An AI readiness assessment tells you whether your business has what it takes to use AI in a way that actually helps. For a small business, five questions cover it. Can you write down a recurring workflow? Is the data for it digital and accessible? Is there someone who can brief, check and approve? Are there rules on what data the AI may see and what never goes out without approval? Do you have a small budget of money and time? If the answer to all five is yes, you can start with a first role. Every no tells you what to sort out first.
I run my business with fourteen AI employees. The five questions come from how I work: every role gets a personnel file with its task and its limits, and anything that leaves the building goes through me first. They are my working rule, not a scientific maturity model, and on purpose they don't add up to a score.
What AI readiness means, and what it doesn't
Public readiness checks usually look at several areas. The KI-Readiness-Check run by the Mittelstand-Digital Zentrum Kaiserslautern, a German SME support centre, names four: technology, organisation, people and environment. In other words: data and IT, processes and culture, how staff see AI, and law, data protection and compliance. At the end you get a maturity level.
For a company with several departments, that makes sense. In a business with five people, those areas shrink to a handful of questions you can answer in an afternoon. There's usually no IT strategy to assess, just an inbox, an accounting tool and an owner who is short on time. The question that matters isn't "How mature are we?" but "What's stopping us from starting with one task next week?"
Your AI readiness assessment: five questions
1. Can you write down a recurring workflow?
Pick a task that comes up every week and describe it in five to ten sentences. What triggers it, what information do you need, which steps follow, and how do you recognise a good result? If you can do that, the answer is yes.
If you can't, the knowledge only lives in your head. Then even the best tool won't help, because an AI can take over a workflow but it can't invent one. How to find and describe your workflows is in Process Optimization With AI: The Method.
2. Is the data digital and accessible?
The input for that workflow has to sit somewhere an AI can reach: an inbox, a shared folder, an accounting tool with an interface. Piles of paper, private phones and sticky notes on your desk count as a no.
In my business, Frieda connects to the inbox through official connectors for Gmail or Microsoft 365, with no passwords in plain text. If your receipts are still on paper, the first step is digital filing, as described in Digital Receipt Filing That Holds Up.
3. Is there someone who can brief, check and approve?
An AI employee needs someone who tells it what to do, checks its results and gives feedback. That person needs regular time for this in the first few weeks, not just once at the start, but no lengthy training. What the EU AI Act expects here is in AI Literacy Under the AI Act: Training Staff. If nobody can or wants to take this on, that's a clear no.
In my business, that part stays with me: giving the assignment, approving the result and deciding what goes live. That's also why my community says it plainly under "Not for you, if": if you're looking for a solution that runs without you.
4. Are there rules for data and approvals?
Check whether two rules are written down: who approves before anything leaves the building, and which folders and data the AI may see and which it never may, job applications or health data, for example. Why both need to be in place before the first real access is explained in Bringing AI to a Small Business.
One page is enough. What belongs in it is in AI Usage Policy for Small Teams, and which data an AI should see is in AI and Privacy: What the AI Gets to See. If your team is already using AI tools on their own, read Shadow AI: When Your Team Uses AI in Secret first.
5. Do you have a small budget of money and time?
To get started with one role, a single subscription to an AI tool is usually enough. Time matters more: the briefing, the first review rounds and the feedback take working hours in the first weeks, before you feel any relief. If you can't spare either the subscription or those hours, postpone the start rather than doing it half-heartedly. What an AI employee actually costs is in What an AI Employee Really Costs.
How to read your result
This assessment doesn't produce a score, and that's deliberate. A percentage sounds precise, but it doesn't tell you what to do on Monday. Each no, on the other hand, points to a concrete step:
The order of the rows is also the order in which I'd close the gaps. Without a described workflow and without an owner, a data rule doesn't get you far. Once both are in place, the remaining questions are about tidying up and agreeing on things, not about fundamentals.
From AI readiness to an AI roadmap
An AI roadmap sounds like a project plan spanning months. In a small business it's simply an order: which task comes first, which one next? I recommend one role per week, with a decision at the end of each week on whether it stays. What that looks like over four weeks, covering the inbox, quotes, finding prospects and the website, is in Bringing AI to a Small Business.
Before the roadmap comes the question of where AI pays off at all. Which tasks are worth tackling first in each department is in AI Use Cases for Small Businesses. Which workflow in your business still runs through you the most is what my Bottleneck Check shows you, in ten questions and without signing up.
Public readiness checks compared
"AI readiness check" can mean very different things. Three examples, as of 28 October 2026:
- KI-Readiness-Check (KIRC) by the Mittelstand-Digital Zentrum Kaiserslautern. A guided self-assessment in German for small and medium-sized companies, covering the four areas above and ending in a maturity level. Germany's Federal Network Agency links to it from one of its AI pages. According to the centre's own page, its public funding ran from August 2021 to the end of July 2026; the start page of the check was reachable on 28 October 2026.
- KI-Readiness-Check by Fraunhofer FOKUS. Not a self-assessment but a commissioned review by experts, which according to the page takes at most two weeks up to the final report. It is aimed at German statutory health insurers.
- My Bottleneck Check. It doesn't measure maturity. It shows which workflow depends on you the most and how many hours a week it costs you. It fits after this assessment, when you know you're ready but not where to begin.
Which route is right depends on your size. If you have several departments and your own IT, a maturity model gives you a broader view. If you have five people, the five questions above get you to a first step faster.
Frequently asked questions
What does AI readiness mean?
A business's ability to put AI to good use: described workflows, accessible data, someone in charge, clear rules and a small budget. If one of these is missing, getting started usually fails because of that gap, not because of the AI.
How do I assess my company's AI maturity?
Public checks like the KIRC from Kaiserslautern work out a maturity level across several areas. For a small business, the assessment above is often enough. It doesn't give you a number; it gives you a next step for every no.
Do I need an external AI readiness assessment?
Not to get started. If you can write down a workflow and you know where you lose the most time, you already know your first candidate. Outside support makes more sense when several departments are involved or the workflows depend on each other.
What is an AI readiness check?
A structured test of whether a business has what it takes to use AI. Public checks like the KIRC work out a maturity level; others are commissioned reviews by experts. The self-assessment above is the short version for small businesses.
What if I answer no to almost every question?
Then start with question 1 and write down a single workflow. It costs nothing in licences and makes your office tidier even without AI. After that, the rest of the assessment usually looks different.
Where to go from here
Answer the five questions today for a single task, not for the whole business. Once all five are a yes, take the Bottleneck Check and start with week 1 of Bringing AI to a Small Business. How I map and hand off workflows together with companies is on my process optimization page. Ready-made roles with approval rules are available to load in my community Claude Practitioners.
