MEMO · TO leaders scoping a first AI deployment · RE a five-dimension organisational score
Resources Organisational AI Readiness Quiz
Organisational AI Readiness Quiz
17 statements across five dimensions, Data, Process, Technology, People and Governance. Rate how true each one is of your organisation today and get a score out of 85, a per-dimension breakdown, a tiered roadmap, and a branded PDF report.
Why this matters
Five dimensions, not one, decide whether AI sticks
A five-question check is a fast diagnostic. A seventeen-statement one is closer to an audit: it catches the gap that hides between two dimensions that each look fine on their own, a well-governed process running on ungoverned data, or a capable team blocked by a stack nobody mapped.
That resolution matters because the dimensions do not fail independently. A business can score well on People and Governance and still stall in deployment because Technology integration was never scoped, or score well on Data and Process and still take on real risk because nobody wrote down what happens when the system is wrong.
This quiz scores all five at once, so the roadmap it returns is sequenced by where the business actually stands, not by which dimension is easiest to talk about first.
Exhibit · What's actually missing
Adoption outpaces the governance and integration work behind it.
-
78%
of companies use AI frequently, per McKinsey
-
33%
have scaled it company-wide, per McKinsey
-
63%
of gen-AI users have no governance structure for the risks, per McKinsey
-
~60%
cite legacy-system integration as a top AI adoption challenge, per Deloitte
Your result
Your five-dimension shape
Section 02 · What your score means
Section 03 · The business impact
Your immediate action plan · next 14 days
Grounded in
95%
of generative AI pilots fail to produce a measurable profit-and-loss result
MIT NANDA, The GenAI Divide, 2025
60%
of AI projects fail on poor data foundations, and over 40% of agent projects fail on legacy-system incompatibility, both detectable before a dollar is spent
Gartner, 2025
74%
of companies have yet to show tangible value from their AI investment, despite the spend behind it
BCG, Where's the Value in AI?, October 2024
33%
of companies have scaled AI company-wide, against 78% who use it frequently, the gap this framework exists to close
McKinsey, The State of AI, 2024
Section 04 · Score by dimension
Each statement is rated 1 (strongly disagree) to 5 (strongly agree); the dimension score is the sum across its statements. Status below reflects the dimension's score as a share of its own maximum: under 40% is a deployment blocker, 40–70% is developing, above 70% is on track.
Definition — Whether the data behind your top AI use cases is trustworthy, current, accessible, and clearly owned.
Why it matters — Every other dimension depends on this one. If the underlying data cannot be trusted, nothing built on top of it, including governance and measurement, can be trusted either.
Research 60% of AI projects will be abandoned by 2026 for lack of AI-ready data, Gartner projects. Gartner, February 2025.
Definition — Whether the process you'd point AI at is documented, measured against specific outcomes, and open to being redesigned rather than just automated as-is.
Why it matters — AI automates what is defined. A process layered with AI without being documented or measured usually repeats its own defects, faster and at volume.
Research 2 in 3 leaders call their own organisation overly complex, and simplifying the process is the bigger lever, McKinsey finds, not layering technology onto it unchanged. McKinsey, The State of Organizations 2026.
Definition — Whether your systems and IT function can integrate an AI tool without a long, custom build, including API access, stack visibility, a safe sandbox, and early security involvement.
Why it matters — Integration effort is often the hidden cost line in an AI project. Legacy systems that were never built to expose data cleanly turn a tool evaluation into a systems-integration programme.
Research ~60% of organisations cite integrating AI with legacy systems as one of their primary challenges to AI adoption. Deloitte, AI adoption pulse survey.
Definition — Whether ownership, hands-on exposure to how AI tools work and fail, and leadership alignment exist beyond a single enthusiast championing the effort alone.
Why it matters — A deployment nobody in the organisation can operate, question or correct is a dependency on the vendor, not a capability the business has actually built.
Research 35% of staff received AI training in the past year, even though 75% of companies are adopting AI. Randstad, AI Skills Gap, November 2024.
Definition — Whether a written policy, a review step before output reaches a customer or decision, and a way to detect unexpected behaviour all exist before deployment, not after an incident.
Why it matters — Governance determines how much damage an error does before someone notices it. Writing down what happens when the system is wrong, and who is accountable, has to happen before launch to matter.
Research 63% of companies using generative AI have no governance structure in place to manage the risks it creates. McKinsey, The State of AI, 2024.
Section 05 · What to do in each dimension
Data
Every other dimension compounds on top of data. If this is your weakest score, fixing it unlocks the rest faster than working on anything else first.
Process
AI layered onto an undocumented or unmeasured process usually just repeats the defect, faster and at volume. Fix the process before you automate it.
Technology
Integration effort is often the hidden cost line in an AI project. Get a clear view of your stack and a sandbox to test in before committing to a tool.
People
A deployment nobody in the organisation can operate, question or correct is a dependency on the vendor, not a capability of the business.
Governance
Write down what happens when the system is wrong, and who is accountable for catching it, before it goes live, not after.
Reference
This quiz is adapted for Praxora Lab from the original Organisational AI Readiness Quiz at terencekok.com/resources › . For the shorter, five-question version, see the AI Readiness Self-Assessment ›.
