AI readiness assessment
What is an AI readiness assessment?
Most readiness assessments check whether you can build AI. The more useful question is whether you can decide which AI to build, because that is where the money is actually lost.
The short answer
An AI readiness assessment evaluates how prepared an organisation is to create value from AI. Traditional assessments score capability, data, talent, infrastructure, and governance. The more decisive readiness question is whether you can choose the right bets: whether you can compare AI initiatives on evidence, fund the few that will pay off, and stop the rest before they waste capital.
- It measures preparedness to create value from AI, not just the ability to deploy it.
- Capability readiness covers data, talent, infrastructure, and governance.
- Decision readiness covers whether you can rank AI bets and fund them on evidence.
- Most AI value is lost at the decision layer, not the capability layer.
Two kinds of AI readiness
Capability readiness is what most assessments measure: do you have clean data, the right talent, the infrastructure to deploy models, and the controls to run them safely? It matters, and it is well covered by existing maturity models.
Decision readiness is the half that gets skipped, and it is where the money leaks. Can you see every AI bet on one page? Can you compare them on evidence rather than enthusiasm? Can you stop a losing bet without a political fight? An organisation can be highly capable and still waste its AI budget because it cannot make these calls.
Why decision readiness matters most
The evidence points squarely at the decision layer. MIT found 95% of enterprise generative-AI pilots produce no measurable P&L impact, and McKinsey found that while 88% of organisations use AI, only 39% see any enterprise EBIT impact. Those are not capability failures. Plenty of capable teams are on the wrong side of them.
They are prioritisation and governance failures: funding bets that had no evidence, and leaving capital in ones that were never going to pay off. Readiness that ignores the decision layer measures the wrong thing.
What a useful AI readiness assessment covers
A readiness assessment worth running looks at both halves:
- Portfolio visibility: can you list every AI bet, live and proposed, in one place?
- Evidence standard: do bets need behavioural proof before they earn significant capital?
- Scoring discipline: are bets compared on the same criteria, or judged one pitch at a time?
- Kill discipline: are there pre-agreed thresholds to stop a bet, and are they used?
- Capability basics: data, talent, infrastructure, and governance controls.
From readiness to a funded decision
A readiness assessment tells you where the gaps are. The next step is acting on them for your actual portfolio. That is what an AI Bets Audit does: in two weeks we map and rank your AI bets, pretotype the highest-stakes ones for real evidence, and hand the board a fund / fix / kill call on each.
It draws on the same method, pretotyping, created at Google and taught at Stanford, that we have run across 4,000+ enterprise experiments. Readiness becomes a decision you can defend, not a score in a slide deck.
Sources
- 95% of enterprise generative-AI pilots produce no measurable P&L impact. MIT NANDA, State of AI in Business 2025
- 88% of organisations use AI, but only 39% report any enterprise-level EBIT impact. McKinsey, The State of AI 2025
- More than 80% of AI projects fail, roughly twice the failure rate of non-AI IT projects. RAND Corporation, 2024