The AI Readiness Checklist For UK SMEs Before Hiring A Consultant

More UK businesses are using AI than ever, and more of them are thinking about paying someone to help. The government’s AI adoption survey report 2026, published on 29 October 2026, found that 32% of UK businesses now use at least one AI technology, up from 16% a year earlier. It also found that limited skills, expertise or understanding of use cases is the top barrier, cited by 67% of adopters and 78% of businesses planning to adopt.

That gap is exactly what AI consultants are selling into. But a consultant is only as useful as the starting point you give them. If you hire one before you know your processes, your data, your existing licences and what success looks like, you will spend the first weeks of the engagement paying for discovery you could have done yourself. This checklist is designed to be worked through before you sign anything in 2027.

How To Use This Checklist

Work through the seven sections with two or three colleagues who know how the business really runs, not just the leadership view. You do not need a yes to every item. The aim is to know where you stand, so that the conversation with a consultant starts at your real problems rather than at a generic workshop. Anything you cannot answer becomes a question for them, and their response tells you a lot about whether they are the right fit.

1. Processes And Data

AI works on processes and data. If neither is clear, the first project will stall no matter how good the tool or the adviser. Before you look at tools, check the following.

• Can you name the five processes that eat the most staff time each week, and roughly how many hours each takes?
• Is each of those processes written down, even as a rough list of steps, or does it live in one person’s head?
• Do you know where the data for each process lives: which systems, spreadsheets and inboxes?
• Is that data reasonably clean and consistent, or would it need tidying before any automation could rely on it?
• Do you know which of it is personal data under UK GDPR, and which is commercially sensitive?

The processes that make the best first projects are high-volume, rule-based and painful: chasing invoices, re-keying data between systems, triaging inbound enquiries, producing the same report every week. Our analysis of why UK small businesses are using AI but not seeing results found that firms bolting AI onto the edges of unchanged processes see the least benefit, which is why this section comes first.

2. Tools You Already Pay For

Many SMEs already hold AI licences they have never switched on or configured. Before you budget for new software, audit what is already in your stack.

• Microsoft 365. Microsoft 365 Copilot Chat is included at no extra cost for Microsoft 365 customers. Since 1 December 2025, organisations with 300 or fewer employees can also buy Microsoft 365 Copilot Business, listed at $21 per user per month in Microsoft’s US pricing. Check UK pricing with your reseller.
• Google Workspace. Since January 2025, Google has included its AI features in Workspace Business and Enterprise plans without the need for a separate add-on, and admins can switch the Gemini app on or off for the organisation.
• Line-of-business software. CRM, accounting, helpdesk and HR platforms increasingly ship AI features inside existing plans. List each system and note what AI features it offers and whether they are enabled.
• Automation tools. If you already pay for an automation platform, note who uses it and what it connects. Existing integrations are often the fastest route to a first AI workflow.

A consultant who recommends a new platform before asking what you already own is a warning sign. A good one will want this list on day one.

3. Governance And Shadow IT

Whatever your official AI policy, your staff are probably already using AI. Research commissioned by Microsoft and carried out by Censuswide in October 2025 found that 71% of UK employees have used unapproved consumer AI tools at work, and 51% still do so every week. Only 32% said they were worried about the privacy of company or customer data entered into those tools.

That is Shadow IT, and banning it rarely works. The better route is to give people approved tools and clear guardrails. Check where you stand:

• Do you know which AI tools staff are using today, approved or not? A short anonymous survey is often the quickest way to find out.
• Do you have a written AI acceptable use policy that says which tools are approved and what data must never go into them?
• Is there a named person responsible for approving new AI tools and reviewing their terms, data hosting and retention?
• Have you checked whether any planned use involves personal data? The ICO’s guidance on AI and data protection says that in the vast majority of cases AI use will be high-risk processing that requires a data protection impact assessment.
• Do you have basic security controls in place? The government’s Cyber Essentials scheme is a sensible baseline before you connect AI tools to business systems.

If you plan to use AI agents that act on your systems rather than just answer questions, our AI agent security checklist covers the extra controls you will need.

4. Budget

The government survey makes a useful point about cost: businesses that named it as a barrier were mostly talking about the time, effort and resources needed to implement AI, not the price of the tools. Budget for all three.

• External help. Decide whether you want a fixed-scope project, a short readiness audit or ongoing fractional support, and set a ceiling for the first phase.
• Licences. Include any new seats or usage-based charges, and check whether existing licences cover what you need first.
• Staff time. Process mapping, testing and training all take hours away from the day job. If nobody has time set aside, the project will drift.
• Running costs. Automations and AI apps need maintaining. Allow for someone to own them after the consultant leaves.

A pilot with a clear budget and a decision point at the end is almost always better value than a large programme agreed up front.

5. Internal Champions

Every successful AI rollout has people inside the business who care about it. Consultants can design and build, but they cannot make your team adopt what they build. Before you hire, identify:

• An executive sponsor who can make decisions, unblock budget and say no to scope creep.
• One or two champions in the teams whose work will change, ideally people colleagues already go to for help with technology.
• Time in those champions’ week, formally agreed with their managers, to learn the tools and support others.
• A plan for training the wider team, so that AI skills do not stay with the champions alone.

6. Success Metrics

Among non-adopters in the government survey, 65% cited an unclear business case or return on investment. The fix is to decide how you will measure success before the project starts, not after.

• Have you measured a baseline for each target process: hours per week, turnaround time, error rate or cost per transaction?
• Have you agreed one or two headline metrics per project, rather than a long list nobody tracks?
• Do you know who will collect the numbers and when you will review them?
• Have you defined what result would make you stop, change course or scale up?

Hours saved per week is often the simplest and most persuasive measure for an SME. It is easy to baseline, easy to explain to a board and easy to convert into a cost figure.

7. Questions To Ask Any AI Consultant

Once you have worked through the checklist, you are in a strong position to judge advisers. These questions separate consultants who will deliver lasting change from those who will leave you with a slide deck.

1. How do you assess readiness before recommending tools, and what do you need from us to do it?
2. Which of our existing licences and systems would you use before suggesting anything new?
3. How do you handle governance, data protection and Shadow IT, and will you help us write an AI use policy?
4. Can you show results from a business of our size, with measured outcomes rather than general claims?
5. How will you train our team and champions so we can run things without you?
6. What does the first 90 days look like, and what will we have at the end of it?
7. Who owns the automations, apps and documentation you build, and what happens to them if we stop working together?
8. How do you price the work, and what is outside the quote?

If you are still building a shortlist, our roundups of AI consulting firms in Europe and generative AI consulting firms are a useful starting point.

AI Readiness Checklist At A Glance

Area	You are ready when	Red flag
Processes and data	Your top time-consuming processes are listed, roughly mapped and their data sources known	Nobody can say where the hours go
Tools already licensed	You have a list of current licences and which AI features are switched on	Paying for AI features nobody has enabled
Governance and Shadow IT	There is an approved tools list, a use policy and a named owner	Staff pasting customer data into personal AI accounts
Budget	A first-phase ceiling covers external help, licences and staff time	Budget for tools only, with no time set aside
Internal champions	A sponsor and team champions are named, with agreed time	The project belongs to whoever has the most spare time
Success metrics	Baselines are measured and one or two metrics agreed per project	Success defined only as “using AI”
Consultant selection	You have a shortlist and a set of questions to compare them	Hiring the first firm that offers a free workshop

Frequently Asked Questions

What is an AI readiness assessment?

It is a structured review of whether a business is in a position to use AI well. It usually covers processes, data quality, existing systems, skills, governance and how success will be measured, and it ends with a shortlist of use cases worth pursuing first.

Do small businesses need an AI consultant?

Not always. Many SMEs can get started with tools they already pay for and a clear policy. A consultant earns their fee when processes span several systems, when personal data is involved, or when the business lacks the time or skills to design and maintain automations itself.

How long does it take to see results from AI?

It depends on the starting point. A well-chosen pilot on a single process can show measurable time savings within a few months. Programmes that start without baselines or clear ownership often take much longer to show anything at all.

What should an AI policy for an SME include?

At minimum: which tools are approved, what data must never be entered into them, who approves new tools, how outputs should be checked before use, and how the policy fits with your existing data protection obligations under UK GDPR.

Conclusion

Hiring an AI consultant can be one of the best investments a UK SME makes in 2027, but only if the business is ready to use the help. Knowing your processes and data, auditing the licences you already hold, getting ahead of Shadow IT, setting a realistic budget, naming internal champions and agreeing how success will be measured will make any engagement faster, cheaper and more likely to stick.

Work through this checklist first, take the gaps to your shortlist of consultants as questions, and choose the adviser whose answers show they understand your business rather than just the technology.