All resources
AI Readiness 8 min read

AI Readiness Assessment: What Good Looks Like

Readiness assessments fail when they list capabilities without judging them. Here is what a defensible readiness score actually looks like — and how the scoring should work.

Readiness is a stage, not a checkbox

AI readiness is a spectrum: Nascent → Developing → Emerging → AI-Ready. Most organizations are Developing or Emerging. Very few are actually AI-Ready. Pretending otherwise leads to funded initiatives that collapse in delivery.

Eight dimensions cover the real foundation

Strategy & vision, Data foundation, Architecture & integration, Semantics & knowledge, Governance & policy, Security & risk, Operating model, and Adoption & change. A readiness score that only covers Data and Technology is a technology audit, not a readiness assessment.

Score 1-5 per dimension, weighted equally

Equal 12.5% weighting keeps the score honest — no dimension can dominate. The weighted total lands in a band that maps to the readiness stage.

AI-READY: 4.0 and above

Board-approved AI strategy tied to business goals, governed data with lineage, an MLOps platform, canonical semantics, signed-off AI policy, AI-specific security controls, defined operating model, and an active adoption community.

EMERGING: 3.0 to 4.0

Most of the foundation is in place but with visible gaps — often semantics/knowledge and governance/policy. Emerging organizations can run pilots, but scaling to production stalls until the gaps close.

DEVELOPING and NASCENT: below 3.0

Fund foundation work before initiatives. Attempting production-grade AI when foundation gaps exceed 2 dimensions consistently produces expensive pilots that never scale.

What good governance looks like

Written AI policy signed off by Legal and Security, an approval process for production launches, a model registry, an evaluation harness, and a clear rollback path. Absent any of these, an AI-Ready score is aspirational, not real.

Turning the score into a roadmap

The lowest-scoring dimensions are usually the sequencing constraint. A 6-month foundation program that closes the two weakest dimensions typically unblocks 3+ downstream initiatives. Run the free /readiness assessment or buy the /readiness toolkit for the editable scorecard, findings doc and deck.

Put this into practice.

Run the numbers for your own initiative in the free AI ROI calculator.