AIDA is a field instrument, not a maturity poster

AIDA is a field instrument, not a maturity poster

Most 'AI maturity' artifacts are posters. Five stages, a rainbow, a workshop. They cannot be wrong, which means they cannot be useful. AIDA was built the other way: six dimensions, a 0–100 score, and a forced ranking of the constraint. If governance is 36 and technology is 74, you do not buy another model. You write down who is accountable when the model is wrong.

The six are not arbitrary. They are the places programs actually die. Strategy & Vision: is there a board-aligned plan with owners, dates, and numbers — or a slide with aspirations? Data Infrastructure: can the data you already have support production, or only a demo? Talent & Capabilities: does the organization have the literacy, time, and permission to use AI, not just buy it? Technology Stack: will current systems, identity, and integrations carry a workload? Governance & Ethics: who is accountable when a model is biased, leaked, or simply wrong — and is that written down? Change Readiness: have past technology programs actually changed how people work?

Scoring is deliberately blunt. Three statements per dimension, 1–5. Average, scale to 100. The public assessment is directional — no email, five minutes, a radar. The contracted audit is evidentiary: interviews, documents, architecture, a 20–80 page report, a 12-month roadmap. Confusing the two is how firms overclaim. We do not.

Bands matter because they change the work. Nascent (0–35): do not fund a platform. Fund a baseline and a named owner. Emerging (36–55): pick one use case that returns cash or time in 90 days. Capable (56–75): sequence, govern, and stop collecting pilots. Ready (76–100): the risk is now complacency and vendor lock-in, not ignorance.

The instrument is proprietary because the statements, the bands, and the recommended moves are the practice. It is not proprietary because we want a trademark on common sense. You can steal the six names. You cannot steal the habit of treating the lowest score as the only score that matters.

Use the 5-minute version if you need a number for Monday. Use the audit if you need a number that will survive a board. Both exist so that 'we should do AI' stops being a sentence and starts being a sequence.