By John Morrison on Sep 02, 2007
What's the "best" technique to estimate the time it will take to do something .. especially when it comes to Testing Software, this is one of the questions that gets asked quite a bit ... so whats the "best" answer to give to this question ?
Depends ... a likely and quite a workable + stick-able answer at times, might seem to be to go and actually ask the person who will be doing the actual testing. Well, this is the first step .. once you have a reasonable estimate from the actual person who will be doing the work and who would also be held responsible for meeting the deadlines, the next step is to address any mis-match in expectations from what the Manager "thought" it should take and what the Engineer has expected it to take.
In case of gross discrepancies, rather than dismiss the engineer's estimate, the recommended practice is to understand why and all of the variables involved in arriving at the Engineer's estimate. This could also be an opportunity for the Lead / Manager to provide any advice (only if needed) on tweaking the processes used to arrive at the estimate as well as ... getting to really know what it takes to get the work done !
Ultimately, it "helps" to get inputs from - the person with the greatest knowledge and the the person who is most likely to face the music if the estimates are not met.
So, is this the "best" technique to make estimates ? ... hmm ... let me put it this way ... think of it as one more weapon in your arsenal and use it appropriately to your advantage.
| Slashdot It! | Submit to del.icio.us