The Automation Readiness Checklist

Automation pays off on work that is repeatable and rule-driven. If a step changes shape every time, it usually needs a person — or a rethink — before it needs software.

• The process follows the same broad steps almost every time it runs.
• You can write down the rules a competent new hire would need to do it.
• Most cases are routine, and the genuine exceptions are a small minority.
• Success and failure are objective — you can tell when an output is wrong.

A system can only act on inputs it can actually reach and trust. Missing, messy, or locked-away data is the most common reason an automation stalls after launch.

• The data the process needs arrives in a consistent, machine-readable form.
• The source systems expose an API, export, webhook, or shared inbox you can connect to.
• Someone owns each input and can answer questions when a record looks wrong.
• You have permission to use this data for an automated workflow.

Effort should follow impact. The best first candidate is high-volume or high-cost enough that a reliable system clearly earns back the build — without being so high-stakes that an early mistake is unacceptable.

• It runs often enough — daily or weekly — that small time savings compound.
• The manual version is slow, error-prone, or pulls senior people off better work.
• A wrong output is recoverable and caught before it causes real harm.
• You can name the outcome you want: faster turnaround, fewer errors, freed-up hours.

A system that no one owns drifts out of date the moment the process changes. Readiness includes the people and habits that keep it running after launch.

• Someone on your side will own the system and notice if it stops behaving.
• There is a clear path for handling the exceptions the system escalates.
• You can tolerate occasional changes when upstream tools or rules shift.
• Leadership agrees this is worth automating now, not just in theory.