I automated the wrong thing first: measure before you automate

The first thing I ever automated was the task I hated most.

Wrong instinct. I hated it because it was tedious, not because it was frequent. I spent a week building something that saved me about twenty minutes a month.

The thing I should have automated was boring enough that I never minded doing it, which is exactly why I never noticed it was eating four hours a week in fifteen-minute pieces.

Tedium is loud. Frequency is quiet.

I still catch myself reaching for the loud one. Most AI tooling on the market right now is sold against the loud one, because a demo has to make you feel something in ninety seconds, and “this saves you fifteen minutes, forty times a month” does not.

Measure first. The thing worth automating is almost never the thing you complain about.

Originally posted on LinkedIn.