Free scorecard

How ready is your company for AI? Ten questions, three minutes

This is for the owner of a company with roughly USD 2 to 50 million in revenue. Ten plain questions about what you know, what is written down and who is responsible. Pick the answer that is true today, not the one you hope to be true. You get a score out of 20 and the areas to look at first.

A score is a starting point, not an audit.

1. Do you know which AI tools your staff already use for work?
2. Is there one named person who makes the decisions about AI in the company?
3. Do you have a written list of what AI may never do on its own?
4. If a lender, buyer or board member asked in writing how your company uses and controls AI, could you answer this week?
5. Do you have written rules on what customer and employee information may be put into AI tools?
6. Are staff using company-approved AI accounts rather than personal ones?
7. Does a person check AI output before it reaches a customer, a payment or a record?
8. Have your people been given short, plain guidance on using AI at work?
9. Do you know what your software vendors and suppliers do with AI and with your data?
10. Do you review AI use, and know what to do when it gets something wrong, at least twice a year?

How the score is built

There are ten questions. Each one is about something an owner can check without a consultant: whether you know what tools people use, whether one person owns the decisions, whether the rules are written down, whether a person checks the output, and whether anyone reviews it. Each answer is worth 0, 1 or 2 points: 0 for "no" or "I do not know", 1 for "partly" or "informally", 2 for "yes, and it is written down or in practice". Add them up and you have a score out of 20.

The questions are weighted equally on purpose. Nobody has shown which of the ten matters most for every company, so the scorecard does not pretend to. The areas listed under your result are simply the ones where you answered lowest.

The three bands

Starting point (0 to 7). Most of the basics are not yet in place. That is common and it is cheap to fix. It means that if someone asked how AI is used and controlled in your company, the honest answer today would be "I am not sure". The risk is not that you are behind. The risk is that people are already using these tools and nobody has decided anything.

Partly in place (8 to 14). You have some foundations, but a lot of them live in people's heads. The gap shows when a lender, a buyer, an insurer or a board member asks a question in writing. The work here is turning habits into short written rules with a name next to each.

Well covered (15 to 20). You have most of the groundwork. The risk now is that it goes stale, or that it exists in conversation but not in a form you could hand to someone outside the company. The work is keeping it current.

What to do in the first week, by band

If you are at the Starting point: spend the week on three things and nothing else. Ask your managers, in one short message, which AI tools their teams use for work. Name one person who owns decisions about AI, and tell them. Write a half-page list of what AI may never do without a person, such as sending money, signing or deleting records. Do not buy or build anything until those three exist.

If you are Partly in place: take your lowest three areas from the result and write one short paragraph for each, with a name and a date. Then put the data rule in front of staff: which tools are approved, what customer and employee information may go into them, and who to ask. A page that staff have actually seen is worth more than a long policy nobody has read.

If you are Well covered: put a review date in the calendar, twice a year. Check that your written answer for lenders, buyers and the board is current, and that it matches what people do. Ask your main vendors again what they do with AI and your data, because their terms change.

Who this is for

The scorecard is written for owners and senior leaders of companies with roughly 50 to 500 people. At that size AI use is usually already happening, mostly without a decision behind it, and nobody has the job of noticing. It also suits the people who advise those companies: an association, an accountant or a banker who wants members or clients to get a quick, honest first read.

It is not built for a company of five, where the owner already knows everything everyone does, nor for a regulated enterprise with a compliance department. If a lender, a regulator or a sale process is already asking you for evidence, a score is not enough and you will want a proper review.

How to use this with your leadership team

Take it yourself first, quickly, and write down your score. Then ask each member of your leadership team to take it separately, without comparing answers, and bring their score to a thirty-minute meeting.

The useful part is not the totals. It is the questions where people answered differently. If you said "yes, a current list" and two managers said "no idea", you have found the real gap: what the owner believes is in place and what the company experiences are not the same thing. Pick the three questions with the widest disagreement, name an owner for each, and agree a date to look again. Repeat the exercise in six months and compare.

If you want the thinking behind the questions, read Do I need an AI strategy?, Who owns AI in a company? and What is shadow AI?. If you would like the board to see the answer in one place, the AI Board Briefing is built for that.

Put this scorecard on your association's or firm's site

If you run an association, a peer group or an advisory firm, you are welcome to put a compact version of the scorecard on your own site, free. It has the same ten questions and the same scoring, collects no email and sets no tracking of its own, and links back to this page for the full result and next steps.

Copy and paste this where you want it to appear:

Questions owners ask about the scorecard

What does the AI readiness scorecard measure?

Whether the basic groundwork for using AI safely is in place: who knows what is in use, who decides, what is written down, what data may go into tools, who checks the output and who reviews it. It does not measure how advanced your tools are.

How is the score calculated?

There are ten questions. Each answer is worth 0, 1 or 2 points, so the total is out of 20. A score of 0 to 7 is Starting point, 8 to 14 is Partly in place and 15 to 20 is Well covered.

Is the scorecard an audit?

No. It reflects only your own answers to ten questions and takes about three minutes. It tells you where to look first. It does not test your systems, read your contracts or replace advice from your lawyer or accountant.

Who is the scorecard for?

Owners and senior leaders of companies with roughly 50 to 500 people, and the leaders who report to them. It also works for an association or advisory firm that wants to offer members or clients a quick first read.

Do I have to give my email to see my result?

No. The result shows as soon as you answer the tenth question. The email form below the result is optional and only used to send you the written result with the next steps for your band.

Benchmarks by industry or company size are not published here. If we ever publish them, it will be only when there are enough real responses to make them mean something.