Answer · AI Transformation
AI Readiness Assessment: What It Is and How to Run One
Aiden Wayne · Updated August 17, 2026
Short answer
What it is measuring
Most AI initiatives do not fail on model quality. They fail because the work the model was supposed to improve was never mapped, the data it needed lived in three systems with different definitions, or nobody owned the decision the model was meant to inform. A readiness assessment measures the conditions around the model, not the model.
- Workflow clarity — can you draw the process end to end, with handoffs and rework loops named?
- Data availability — does the data exist, is it accessible, and does one field mean one thing?
- Decision ownership — who acts on the output, and do they have authority to act differently?
- Tooling and integration — can a result reach the system where work actually happens?
- Skills and capacity — who maintains this in month six, and is that role funded?
- Governance and risk — what is the review path for customer-facing or regulated output?
The six-dimension score
Score each dimension for the specific workflow you are considering, not for the company as a whole. Company-level readiness scores are comforting and useless; a company can be ready for AI-assisted support triage and completely unready for AI-assisted pricing.
| Score | What it means for that workflow |
|---|---|
| 0–1 | Blocked. Fix the underlying condition before any AI spend. |
| 2–3 | Workable with effort. Sequence the fix into the implementation plan. |
| 4–5 | Ready. This dimension will not be the thing that stalls you. |
How long a credible assessment takes
One workflow can be assessed properly in about a week. A whole operating area — support, onboarding, revenue operations — takes one to two weeks with interviews, system walkthroughs, and a data spot-check. Anything sold as a same-day AI readiness score is a lead-generation quiz, not an assessment.
The output should be short: a ranked opportunity list, a readiness score per dimension, the blockers, and a 30/60/90-day sequence. If the deliverable is a fifty-slide deck with no sequence in it, it will not survive contact with a quarter.
How to run one internally
- Pick the three workflows leadership already complains about. Do not survey the whole company.
- For each, interview the person who does the work — not the person who owns the budget.
- Draw the workflow on one page, including rework loops and manual re-entry.
- Mark every point where a human is transcribing, reconciling, summarising, or chasing.
- Estimate frequency and time cost per point. Rough is fine; consistent is what matters.
- Score the six dimensions per workflow, then rank by value ÷ effort, not by AI novelty.
- Write the 90-day sequence before you talk to a single vendor.
Assessment versus maturity model
An AI maturity model tells you where you sit on a curve relative to other companies. That is useful for a board slide and largely inert operationally. A readiness assessment is workflow-specific and ends in an action sequence. If you have to pick one, pick the one that names what you will do in the next 90 days.
Frequently asked
- What is an AI readiness assessment?
- A structured review of whether a company can absorb AI into how it operates — scoring workflow clarity, data availability, decision ownership, tooling, skills, and governance for a specific workflow, and ending in a ranked opportunity list with a 90-day sequence.
- How long does an AI readiness assessment take?
- About a week for a single workflow, and one to two weeks for a whole operating area including interviews, system walkthroughs, and a data spot-check.
- Who should be involved?
- The people who do the work daily, the person who owns the decision the AI would inform, and whoever controls the systems the output has to land in. Budget owners should see the result, not shape the input.
- Do we need clean data before we start?
- No. The assessment is partly how you find out which data is clean enough for which use. Waiting for clean data before assessing readiness is the most common reason companies stall for a year.