Are you ready for AI? Seven questions to answer first

Most leadership teams now feel they should be doing something with AI. A competitor mentions a chatbot, a board member forwards an article, a vendor offers a free pilot. The pressure is real, and so is the risk of spending six months and a healthy budget on a project that nobody uses.

The businesses that get value from AI rarely start with the technology. They start with a short, honest conversation at leadership level about whether the organisation is ready. Below are the seven questions we ask first. If you can answer them clearly, you are ready to move. If you cannot, you have found your real starting point.

The problem and the raw material

1. What problem is actually worth solving?

"Use AI" is not a goal. A useful problem has a cost you can feel: enquiries that wait two days for a reply, a finance team that rekeys supplier invoices every month, sales staff who spend Friday afternoons building reports. Write the problem in one sentence, in the language of the business, with no mention of AI. If the sentence would still matter without the technology, it is worth pursuing.

2. What data do you actually have?

Not the data you wish you had, or the data that sits in someone's head. Look at what is written down, where it lives and how clean it is. A typical distributor has years of orders in an accounting system, price lists in spreadsheets, and customer conversations scattered across personal WhatsApp accounts. AI can work with imperfect data, but it cannot work with data that does not exist or that nobody is allowed to access.

The work and who owns it

3. Is the process clear enough to describe?

If two people in your team handle the same task in different ways, AI will not settle the argument. It will copy the confusion faster. Before any tool is chosen, someone should be able to write the process down in steps: what comes in, what decisions are made, what goes out and who checks it. This is often the most valuable part of the exercise, and it costs nothing but time.

4. Who will own it?

Every AI project needs one named person who is accountable for the result, not just the installation. In family businesses this is often where things stall: the founder is enthusiastic, the IT person is cautious, and the department head who would use it was never asked. Ownership belongs with the person whose numbers improve when the work improves.

The guardrails and the people

5. How much risk can you accept?

Be specific. Drafting internal meeting notes carries little risk. Sending price quotes to customers without a human check carries a lot. Handling patient records at a clinic or student data at a school carries legal and reputational weight. Leadership should agree, in advance, which kinds of work AI may assist with, which need human review, and which are off limits for now.

6. Do your people have the skills, or the time to learn?

Skills here means more than technical ability. It means knowing how to give a clear instruction, how to check an output, and when to distrust it. Most teams can learn this quickly, but only if leadership protects time for it. A tool handed to an overloaded team with no training will be tried once and then quietly abandoned.

The scoreboard

7. How will you know it worked?

Decide the measure before you start. Hours saved per week on a named task. Average reply time to enquiries. Errors caught before invoices go out. Pick one or two numbers you already track, or can start tracking this month, and record where they stand today. Without a baseline, every AI project ends in opinions rather than evidence.

If you cannot describe the problem without mentioning AI, you have not found the problem yet.

How to read your answers

Go through the seven questions with your leadership team in a single sitting. Be honest about where the answers are vague. Your results will usually fall into one of four patterns:

• Clear on most questions: you are ready for a focused pilot on one process, with a named owner and an agreed measure.
• Clear on the problem, unclear on process or data: spend the next few weeks documenting the workflow and tidying the data. This work pays off whether or not AI follows.
• Unclear on ownership or risk: the gap is at leadership level. Settle who decides and what is allowed before any tool is bought.
• Unclear on the problem itself: pause. Talk to the people doing the work and ask where their time goes. The right use case usually appears quickly once you listen.

None of these outcomes is a failure. Each tells you what to do next, which is more than most AI conversations achieve.

A sensible next step

Readiness is not about being advanced. A mid-sized manufacturer with tidy order data and one clear bottleneck is often more ready than a larger firm with an innovation team and no agreed problem. The questions above favour clarity over size.

If you would like a second pair of eyes on your answers, our complimentary one-week Cost Review looks at where time and money are leaking across your operations and sets out a 90-day plan, including whether AI belongs in it yet. The plan is yours to keep, whether or not we work together.