Definition
An AI readiness assessment is a structured review of whether an organisation has the conditions it needs to create value from artificial intelligence: a clear ambition, usable data, the right technology, skilled and willing people, an operating model that can deliver, governance that keeps AI safe, and the discipline to measure results.
It is not an inventory of AI tools, and it is not a technology audit. The question it answers is a leadership question: if we decided to scale AI tomorrow, what would stop us?
A good assessment produces three things: a baseline score that the leadership team agrees on, a short list of the gaps that matter most for your ambition, and a sequenced plan to close them.
Why it matters now
Generative AI has made experimentation cheap. Any team can try a chatbot, a copilot or an agent in an afternoon. The result, in many companies, is a long tail of pilots that never reach production, unclear spending, and growing risk from tools nobody approved.
The constraint has moved. It is rarely the model or the vendor. It is ownership, data quality, process redesign, skills, and the board's ability to see what is working. Those are organisational questions, and they are exactly what a readiness assessment makes visible.
An assessment is most useful at three moments:
- Before a strategy or budget cycle, to decide where AI should create value and what has to be fixed first.
- When pilots are not scaling, to find out whether the blocker is data, technology, people or ownership.
- When the board asks "where do we stand on AI?" and the honest answer is "it depends who you ask".
What it measures: seven dimensions
Frameworks differ in their labels, but the serious ones cover the same ground. The AI Readiness Check uses seven dimensions in two families.
Foundations: the enabling conditions
- Data foundations. Is key business data accessible, trusted and owned? Can AI use your documents and company knowledge?
- Technology & platforms. Are core systems modern enough to connect to? Are approved AI tools available? Can you deploy and run AI at scale?
- Governance, risk & security. Is there a usage policy, a risk review before deployment, security controls and a clear view of the regulation that applies?
Differentiators: the leadership choices
- Strategy & ambition. Is AI tied to business goals, focused on a few domains, funded and on the board's agenda?
- People & culture. How many people use AI daily? Are there specialists, a training programme and real change management?
- Operating model. Does the business own AI initiatives? Are processes and roles redesigned? Are there rules for what AI may decide on its own?
- Use cases & value. Is there a prioritised portfolio, are use cases in production, is value measured and do successful pilots get scaled?
Splitting foundations from differentiators matters because two companies with the same overall score can need opposite programmes. One has strong data and platforms but no business ownership; the other has an energised leadership team building on weak foundations. The archetypes in the methodology describe these patterns.
Readiness vs maturity
The two terms are often used interchangeably. A useful distinction:
- Maturity describes where you are today, usually on a five-stage scale from exploring to AI-native.
- Readiness compares where you are with where you need to be. A company that only needs to be competent in AI is ready at a lower score than one whose strategy depends on leading its industry.
That is why the AI Readiness Check asks you to set an 18-month target for each dimension. Priorities are ranked by the gap to your own ambition, not only by your weakest scores.
How to run one with your leadership team
- Pick the respondents. Start with one senior leader who sees across functions: the CEO, COO, CIO or chief data officer. For a richer view, ask two or three members of the leadership team to complete it independently.
- Answer what is true, not what is planned. Good questions describe observable states, such as "a written policy that staff have acknowledged", rather than sentiment. If you are unsure, choose the lower option and flag low confidence.
- Set the target. For each dimension, agree the level the business needs in 18 months. This is where the strategy conversation happens.
- Compare and discuss. Where respondents disagree by more than one level, you have found a blind spot. That disagreement is often the most valuable output.
- Validate the priorities. Before committing budget, check the top three gaps against evidence: data samples, the list of pilots, the policy that exists or does not.
A self-assessment takes about ten minutes. An independent review, with stakeholder interviews and an evidence check, typically takes two to four weeks and is worth doing before a major investment.
Reading the results
Read the results in this order:
- The overall index tells you the stage you are at. It is a conversation starter, not a verdict.
- The archetype tells you which family of dimensions is holding you back, and therefore what kind of programme you need.
- The gaps to target are your priorities. Three is the right number: more than that and nothing moves.
- Peer and industry benchmarks tell you whether a gap is a competitive risk or simply where your sector is today.
The sample report shows each of these for a fictional company, including the 90-day roadmap and the board deck.
Common mistakes
- Treating it as an IT exercise. Technology is one dimension of seven. Most stalled programmes stall on ownership, data and adoption.
- Scoring the best team, not the company. One advanced analytics team does not make an AI-ready organisation.
- Stopping at the score. An assessment that does not change next quarter's priorities was a survey, not a diagnostic.
- Fixing everything at once. The point of ranking gaps is to choose. Start with the two or three that block the domains you care about most.
- Never repeating it. Re-run the assessment every six to twelve months to show the board progress against target.
Next steps
Start the assessment See a sample report