Two Weeks As A Software Estimation Rule of Thumb?
- by Todd Williamson
I saw a blog posting that spoke to me: http://james-iry.blogspot.com/2010/10/how-to-estimate-software.html
Oddly, this is the kind of estimate that I tend to do on smaller projects. Just about everything is "two weeks" as that is comfortably far enough out.
I once had an instructor walk us through how to create a more detailed estimate, wherein we already had the requirements up front, etc. and even after all the careful tabulation and such the final instruction was "Now that you have all this documentation go ahead and double it."
Agile practitioners seem to like two weeks also as a sprint length. Is there something magical about two weeks? Is it a hrair number for our psyches or some other kind of crutch? Do you have an immediate default fall-back schedule strategy when you are pressed for an initial delivery date?