AI Readiness Assessment: A One-Hour Score for Leadership Teams

The five lessons

If you read nothing else, read these.

1. Score six dimensions before you spend.
2. Each scores 0 to 3, out of 18. The band sets how narrow the first task must be.
3. A low score does not rule out AI. Start narrow and fix the lowest dimension first.
4. Teams skip measurement. With no baseline there is no ROI, only a feeling.
5. Score in an hour, with the people who run the task. Not with the vendor.

Why score readiness before choosing a tool

Most AI adoption trouble sits in the gap between the vendor demo and what you have: a documented process, clean data, an accountable person, a way to stop it and a number that proves it worked. The assessment finds that gap before money moves.

Score one candidate task, not the business. You can be ready to automate a weekly report and unready to automate customer replies, in the same week. For marketing tasks, see when AI belongs in marketing operations.

Put three people in the room: the manager, the person doing the work and whoever owns the number. Leadership alone scores too high, because it scores the process as written, not as run. Then test any outside help with the questions for choosing an AI consultant.

The six dimensions

Score the task 0 to 3 on each. Zero is absent. Three would survive a skeptical outside review.

1. Process definition. Would two people following a one-page write-up get the same result? 0 if it lives in one head. 3 if written and followed.
2. Data quality. Pull 50 records and check the fields this task reads. 0 if a third or more are wrong, missing or stale. 3 if reliably populated and current.
3. Ownership. 0 if it is a shared inbox or nobody. 3 if one named person reviews output and can turn it off.
4. Approval and risk controls. Does anything reach a customer or spend money unchecked? 0 if it goes straight out. 3 if every external action waits for a named approval.
5. Measurement baseline. Do you know, from running it by hand, how long it takes and how often it goes wrong? 0 if nobody measured. 3 if recorded within the last month.
6. Review capacity. Does someone have time each week to check a sample? 0 if the team has no slack. 3 if review is built into someone's week.

What each total means

Add the six for a total out of 18. The band describes this task only. Another task can land in a different band the same afternoon.

• 0 to 6, not ready. Do not pilot. Fix the lowest score and rescore. A pilot here makes an undefined process fail faster.
• 7 to 11, ready for a narrow pilot. Run the narrowest version, fix the weak dimension in parallel and gate every output.
• 12 to 15, ready with normal governance. Use a standard gate, sample after a full cycle and rescore before a second task.
• 16 to 18, ready to widen carefully. Add a second and third task, each with its own score.

The dimension that gets skipped: measurement

Teams score this one generously without evidence. It decides whether AI ROI is a number or an opinion.

Example: a team says the weekly report takes about a day. Timed over two weeks, it is 90 minutes on a normal week and closer to four hours when a client review is due.

If your answer is an estimate, score it 0 or 1 and spend a week timing it.

From score to adoption plan

Score the task, fix the lowest dimension, rescore, pilot, then measure against the baseline. Score the next task fresh. Readiness does not carry over.

This turns one large risk into small scored decisions. Three tasks over two quarters will usually beat one big rollout in a month, because each step rests on a measured result, not a demo. The AI implementation roadmap lays out the phases that follow. For help running them, see AI implementation for your business.

What low scores tell you

A low total is diagnostic, not disqualifying. Process at 0 means writing the page down, usually a day. Ownership at 0 means naming a person.

Approval and risk controls usually take longest, because you must agree what counts as an external action and who approves it. Start that conversation early. Governed AI operations explains how to set those gates.

What to check this week

Pick one candidate task. Time it by hand for a week, then score the six dimensions with the manager and the person doing the work.