When automation is worth it (and when it isn't)

The fastest way to waste money on automation is to automate the wrong thing first. The flashiest task is rarely the highest-leverage one. Here's the framework we use with clients before writing a line of code.

Three questions that decide it

• How often does it happen? Daily beats quarterly, frequency compounds the savings.
• How rules-based is it? If a human follows the same steps each time, a machine can too. Judgement-heavy work is a worse fit.
• What does an error cost? High-volume, low-stakes tasks are the safest place to start.

Score a task high on all three and it's an obvious win. Score it low and you're often better off leaving it manual, or redesigning the process before automating it.

Start with the boring, frequent thing

Reporting, data entry, follow-ups, reconciliation: the unglamorous work that eats a few hours every week. It's predictable, it's high-frequency, and nobody enjoys it. Automate that first; it pays for the whole engagement and frees the time to tackle the harder, judgement-heavy work later.

If deleting a step entirely would help more than automating it, delete it first.

Automation amplifies whatever process you point it at. Point it at a bad process and you get bad outcomes faster. The cheapest optimization is often removing a step, not speeding it up.