When will this be done?
I don’t know yet.
When will you know?
I don’t know that either.
Just guess how long it’s going to take.
Um, two weeks?
Ah, the magic estimate of two weeks. Just far enough out that maybe we could get it done by then, but not so far out, that anyone is concerned. Worst case, we can just slip another two weeks if we don’t make it this time.
There are some things where we can accurately estimate the amount of time they will take. How long will it take to do the dishes? How long will it take to get to the grocery store? How long until it’s 5pm? 🙂 These tend to be non-creative endeavors that require no research or necessarily though to accomplish.
Creative work, whether it’s writing software or making art, thwart predictive schedules. Unless the task is simple with a known solution, we struggle to pin down exactly how long it’s going to take us. We might guess, but often find that our best guess is wildly inaccurate.
My project leader once asked me to do something and I thought, oh, that’s really hard. That will take me, um two weeks. “Two weeks?!” he replied. “Well, it’s really important so I guess we should just do it.”. On my way back to my desk after talking to him, I realized that actually there was an easy way to do it and was back telling him it was done 30 minutes later. This doesn’t mean all my projects were like that. Many projects follow the general rule of thumb for software schedules: Take the amount of time estimated, double it, and round-up to the next unit of time — 2 hours becomes 4 days.
Inaccuracy in either regard leads to inefficiency for those depending on the deliverable. But we fight against living with the uncertainty of not knowing. And so we press to have a schedule for something that relieves the uncertainty without actually providing us any additional amount of predictability. The best approach in situations like this is to recognize that there is some unknown-unknowns and try to break the project into next steps and then schedule those.
The next time someone estimates two weeks raise an eyebrow, but also question whether creating such a schedule is really in the best interest of everyone involved. We’re not just making widgets anymore.