It's not limited to software. In the early 90s - 1990 to be exact - I open a retail store. I don't remember where I read it, but this heuristic has served me *

It's not limited to software. In the early 90s - 1990 to be exact - I open a retail store. I don't remember where I read it, but this heuristic has served me *very* well ever since: It - whatever it is - will take you twice as long and cost twice as much as your original estimate. Sure occasionally you'll do better or worst, but on average it's spot on. Or cost might not double as the raw goods (e.g., lumber) are fixed, but time is nearly always 2x'ed. Needless to say, I often get into deep (read: heated) discussion over estimates that don't use this framework.