What an AI readiness assessment actually covers

“AI readiness assessment” has become one of those phrases that means something different in every proposal. Some firms hand you a maturity score out of five and a slide deck. Others use it as a discovery call for a much bigger project.

We run these for businesses every week, so we can tell you what a useful one actually involves. It’s a short, fixed-scope piece of work with a specific output: a ranked list of use cases, the risks for each, and a plan for the first one.

What we look at

The assessment has four parts.

Your workflows. Where does work actually happen, and where does it slow down? We talk to the people who do the tasks, not just their managers. The best use cases come from the people who know the work.

Your data. Is the information an AI system would need written down somewhere it can read? Is it reliable? Can we reach it, or is it locked in a system with no API? This is usually where ideas live or die.

Your risk appetite. What’s acceptable for a first project? A business that handles payments or health data scores ideas differently from one that schedules internal meetings.

Your governance. What policies already exist, and what would a new AI system need to satisfy? For most regulated businesses this means ISO/IEC 42001 and the data rules that already apply to you.

What you get at the end

Three things, in plain language.

A ranked shortlist of use cases, scored the same way we describe in Choosing your first AI use case. A risk summary that names the data, privacy and security issues for each idea, before anything is built. And a pilot plan: scope, team, and a written measure of success for the first project.

That last one matters more than it sounds. Most AI projects don’t fail technically. They fail because nobody agreed what “working” meant before they started.

What it isn’t

An assessment isn’t a technical audit, and it isn’t a build. It won’t produce working software, and a good one won’t sell you a large project on the back of it. It should leave you with a decision you can make on your own, even if that decision is “not yet”.

A self-check you can run today

If you want a rough version before committing to anything, try this. List every repetitive task that touches text or documents. For each one, write down who does it, how often, and what a mistake costs. Then score each on value, effort and risk. The method is the same one we use, and it’s in our first post.

Most teams can turn their forty ideas into a shortlist of three in an afternoon. If you’d like us to run the fuller version with your people, our AI Readiness & Risk Assessment is a fixed-scope engagement that starts with a free call.