Skip to content
← All writing

What an AI audit actually produces

5 min read

Rows of labelled electrical connectors, each block coded with its own reference

Not a slide deck of possibilities. A ranked list with a cost against each line and a price against each phase.

Most audits produce a document that describes what a company could theoretically do. That is a survey, not an audit, and it is why the word has a bad reputation among people who run operations for a living.

Three things, and they are all arithmetic

A ranked opportunity matrix: every workflow found during the interviews, scored on what it costs now and what it would cost to fix. A payback estimate against each line, built from the company's own numbers rather than an industry benchmark. A phased roadmap with a price against each phase.

The ranking matters more than the list. Any consultant can produce twenty ideas. The useful output is the one that says which of the twenty to do first and why the other nineteen wait.

Why the interviews are with the wrong people, deliberately

Leadership knows what the business is supposed to do. Dispatchers, coordinators and technicians know what it actually does, including the workarounds nobody has written down. Both get interviewed, and where the two accounts disagree is usually where the money is.

The findings get walked through before anything is scoped, with no deck in the room, and the client reorders them. Sequencing belongs to whoever has to live with it.

Photograph: Pavel Danilyuk / Pexels

A 30-minute call. Bring your numbers.

Bring your job numbers and we'll tell you where AI is likely to pay back in your shop — and where it isn't. No deck, no slides, and nobody promising you visibility in six months. If we build, the audit fee comes off it.