Home › Blog › AI readiness

AI readiness: the twelve things worth checking first

If the mandate to adopt AI has arrived and you are not sure where to look first, it is faster to assess your own organisation before choosing a tool. Here are twelve questions across four axes.

12 checks 4 axes Free assessment

The short answer

Most failed AI adoptions begin not with the wrong tool but with a misjudged level of readiness. When an organisation whose data is not organised bolts on generative AI, the failure is not that the answers are wrong — it is that there is nothing for it to base an answer on.

The twelve below split into four axes: work, data, technology and organisation. Two or more "no" answers on any axis means that axis is your current bottleneck. Fixing it costs far less than fixing the tool you bought instead.

Axis 1 — Work: is there actually something to automate?

Repetition — Can you name three tasks that repeat in the same sequence every week? If not, the target for automation has not been identified yet.

Standardisation — Is that task handled the same way when a different person does it? If everyone does it differently, it is a candidate for standardisation, not automation.

Time spent — Can you state how many hours a month it consumes? Without a number you cannot prove the benefit either.

Axis 2 — Data: is there anything for AI to reference?

Digitisation — Is the information needed for a decision held in files and systems, or in someone’s head? If the latter, getting it out comes first.

Accessibility — Could one person gather the data they need within thirty minutes? If it takes requests to three departments, that is itself an automation project.

Consistency — Is the same field stored under different names in different systems? Where naming diverges, not even a person can aggregate it, let alone AI.

Axis 3 — Technology: is there somewhere to attach it?

Current use — Is anyone in the organisation already using an AI tool for real work? If nobody is, do not be optimistic about how fast it will spread.

Collaboration tools — Are documents and communication consolidated into one tool? If they are scattered, the places automation can attach are scattered too.

Integration — Do your core systems have an API — and if not, has file-based integration been confirmed as possible?

Axis 4 — Organisation: is there someone who will see it through?

Leadership backing — Is there enough support to try a second time after a failure? Demand results from the first attempt and most projects end at the pilot.

A named owner — Is there a person assigned to this by name? Projects run on the assumption that somebody will pick it up stall without exception.

Budget — Is there a plan for the monthly running cost, not just the setup cost? AI tools are a subscription, not a purchase.

If you scored low

Most organisations running these twelve checks for the first time start below half. That is normal, and it is not a problem. What matters is not the total but which axis is lowest.

A low work axis means starting by identifying what to automate. A low data axis means starting with organisation and access. A low technology axis means starting with tool consolidation. A low organisation axis means starting by naming an owner and securing budget. Skip the order and buy a tool first, and the tool will look like the cause of failure — when the real cause was the axis you skipped.

Frequently asked questions

Do we need all twelve before we can adopt anything?

No. Almost no organisation satisfies all of them. If even one axis is ready, you can start small on that axis. The point of this list is to find a starting point, not to gate you.

How does a self-assessment differ from a formal one?

A self-assessment is twelve questions you answer yourself, to understand where you stand. A formal assessment produces automation priorities and an execution roadmap as documents, based on real interviews and your operational data.

If our data is not organised at all, can we do nothing?

Not at all. Plenty of automation needs no data preparation. Work that takes its input on the spot — summarising a meeting, drafting a document — applies immediately regardless of the state of your internal data.

We need evidence to persuade leadership.

A readiness score is more persuasive with an estimate of time and cost saved attached. Using a tool that shows its calculation as a table means you can put it straight into a proposal.

Twelve questions, three minutes

Get the score first, then start

The free self-assessment gives you a score on each of the four axes plus your top three automation priorities. If you need the formal version, tell us your situation and we will review it and reply.

Get in touch