Ten honest questions. Two minutes. A score, a band, and the three things worth fixing before you spend anything.
Readiness has very little to do with which AI tool you pick. The businesses that get value from AI first are the ones with repeatable processes, accessible data and a specific task worth automating. The ones that struggle usually have none of those and are hoping a tool will supply them.
The most common failure pattern in small businesses is adopting a tool with no defined job to do. The strongest starting point is the opposite: one task that happens often, follows roughly the same shape each time, currently eats hours, and has a clear right answer. Fix that first, measure the time saved, then widen.
Answer honestly. An inflated score helps nobody.
Here is exactly what the tool does with your numbers.
Each question scores 0 for "not yet", 1 for "partly" and 2 for "yes". Your score is the percentage of the available points across the questions you answered, so it only appears once all ten are in.
| Band | Score |
|---|---|
| Ready to build | 80 to 100 |
| Nearly there | 60 to 79 |
| Foundations needed | 40 to 59 |
| Start with the basics | Under 40 |
The ten questions split into three groups. Foundations covers whether you have documented, repeatable processes and data a system could reach. Capability covers practical skills and the governance to use AI safely. Commitment covers whether anyone owns the work and whether there is money behind it.
Foundations carries the most questions deliberately. In practice it is the group that decides whether an AI project produces anything, and the one most often skipped.
Three things, none of which are a tool. Repeatable processes, because you automate a process rather than a vague intention. Accessible data, because a system cannot use information locked in paper or someone’s head. And a specific task worth automating, chosen because it happens often and eats real hours.
Businesses with those three get value quickly regardless of which product they choose. Businesses without them usually buy a subscription that goes unused.
Pick a task that is frequent, roughly the same shape each time, currently manual, and has a clear right answer. Drafting replies to common customer questions, summarising notes into follow-ups, extracting information from documents and preparing first-draft quotes all fit.
Avoid anything where a wrong answer is expensive and hard to spot. Early projects should be easy to check, so mistakes surface immediately rather than compounding.
No, but it needs to be reachable. Data that is digital and searchable is enough to start, even if it is messy. Data that exists only on paper, in email threads or in someone’s memory is the real blocker.
Start with the one area where your records are already in reasonable shape rather than waiting for a full data clean-up that may never finish.
Yes, and it can be one page. The essentials are what may not be pasted into public AI tools, who checks output before it reaches a customer, and which tools are approved.
Small teams need this more than large ones, not less, because there is rarely anyone reviewing work before it goes out. The realistic risk is not dramatic misuse but a confidential detail pasted into the wrong box, or a confident wrong answer sent to a client unchecked.
Built by Veris Labs
These calculators are a small, public version of what we do. The Veris Labs suite covers marketing and delivery, and where nothing off the shelf fits, we build it around your business instead.
