Free framework · 20 minutes

AI Readiness Assessment for UK Small Businesses

Find out if your business is ready to use AI safely, usefully, and without wasting money. Seven pillars, plain English, no jargon.

Jump to the scorer ↓

What it is

An AI readiness assessment is a 20-minute review of whether your business is set up to use AI well. It scores seven areas (strategy, data, skills, workflows, tools, governance, adoption) and tells you the one thing to fix first.

The framework

The seven pillars

Read all seven, or jump to the one most relevant to your business. Each pillar has a short explanation, the assessment questions, and a worked example from a UK small business.

Pillar 1 of 7

Strategy

Do you know what you would use AI for?

The most common barrier to AI adoption in the UK is not cost or skill. It is not knowing what to use AI for. Government research found 71% of UK businesses with no AI plans cite a lack of identified use case as the top reason. The fix is not more AI tools. The fix is a sharper question.

A strategy answer is not "we want to use AI for marketing". It is a specific job with a measurable outcome: "we want to reply to overnight enquiries within five minutes during peak season" or "we want to forecast stock for Christmas". The narrower the answer, the better the chance of getting somewhere.

This pillar checks whether you have a defined use case worth pursuing.

Ask yourself

  • Can you name three workflows where you have thought about using AI?
  • For each, can you describe the current pain in plain English?
  • Is there an obvious save-time or save-money measure you would use to know if AI worked?
  • Have you tried using ChatGPT or Copilot on any of these workflows yet?

Pillar 2 of 7

Data

Is your business's information findable and clean?

AI works on top of your data. If your information is fragmented across email threads, paper, three spreadsheets, and your team's heads, AI cannot do much. The phrase that recurs across SME interviews is "your best knowledge is stuck in someone's head."

This pillar is not about big data. It is about whether your AI assistant can find the customer record, the order history, the brand guidelines, the standard operating procedures. Could a new hire find your top twenty documents without asking? If not, neither can an AI tool that depends on them.

This pillar checks the readability of the information your business already holds: readable by another human, never mind AI.

Ask yourself

  • If a new hire joined Monday, could they find your top twenty documents without asking?
  • Is your customer information in one place, or scattered across email, phone notes, and one person's memory?
  • Do you have written processes for the things you do most often?
  • How much of your business's knowledge would walk out the door if your most senior team member left?

Pillar 3 of 7

Skills

Who on the team can actually use these tools?

60% of UK businesses cite limited AI skills as a top barrier to adoption. The fix is not a five-day course; it is a few people who can use AI tools as part of their normal working week.

Skills here are practical, not theoretical. Can someone open ChatGPT, Claude, or Copilot, write a useful prompt, and get a reliable output? Can they show a colleague how to do the same thing? Your team needs rules, examples, and confidence, not a lecture.

This pillar checks whether AI capability is concentrated in one person (risky) or distributed across the team (durable).

Ask yourself

  • Is at least one person on your team comfortable using ChatGPT, Claude, or Copilot for real work?
  • If that person left tomorrow, would anyone else know the prompts they use?
  • Are you the only person in the business who has tried AI tools so far?
  • Has anyone documented the AI prompts that work, so the team can reuse them?

Pillar 4 of 7

Workflows

Which processes are mappable and worth automating?

Most AI failures start with the wrong workflow. You cannot automate what you cannot describe step by step. Map the workflow before buying the tool.

A mappable workflow has a clear trigger (an enquiry arrives), a sequence of steps (look up history, draft reply, send), and a clear output (a sent email and a tracked record). The job in this pillar is to separate workflows that fit that pattern from workflows that do not.

The principle worth remembering: automate the admin, not the judgement. The packing of a Christmas hamper is a judgement call about gift presentation. The printing of the label is admin. AI should touch the second, not the first.

Ask yourself

  • Could you draw the steps of your quoting process on a single page?
  • Are there steps you skip when you are tired, or duplicate when you forget?
  • Which workflows are automate-the-admin (safe) and which are automate-the-judgement (do not)?
  • Where do bottlenecks form when the business gets busy?

Pillar 5 of 7

Tools

Do you have the right software stack?

AI is most useful when it plugs into the software you already run. If your CRM is Excel and your bookings are paper, you will need to fix the foundation before adding AI. The aim is to build around the tools your team already uses, not to add more tools.

This pillar is not about expensive enterprise software. It is about whether you have a system of record (CRM), a way to capture work (project tool), a way to communicate (email or chat), and a way to store documents that the rest of the team can find.

This pillar checks whether your foundation is shaped for AI to attach to.

Ask yourself

  • Where is your customer information? (One name only.)
  • Where is your project or job tracking? (One name only.)
  • Are your tools connected, or do you copy-paste between them?
  • Have you accumulated three tools that do the same job and never picked a winner?

Pillar 6 of 7

Governance

How will you handle risk, privacy, and AI mistakes?

AI makes mistakes. The question is not whether but how often, and what your business does when it happens. Use AI without making the business less human.

Governance is not a fifty-page policy document. It is a few clear rules: what AI is allowed to see, what AI is allowed to send, and who reviews AI output before it goes to a client. UK GDPR applies. The ICO publishes specific guidance on AI and data protection for UK organisations.

This pillar checks whether you can use AI safely under UK GDPR and your own client commitments.

Ask yourself

  • Do you have a one-page rule for what client data can be put into ChatGPT or Claude?
  • Is there a human-reviews-this-before-it-sends step on AI-generated client communication?
  • Have you signed any client contracts that mention AI use or restrict it?
  • If your AI tool sent something incorrect to a client tomorrow, who would notice, and how quickly?

Pillar 7 of 7

Adoption

Will your team actually use what you build?

The most common cause of failed AI projects is not the tech. It is that nobody uses it. A workflow that takes thirty seconds longer than the old way will be quietly abandoned within two weeks.

Adoption is not training. It is making the AI version the default path: faster, more useful, embedded in the tool your team already opens. For businesses too busy to become AI experts, this matters more than the model choice.

This pillar checks whether you have the operational habit to embed a new step in your working week.

Ask yourself

  • When you have added a new tool before, did the team adopt it within a month?
  • Is there a person whose job includes making sure we actually use the new thing?
  • Are you willing to drop a tool that does not earn its place?
  • How will you know within four weeks whether the new AI workflow is being used?

The quick scorer

Take the assessment

The framework above is the slow read. The quiz below is the quick scorer. Twelve questions about your business, two to three minutes, and you get a tier (Not ready, Partially ready, or Ready) plus an estimate of the hours per week AI could save your team.

1 of 12×

1 →

What kind of business do you run?

This helps us tailor your recommendations.

chanikul.com

Want the £-quantified version? Try the 8-minute audit →

How this compares

How this framework compares

Other frameworks reach the same destination by different routes. Microsoft's is deeper on enterprise governance. Cisco's leans on infrastructure (which they sell). TDWI's 75 questions are essential if you have a data team to answer them. Naheed et al.'s academic model carries peer-reviewed credibility. This framework is the shortest and the most directly written for owner-operators who do not have a CTO and do not want a fifty-page PDF.

FrameworkDimensionsOriginCostBest for
You are here
This framework		7 pillars: Strategy, Data, Skills, Workflows, Tools, Governance, AdoptionUK SME consultant, DevonFreeOwner-operated UK SMEs, 10-100 staff
Cloud Adoption Framework for AI6 steps: AI Strategy, AI Plan, AI Ready, Govern AI, Manage AI, Secure AIMicrosoftFree (Azure context)Enterprises adopting Microsoft and Azure AI
Cisco AI Readiness Index6 areas: Strategy, Infrastructure, Data, Governance, Talent, CultureCiscoFree (registration)Large IT-led organisations
TDWI AI Readiness Assessment5 categories, ~75 questions: Organizational, Data, Skills, Operational, Governance ReadinessTDWI (1105 Media)Free (registration)Data-mature organisations with analyst teams
Multidimensional AI Readiness Model for SMEsTOEH dimensions: Technology, Organization, Environment, HumanNaheed, Pinto & Pirola (2025), peer-reviewedFree (academic PDF)SMEs wanting academic rigour

What to do next

What you do after the assessment

If you score 0 to 30 · Not ready

Most of your gaps are foundational. Read the field guides linked in the data, skills, and adoption pillars before spending money on AI tools. Save the time you would spend evaluating software, and invest it in fixing one foundation first.

If you score 30 to 60 · Partially ready

You have one or two strong pillars and a few weak ones. Book the AI Opportunity Audit. We will map your specific workflows, quantify the savings in pounds, and tell you whether AI is worth building before you spend a penny on implementation.

If you score 60 to 100 · Ready

Your foundation is set. The audit will focus on workflow prioritisation, not foundation repair. A faster path from assessment to first pilot: usually six to eight weeks rather than three to six months.

Questions about this assessment

Sources

Written by Chanikul Dechpholkrang, AI consultant, Plymouth, Devon.

Updated 19 May 2026

The next step

Ready for a paid audit that quantifies the savings?

The AI Opportunity Audit maps your specific workflows, finds the AI savings, and delivers a prioritised roadmap. Yours to keep, even if you never hire me to build.

Book the AI Opportunity Audit

Find out what manual work is costing you

Ninety seconds, six questions. We'll send the £ figure and three AI quick wins for your industry.