ended up tracking time to measure baselines for the work, however, because these more detailed indicators of what I was doing was more believable to persons we were reporting back to.
Estimating similar work got a lot easier from then on out, even if the higher-ups never could believe how long detailed work takes.
______________
Well, that was boring. This is one reason why data analysts are a project manager’s best friend. They can help you decipher data in so many ways. And it can help you estimate time on projects more accurately. Seems like businesses want to know how much time and money will go toward project efforts more than any other concern they have.
The closer you can get to getting it right the happier everyone will be. Factor in everything you know, be clear about what you don’t know, and provide a buffer for it in the schedule (Note: The PERT formula and other methods attempt to account for some unknowns in the equation, but you may never fully pin down some creative things like “original” cataloging.).