What to automate first — a decision framework

Three numbers

Deciding what to automate first is usually treated as a matter of judgement. It does not have to be. Score every candidate workflow on three axes and the ranking falls out on its own.

Frequency is how many times a week it runs. Fragility is how often it goes wrong or produces something that needs correcting, on a one-to-five scale. Hours is the time it consumes per run. Multiply the three. The workflow with the highest score is your first sprint, and it is very often not the one anybody nominated.

Why fragility carries the most weight

A task that runs constantly but never breaks is cheap to keep manual — people get fast at it. A task that runs less often but fails unpredictably is expensive twice over: once for the work, and again for the checking that the failure forces everyone else to do.

Rescore quarterly. Automating the top item changes the shape of everything below it, and the second sprint is rarely the workflow that ranked second the first time around.