AI readiness assessment: what to prepare before you start

What an assessment is for

A readiness assessment looks at where automation or AI could help in your business, how ready your processes, data, systems and people are, and what should be tackled first. The output should be a ranked shortlist with rough effort and risk, not a sales proposal for everything.

You get far more from it if you arrive prepared. Most of the preparation is simply gathering what you already know in one place.

Pull together your pain points

Ask team leads a few direct questions and collect the answers in a simple list.

• Which tasks do people dread or put off?
• Where do things regularly get stuck, chased or redone?
• Which jobs involve copying information from one place to another?
• Where do customers wait the longest for a response?
• What would you stop doing by hand tomorrow if you could?

List your systems and data

Note the main systems involved in each pain point: accounting package, CRM, email, shared drives, line-of-business software, spreadsheets. For each, jot down who administers it, whether it has an export or interface, and roughly how clean the data is.

Where possible, bring a few anonymised samples of the documents or records involved, such as an invoice, an enquiry email or a spreadsheet extract. Remove personal data you do not need to share. Real examples reveal more than descriptions do.

Identify owners and decision makers

Every process needs someone who knows how it truly runs and someone with authority to change it. Name both, and make sure the people who do the day-to-day work can take part in the conversation. Include whoever looks after IT and data protection, as their input will shape what is feasible.

Also think about capacity. Even a good project needs time from your team for explanation, testing and feedback, so check that the right people can spare it.

Decide what success means

• Time: hours saved per week, or faster turnaround for customers.
• Quality: fewer errors, less rework, fewer complaints.
• Capacity: handling more work without adding headcount.
• Risk: better audit trails, consistency and compliance.
• Constraints: budget range, timing, systems that must not change.

What you do not need

You do not need a perfect data warehouse, a formal strategy document or an in-house data science team. Messy spreadsheets and half-written procedures are normal. The assessment is where those gaps are identified and sized. What matters is honesty about how work really happens, including the workarounds.

Frequently asked questions