FREE TOOL

AI suitability quiz

Twelve questions about where you would like AI to help, what shape your data and processes are in, how much is at stake when it is wrong, and who would own it. The result tells you which of your areas are suited to AI now, which need groundwork first, which are better served by ordinary automation, and where to start. No prices and no promised savings.

HOW IT DECIDES

Data, written processes and an owner decide, not the model

Each area gets its own verdict for a reason you can read next to it.

Data in systems with APIs, consistent records and written steps raise readiness; information in people's heads and messy data lower it.
Rule-based work such as data entry and reporting from systems gets 'ordinary automation': cheaper and exact.
Documents, email, knowledge, requests, sales and content get AI with review when their inputs are usable, and groundwork when they are not.
Regulated data and direct customer contact never combine; high error cost adds logging and rollback.

WHY MONTHLY CARE

AI output drifts; someone has to notice

Products, policies, customer language and the models themselves change.

Weekly sampled checks catch wrong outputs before customers or decisions do.
Source and prompt updates keep answers current; stale ones are retired.
Adoption review shows what people correct or bypass, which is where the next improvement is.
Cost control and a monthly report keep usage and value visible.

HONEST LIMITS

What this quiz does not do

It does not promise savings. How much AI helps depends on process clarity, source quality, access design and review points, and we assess that together.
It does not give a price. Any number without seeing your data and systems would be a guess.
It does not pick tools or models. The fit of the task comes first; the tool is chosen for the task, not the other way round.
It does not replace the first conversation, where we look at real examples with you and the first pilot usually becomes smaller and sharper.

Questions about the quiz

Why does it say 'ordinary automation' for some areas?

Because if a stable rule, an exact formula or a fixed field mapping can produce the result, ordinary automation is cheaper and easier to verify than a model. Data entry between systems and reports from system data are the usual cases. We tell you so instead of selling AI where it does not belong.

Can we start with AI even if the quiz says groundwork first?

You can, but the pilot will fail for reasons that have nothing to do with AI: nobody can say what 'correct' is, the data contradicts itself, nobody owns the corrections. The groundwork takes weeks, not months, and makes the pilot cheap.

What about the EU AI Act and GDPR?

For the use cases here, mostly transparency when AI interacts with people, human oversight where output has real impact, and the usual data protection duties: legal basis, data minimisation, a processing agreement with the provider. We design those in before the first real record is processed.

What happens to my answers here?

They stay in your browser until you choose to send the result through the contact form. We do not store questionnaire answers otherwise.