To estimate or not to estimate? That is the question! Is it better to use the No Estimates approach as a way of freeing yourself from unnecessary formalities, or are you only setting yourself up for a more complicated and less productive workflow? Glen B. Alleman sticks by his defense of estimation, while telling us how exactly a No Estimates approach works.
He talks about Vasco Duarte’s No Estimates theory called Little’s Law, which calls for users to break down work into same-sized chunks and then attack the work in a systematic manner. Using this approach, you should still have a pretty good idea of when you will be done with the project, just multiply the average time spent in the queuing system with the number of objects left.
It might not be actually estimating, but it’s still estimating. And it’s still flawed. There are still variables that need to be estimated before applying Little’s Law, and there is still the question of going over schedule or over budget. So maybe no approach to estimation is perfect. You just have to do what works for you. Maybe it’s Little’s Law, maybe it’s regular estimation, or maybe it’s no estimation at all. Whatever it is, you will find flaws, but that’s what estimation is. You can’t escape the variables, even when you get rid of estimation itself.