How to Run an AI Readiness Assessment for Your Business

An AI readiness assessment is a structured check of whether your business is set up to adopt AI, scored across six things: strategy and leadership, people and skills, data, tools and technology, process and workflows, and governance. Score each from 1 (not ready) to 3 (ready) for a total out of 18, which tells you whether to fix foundations, run a focused pilot, or scale. You can do a rough version yourself in an afternoon. Below is the framework, what each score looks like, and the gap that catches most small businesses out.

Quick links

• What is an AI readiness assessment?
• Readiness vs maturity: what is the difference?
• The six dimensions to score
• How do I score it and what does the total mean?
• What is the most common SME readiness gap?
• FAQ

What is an AI readiness assessment?

An AI readiness assessment is a structured way of answering one question: if you started adopting AI tomorrow, would it actually work, or would it stall. It looks at the foundations rather than the tools.

It matters because adoption is no longer the hard part. The British Chambers of Commerce found 54% of UK firms are now actively using AI, up from 35% a year earlier, with around 94% of those surveyed being SMEs. The ONS puts it lower at 23% of all UK businesses using some form of AI in late 2025, depending on how you define use. Either way, the trend runs one direction. The harder question is no longer whether to adopt AI, but whether you are ready to get anything out of it.

That is what the assessment is for. It is a diagnostic, not a sales pitch, and done honestly it usually tells you to fix two or three unglamorous things before you spend a penny on tools.

Readiness vs maturity: what is the difference?

Readiness asks whether you are set up to start. Maturity asks how far along you already are. They get used interchangeably, and they should not be.

An AI readiness assessment checks the foundations: do you have a clear use case, the skills to use the tools, data you can actually feed an AI, and a few sensible rules. It is the right starting point for almost every SME.

An AI maturity assessment measures depth: how widely AI is embedded, whether it is scaled across functions, how well it is governed. Maturity is a question for businesses already a few years in, and most maturity models score on a 1 to 5 scale across dimensions much like the readiness check below.

If you are asking “are we ready”, you want readiness. McKinsey’s 2025 survey found 88% of organisations using AI but only 6% counting as high performers genuinely capturing value. The gap between using AI and getting value from it is mostly a readiness gap.

The six dimensions to score

Six dimensions cover the ground that decides whether AI works for a small business. For each one, here is what not ready, getting there, and ready look like, and the gap SMEs tend to fall into.

1. Strategy and leadership

Does someone at the top actually own AI, with a reason for it that ties to the business.

• Not ready (1): No one owns it. AI is “something we should look at” with no link to a business outcome.
• Getting there (2): A leader is interested and there is a rough sense of where AI might help, but nothing is written down.
• Ready (3): A named owner, one or two specific problems AI is meant to solve, and a budget that matches the ambition.

Common SME gap: AI gets delegated to whoever is keenest, not whoever owns the outcome. Enthusiasm is not strategy.

2. People and skills

Can your team use the tools well, and do they want to.

• Not ready (1): A couple of people quietly use ChatGPT. Most have not touched it. No training, no shared knowledge.
• Getting there (2): Broad curiosity, some informal use, but quality is patchy and nobody is sure what good looks like.
• Ready (3): People across the business use AI for real tasks, share what works, and know its limits.

Common SME gap: confusing access with skill. Everyone has the tools. Far fewer know how to get a reliable result from them.

3. Data

Can an AI actually reach and use the information your business runs on.

• Not ready (1): Data is scattered across inboxes, spreadsheets and someone’s head. Quality is unknown.
• Getting there (2): Core data lives in a few systems and is mostly accurate, but it is not joined up or easy to pull.
• Ready (3): Clean, accessible, reasonably organised data that a tool could draw on without a six-month project first.

Common SME gap: assuming you need a data warehouse. You do not. You need your existing data to be tidy enough that AI is not working from a mess.

4. Tools and technology

Do the systems you already have play nicely with AI, or fight it.

• Not ready (1): Legacy software, no integrations, nothing connects. Adding AI means bolting it on the side.
• Getting there (2): Modern core tools with some AI features you have not turned on, and a few integration headaches.
• Ready (3): Systems that connect, with AI either built in or easy to add through existing integrations.

Common SME gap: buying a shiny new AI tool that cannot see any of your other systems, so it never gets used.

5. Process and workflows

Do you understand your own processes well enough to know where AI would help.

• Not ready (1): Processes live in people’s heads and change depending on who is doing them.
• Getting there (2): The main workflows are documented or at least understood, and you can spot a few repetitive, rules-based tasks.
• Ready (3): Clear processes, and you have already identified the specific steps where AI would save time or reduce errors.

Common SME gap: trying to automate a broken process. AI applied to a mess gives you a faster mess.

6. Governance and policy

Are there rules for how AI gets used, what data goes near it, and who checks the output.

• Not ready (1): No policy. People paste whatever they like into whatever tool they fancy.
• Getting there (2): Informal norms and a general sense of caution, but nothing written or enforced.
• Ready (3): A short, real AI policy covering acceptable use, client and personal data, and how output gets checked before it goes out.

Common SME gap: treating governance as an enterprise problem. It is not. A one-page set of rules prevents most of the trouble small businesses get into. Our AI Charter is built for exactly this.

How do I score it and what does the total mean?

Score each of the six dimensions from 1 to 3 using the descriptions above, then add them up for a total out of 18. The total tells you what to do next, not just how you feel about it.

• 6 to 9: Foundations first. You are not ready to spend on AI yet, and that is fine. Pick the lowest-scoring dimension and fix that before anything else. Usually it is data or governance. Tools can wait.
• 10 to 14: Ready for a focused pilot. You have enough in place to run one well-chosen experiment. Pick a single process where you scored well, set a clear measure of success, and learn from it before going wider.
• 15 to 18: Ready to scale carefully. The foundations are solid. Now the risk is moving too fast and skipping governance. Keep the rules ahead of the rollout.

Two honest notes. Be strict: the temptation is to mark yourself a 2 out of charity, and a self-assessment is only useful if it is uncomfortable. And the total matters less than the lowest number. A business scoring 16 with a 1 on governance is not ready to scale; it is one careless prompt away from a data problem. Fix the weakest dimension first, every time.

That is the logic behind structured AI enablement: get the foundations right, then layer tools on top, not the other way around.

What is the most common SME readiness gap?

Governance, with data a close second. Running AI enablement for small businesses, we see these two dimensions score lowest almost every time, and they are the two that quietly cause the most damage.

Governance scores low because it feels like bureaucracy until something goes wrong. Most SMEs adopt AI tools person by person, long before anyone writes a rule about client data, acceptable use or checking output for accuracy. By the time a policy gets written, there is usually a near-miss behind it. One page of sensible rules first is far cheaper than the alternative.

Data scores low because useful AI sits on top of accessible, reasonably clean information, and most small businesses run on data scattered across inboxes, spreadsheets and the odd legacy system. You do not need a grand data project, just the data you already have tidy enough that a tool is not starting from chaos.

The pattern holds nationally. The UK government’s own research found the biggest barriers to AI adoption are a lack of identified need and limited AI skills and expertise, not the cost of the tools. Readiness is mostly about people, clarity and rules. The technology is the easy part, and rarely where SMEs come unstuck.