It’s well-established within the agile world that delivery teams—and particularly individuals—are terrible at estimation. We do estimate, still, and hopefully we get better with time, comparing forthcoming work with past examples, for instance. But it often remains a wonder to those outside the immediacy of delivery teams just how misjudged our estimations can be.
“It’s software,” is often the explanation, “It’s very complicated. No-one has done this thing before.” But perhaps it’s not always a software thing.
A short while ago I needed to build a flat pack cupboard. All the pieces were in the box. I had the instructions in front of me, and they were very clear. They were almost exclusively pictures, and I read them carefully. I had all the tools I required. “Looks easy enough,” I thought. I estimated 60 minutes. It took 120 minutes.
It was indeed easy enough. Nothing went wrong. I had all the tools, the instructions were set out for me, and all I had to do was follow them… and I did. And yet I was still out by a factor of two.
Sometimes even everyday estimation can be very poor. It can be a good reminder when we’re discussion with people outside our delivery teams looking for certainty.