AI readiness
How ready is the business for AI?
Eighteen questions, six dimensions, a score out of 36, and a plain answer about what to build first.
The score
Most AI projects fail before the first line of code, on data nobody can reach, a process nobody wrote down, or a budget nobody agreed. This measures those six things in five minutes and tells you where to start. No account, no call, no sales deck.
- 18 questions
- 6 dimensions
- Under five minutes
The eighteen questions
Data
Do you have it, is it clean, can you reach it.
- Customer and job records live in one system, not in spreadsheets and inboxes.
- Someone could pull last quarter's numbers in an hour without asking a developer.
- The records are clean enough to trust: no duplicates, no columns nobody fills in.
Systems
Where the work actually lives, and whether those systems can talk.
- The main systems, CRM, accounts, email, calendar, can talk to each other or have an API.
- Staff do their work inside those systems rather than around them.
- You know who owns each system and could change a setting in it this week.
Process
What repeats, what is written down, what is measured.
- The work that repeats every week is written down, step by step.
- You can name the three tasks that eat the most hours.
- At least one of those tasks is measured: how long it takes, how often it goes wrong.
People
Who owns it, who will use it, who wants it.
- One named person would own an AI project and has the time to.
- The team uses the tools it is given rather than working around them.
- The owners want this. It is not one enthusiast's idea.
Money
A budget, a payback you would accept, and the grants you know about.
- There is a budget for a first project, even a small one.
- You would fund a project that pays back inside a year.
- You know which grants and tax credits the business qualifies for.
Guardrails
Personal data, sign off, and the decisions that stay human.
- You know which of your data is personal data and where it is held.
- Someone signs off before a system sends anything to a customer.
- You can name the decisions that must stay with a person.
The four bands
- AI Starter, 0 to 11: The foundations are not there yet, and that is useful to know before spending anything. The first project is not an agent. It is one system of record and one written process, and both are funded.
- AI Explorer, 12 to 20: Parts are in place and parts are not. The right first project is small, inside the strongest dimension, and it should pay back within the year so the second one funds itself.
- AI Builder, 21 to 29: The business is ready for a real first agent. The work now is choosing the task that eats the most hours and scoping it tightly enough that a grant covers most of the build.
- AI Leader, 30 to 36: The foundations are solid. The question is no longer whether, but which three tasks first and in what order, and whether the funding is lined up before the build starts.