AI readiness checklist for UK SMEs
Ten yes/no questions across the seven dimensions that actually predict whether your first AI project will work. Score yourself in five minutes.
0 of 10 ticked
The seven dimensions of AI readiness
Most AI projects in SMEs do not fail because of the AI. They fail because of missing prerequisites. These are the seven we score against.
Data hygiene
Can you find your customer and order data without a treasure hunt? Perfect data does not exist anywhere, but you do need to know where it lives and roughly trust it.
Process maturity
AI works on repeatable patterns. If a task happens regularly and you could write a recipe for it, that is a candidate. If every job is bespoke, the quick wins are harder.
Team capability
You do not need a tech team. You do need a team that is broadly open to new tools when those tools make their work easier.
Tooling baseline
Modern SaaS systems usually offer integrations. Software you log into through a browser is usually fine. Software that lives only on one laptop is usually not.
Leadership engagement
MIT Sloan research consistently finds leadership engagement is the single strongest predictor of AI success in smaller organisations. One champion with authority beats a committee every time.
ROI clarity
If you cannot describe what 'better' looks like in numbers, you cannot tell whether the pilot worked. Spend a week tracking time before you spend a month building anything.
Risk posture
UK GDPR and the ICO's AI guidance apply to anything that processes personal data. A short Data Protection Impact Assessment is normal, not exotic, and your consultant should handle it.
What each score band means in plain English
0 to 3 -- Not ready yet
You are not quite there, and that is good information. Pick the two or three weakest dimensions and fix those first. Usually that means documenting one process end to end, naming a clear owner, and writing down what ‘better’ looks like in numbers. Retake in a quarter.
4 to 6 -- Pilot-ready
The most common band. You have the prerequisites for a useful first project. Pick one painful, repetitive task, scope a pilot of 4 to 8 weeks for under £8,000, and run it alongside the existing process. Most of our first engagements start here.
7 to 10 -- Ready to scale
Multiple parallel projects make sense at this stage. Sequence them so wins land at different times and so different teams get visible benefits. The risk now is spreading too thin, not technical readiness.
Frequently asked questions
Who is this AI readiness checklist for?
UK small and mid-sized businesses considering their first AI project. It is calibrated against the prerequisites that actually predict success or failure in this size of business, not enterprise rollouts. If you are a 5 to 250 person company, this will tell you something useful in 5 minutes.
How is the AI readiness score calculated?
Each ticked item is worth one point. Scores fall into three bands: 0 to 3 means you are not quite ready, 4 to 6 means you are pilot-ready, and 7 to 10 means you are ready to run several projects in parallel. The bands map directly to the readiness research used in our paid Free AI Opportunity Report.
Is this the same as the Free AI Opportunity Report?
No. The checklist is a fast self-score across ten yes/no questions. The Free AI Opportunity Report is a written analysis specific to your business. You answer four longer questions about your operations, and we send back a personalised report with prioritised opportunities, estimated savings and a recommended starting point. The checklist tells you whether you are ready; the report tells you what to do.
What if my AI readiness score is low?
A low score is the most useful result. It tells you which two or three things to fix before you spend money. Usually it is one of three patterns: nobody owns the project, processes are not documented well enough to automate, or no one can describe what 'better' looks like in numbers. Each of these is fixable in weeks, not months. Fix them and the second pilot lands.