Readiness check
Twelve questions before you spend a dollar on AI
Answer honestly; there are no trick questions and nothing you tick leaves this page. At the end you’ll see which of three starting points fits your business today, and why.
How to read your result
The check sorts businesses into three starting points. None is better than another; they are simply different first steps.
Starting point A: tidy the foundations
If you ticked fewer than two of the systems questions, the most valuable step is not AI at all. Get email and files into a managed business account, make sure client folders have sensible permissions, and make sure someone is responsible for security. Every AI tool you might later use reaches your information through those foundations, and if access is muddled, an AI assistant will happily surface a document to someone who should never have seen it. Your IT provider can usually do this work; a consultant should tell you so rather than sell you an AI project on shaky ground.
Starting point B: write it down, then measure it
If your systems are in reasonable shape but you ticked fewer than two of the work questions, the bottleneck is that nobody has pinned down the process. Pick your most irritating weekly task. Write the steps on a page. Time it for two weeks. That document, and that number, are what any sensible AI project starts from — and doing it yourself saves you paying someone else to do it in an audit.
Starting point C: a small, controlled pilot
If you ticked most of the boxes in all three groups, you are in a good position to trial one automation. The usual shape is a short audit to confirm the best candidate and how it will be measured, followed by a pilot that runs inside your existing tools for 30 to 60 days with a person approving each output. The costs page explains what that typically involves.
Why the people questions matter most
Technical readiness is the part that gets talked about, but most stalled AI projects in small businesses stall for human reasons. The owner who commissioned it gets busy and nobody checks the drafts, so trust never builds. Staff who weren’t asked feel it is being done to them. Or nobody agreed what “better” meant, so after two months there is a vague feeling that it’s fine, and no one can say whether to keep paying for it.
The fix is simple and unglamorous: one named person who owns the pilot, a before-and-after number, and the people who do the work involved from the first week. If you can’t find that person right now, wait until you can.
Questions people ask after doing the check
We use personal Gmail accounts. Is that a problem?
For AI it is a real one. Business tools connect to business accounts with admin controls, audit logs and the ability to remove access when someone leaves. Moving to a managed account is a worthwhile project on its own.
We’re already paying for Copilot or ChatGPT seats. Does that count?
It means your team has a drafting assistant, which is useful. It doesn’t on its own connect that assistant to your processes, approvals and other systems. Many businesses find a handful of people use it daily and the rest never open it; a workflow-level project is a different thing.
Our process is mostly on paper. Should we scan everything first?
Not necessarily. Start with new work arriving digitally and leave the archive alone unless there is a specific reason to search it.
Want a second opinion on your result? Book a 15-minute call and bring it with you.