For any project task, it is unlikely that your estimate of how long it will take will be perfect. So, is it worse to overestimate the time it will take to complete a task or underestimate it?
Cost of overestimation
Overestimation creates the problem that the estimate become self-fulfilling. The task takes longer than it would have done with a more accurate estimate in place. There are two ideas behind this linked to how people behave. Firstly, Student’s Syndrome states that people often won’t start working until very close to a deadline. Secondly, Parkinson’s Law states that work expands to fill the time available. Therefore, if you have a task with overestimated length, the impact is the task might take longer than it ‘should’ do.
Costs of underestimation
If a task is assumed to take long time than it actually needs, one of two things will happen. Either the task gets done at lower quality, or the task doesn’t get done on time and any tasks dependent on it are pushed out.
Which is worse?
Whilst obviously accurate estimates are the best outcome, over-estimation is less bad than underestimation. Underestimation can impact dependencies and the overall quality of the project. Overestimation may be wasteful for the resources on a particular task, but it is less likely to impact other tasks or overall quality.
Of course, a third option following the critical chain methodologies is to consider adding buffers to the schedule to allow for some underestimation at the individual task level.