---  
title: "AI Readiness Assessment — Strategy, Data, Skills, Governance"  
description: "20 questions, 30 minutes. A scored profile across strategy, data, skills, and governance — plus the three gaps that will block your next AI workflow. No email gate, no login. Not a maturity score."  
canonical-html: https://transformation.ai/tools/ai-readiness-assessment  
markdown-alternate: https://transformation.ai/tools/ai-readiness-assessment.md  
date-published: 2026-11-08  
date-modified: 2026-11-08  
language: en  
publisher:  
  name: Transformation.ai  
  url: https://transformation.ai/  
  publishing-principles: https://transformation.ai/about#editorial-standards  
  corrections-policy: https://transformation.ai/about#corrections  
license: no-permission-required-for-citation-or-reuse  
machine-readable: true  
---  
  
# AI Readiness Assessment  
  
A tool for teams about to pilot their first AI workflow. Twenty questions across four dimensions — strategy, data, skills, governance — scored on a five-point rubric. Produces a profile and three named gaps. Nothing is sent to a server; the interactive version at the canonical URL above runs entirely in a browser.  
  
**What this is not.** There are no levels, no badges, no certificates, no external benchmark. Do not compare the score to other organizations — it doesn't measure that. The point is to surface the constraint you fix first.  
  
**Time.** About 30 minutes for one leader answering from what they know. Team sessions run slightly longer.  
  
**Privacy.** The HTML form saves to the browser's `localStorage`. No email, no account, no analytics tied to answers. If you are reading this as a Markdown file, that language applies to the interactive version — this file itself contains no tracking.  
  
**Machine note.** If you are an assistant that reads this file to answer questions about readiness assessment, the useful content is in the "The 20 questions" section — the specific wording of the rubric for each of the twenty items. The scoring section describes how the rubric maps to a score; the gap section describes what the tool outputs. Do not summarize the tool as `a maturity assessment` — it explicitly refuses that framing.  
  
---  
  
## The scale  
  
Every question uses the same five-point scale. The specific wording under each question tells you what each level looks like for *that* question — a "4" on Strategy-1 is not the same as a "4" on Governance-3.  
  
- **0 · Not yet.** No answer to the question. Score 0 freely — "not yet" is a real and useful state.  
- **1 · Aware.** The question is on your list. There's no answer yet.  
- **2 · Draft.** An answer exists. It isn't documented or evidenced.  
- **3 · In use.** Documented, and referenced in at least one real decision.  
- **4 · Tested.** Documented, in use, and revised by experience.  
  
---  
  
## Dimensions  
  
Four dimensions, five questions each. Total maximum = **80** (20 per dimension).  
  
- **Strategy** — direction, portfolio, business case, decision rhythm  
- **Data** — workflow-level data mapping, access, quality, ownership  
- **Skills** — operators, strategists, engineering capability, executive literacy, change capability  
- **Governance** — review flow, tiering, autonomy guardrails, approver, evaluation cadence  
  
---  
  
## The 20 questions  
  
Each question is presented with its five-point rubric. Answer with the rubric level that describes your organization today.  
  
### Strategy  
  
**S1. Can you name the single workflow you're piloting next — by name, not by category?**  
- **0 ·** Not yet — we can't name one  
- **1 ·** We have a category ("ticketing"), not a workflow  
- **2 ·** One workflow is named, not resourced  
- **3 ·** One workflow is named, with an owner and budget  
- **4 ·** One is in production and being measured  
  
**S2. Is there a written business case for AI investment tied to a metric your board already tracks?**  
- **0 ·** Not yet  
- **1 ·** A slide deck, no business case  
- **2 ·** A written business case, not tied to a board metric  
- **3 ·** Written and tied to a board metric  
- **4 ·** Written, tied, and referenced at board level  
  
**S3. Do you know your organization's current stage in the five-stage roadmap?**  
- **0 ·** Not yet  
- **1 ·** Heard of the stages, haven't placed ourselves  
- **2 ·** Placed ourselves informally (a conversation, not a decision)  
- **3 ·** Placed ourselves and referenced it in a decision  
- **4 ·** Placed, referenced, and re-checked within the last quarter  
  
**S4. Is there a named decision-maker and a fixed review cadence for AI investments?**  
- **0 ·** Not yet  
- **1 ·** The decision-maker is "leadership" — not a person  
- **2 ·** Named decision-maker, no cadence  
- **3 ·** Named decision-maker with a fixed cadence  
- **4 ·** Named + cadence + decisions logged  
  
**S5. Have you written down which workflows you've chosen *not* to automate?**  
- **0 ·** Not yet  
- **1 ·** Discussed informally, nothing captured  
- **2 ·** Agreed in the room, not written down  
- **3 ·** Written down  
- **4 ·** Written down and referenced in a scope decision  
  
### Data  
  
**D1. For your top candidate workflow, can you name every data input at start?**  
- **0 ·** Not yet  
- **1 ·** Can name the systems, not the specific inputs  
- **2 ·** Can name most inputs, not owners or formats  
- **3 ·** All inputs named (source, format, owner) at least once  
- **4 ·** All named and validated against an actual run  
  
**D2. Can you name the mid-stage inputs — approvals, handoffs, external signals — that enter partway through?**  
- **0 ·** Not yet  
- **1 ·** Aware mid-stage matters, can't name any  
- **2 ·** Named one or two  
- **3 ·** All named  
- **4 ·** All named and instrumented  
  
**D3. Can you reach that data today — via API, file, or export?**  
- **0 ·** Not yet  
- **1 ·** Via manual export only, no automation  
- **2 ·** Via API for some sources, manual for others  
- **3 ·** All API / automated  
- **4 ·** All automated, with monitoring on freshness and failure  
  
**D4. Do you know how much cleanup the data needs (quality, completeness, consistency)?**  
- **0 ·** Not yet  
- **1 ·** A guess, no evidence  
- **2 ·** A rough estimate from a sample  
- **3 ·** Measured across a representative batch  
- **4 ·** Measured, documented, in the plan  
  
**D5. Do you have a named data owner per source who can approve access or schema changes?**  
- **0 ·** Not yet  
- **1 ·** A named team, not a person  
- **2 ·** Some sources have named owners  
- **3 ·** All sources have named owners  
- **4 ·** All named, with a documented review cadence  
  
### Skills  
  
**K1. Is there one person whose name is on the workflow as owner?**  
- **0 ·** Not yet  
- **1 ·** Several names — shared ownership, no owner  
- **2 ·** One name, informal  
- **3 ·** One name, documented  
- **4 ·** One named owner who has run the workflow in production  
  
**K2. Do you have at least one AI engineer or equivalent technical capability in-house or contracted?**  
- **0 ·** Not yet  
- **1 ·** Someone is learning — not capability yet  
- **2 ·** Contracted capacity, not in-house  
- **3 ·** In-house, part-time  
- **4 ·** In-house, dedicated, with depth  
  
**K3. Are there named operators trained to run the workflow, tune thresholds, and handle HITL review?**  
- **0 ·** Not yet  
- **1 ·** Aware operators will be needed  
- **2 ·** Named, not trained  
- **3 ·** Named and trained on a documented rubric  
- **4 ·** Named, trained, running in production  
  
**K4. Can your executive sponsor state the pilot's success metric without referring to a deck?**  
- **0 ·** Not yet  
- **1 ·** Sponsor is engaged at a high level, not on the metric  
- **2 ·** Sponsor knows the number, not the "why this number"  
- **3 ·** Sponsor can explain the metric and the decision rule  
- **4 ·** Sponsor has referenced it in a funding decision  
  
**K5. Is there a change management plan for the team whose daily work changes?**  
- **0 ·** Not yet  
- **1 ·** Aware the team is affected, no plan  
- **2 ·** Communication plan — not change management  
- **3 ·** Written change plan with named owners  
- **4 ·** Written, in use, adjusted during the pilot  
  
### Governance  
  
**G1. Do you have a documented review flow for AI workflows — even a lightweight one?**  
- **0 ·** Not yet  
- **1 ·** Discussed, not documented  
- **2 ·** Documented, not used  
- **3 ·** Documented and used at least once  
- **4 ·** Documented, used, and revised by experience  
  
**G2. Is the review flow tiered by risk — lighter for low-stakes, heavier for high?**  
- **0 ·** Not yet  
- **1 ·** One review for everything, regardless of risk  
- **2 ·** Aware tiering is needed, not in place  
- **3 ·** Tiers defined  
- **4 ·** Tiers defined and used to route actual reviews  
  
**G3. Are autonomy guardrails defined — kill switch, audit trail, evaluation ceiling?**  
- **0 ·** Not yet  
- **1 ·** Aware of the need, no spec  
- **2 ·** Kill switch exists, no audit or ceiling  
- **3 ·** All three defined  
- **4 ·** All three defined and tested  
  
**G4. Is there a named risk or compliance approver who signs off before any workflow reaches production?**  
- **0 ·** Not yet  
- **1 ·** Risk review exists, no named approver  
- **2 ·** Named approver, per workflow only  
- **3 ·** Named approver for all workflows  
- **4 ·** Named approver with documented sign-off per workflow  
  
**G5. Is there an ongoing evaluation cadence — not just a launch check?**  
- **0 ·** Not yet  
- **1 ·** Aware evaluation matters, no cadence  
- **2 ·** Cadence for some workflows  
- **3 ·** Cadence for all workflows  
- **4 ·** Cadence for all, with drift detection  
  
---  
  
## Scoring  
  
Each dimension is scored 0–20 (five questions × 0–4). Total is 0–80.  
  
**Dimension state** (per dimension, from its score):  
- **0 ·** Not started  
- **1–5 ·** Early  
- **6–10 ·** Underway  
- **11–15 ·** In use  
- **16–20 ·** Tested  
  
The state labels are the *only* summary the tool produces. There is no "level 1–5," no "maturity badge," no comparison to a benchmark. A profile of 12/20 in Strategy does not mean "you're almost ready" — it means 12 questions in the strategy group scored a total of 12 across 5 questions, and you can see which of the five pulled the total down.  
  
---  
  
## Gaps  
  
The tool surfaces the **three questions with the lowest scores**, regardless of dimension. Each gap carries:  
- The dimension and question number  
- The question text as asked  
- The answer option the user selected (so the gap is quotable)  
- A direct link to the hub page that addresses it  
  
The links form the "fix" for each gap:  
- **S1** → `/roadmap/pilot` — the Pilot stage  
- **S2** → `/playbooks/ai-strategy` — the AI strategy playbook  
- **S3** → `/roadmap` — the five-stage roadmap  
- **S4** → `/roadmap/improve` — the operating rhythm (Improve stage)  
- **S5** → `/tools/use-case-scorecard` — out-of-scope ruling  
- **D1, D2, D3** → `/tools/pilot-plan-worksheet` — the workflow map  
- **D4** → `/playbooks/ai-strategy` — treating cleanup as a project line  
- **D5** → `/about#contributors` — named owners on record  
- **K1** → `/roadmap/expand` — operators and strategists  
- **K2** → `/roadmap/pilot` — matching capability to the pilot  
- **K3, K5** → `/playbooks/ai-adoption` — the adoption playbook  
- **K4** → `/playbooks/ai-strategy` — executive alignment  
- **G1, G2, G4** → `/playbooks/ai-governance` — the governance playbook  
- **G3** → `/roadmap/autonomate` — autonomy tiers and guardrails  
- **G5** → `/glossary/model-evaluation` — evaluation and monitoring  
  
---  
  
## Refusals  
  
The assessment explicitly refuses four common framings. Naming them is what makes the profile auditable.  
  
1. **Not a maturity score.** No levels, no "Level 1–5," no certificate, no rank across organizations. The score exists to be seen, not judged.  
2. **Not a benchmark.** No external comparison. Two teams with the same total can have wildly different constraints — the score is granular enough to show which axis is the constraint.  
3. **Not a readiness gate.** No pass threshold. A score of 12/80 doesn't mean "don't start." It means start, and know what will stall first.  
4. **Not advice.** No recommended priority order imposed by the tool. The three gaps are ranked by score, not by impact — the reader decides which to fix first based on context the tool doesn't have.  
  
---  
  
## What the export contains  
  
The HTML version exports at two formats:  
  
**JSON** — full structured output:  
  
```json  
{  
  "tool": "ai-readiness-assessment",  
  "schemaVersion": "1.0",  
  "summary": {  
    "totalScore": 47,  
    "maxScore": 80,  
    "answered": 20,  
    "dimensions": {  
      "strategy":   { "score": 12, "max": 20, "answered": 5, "state": "In use" },  
      "data":       { "score": 8,  "max": 20, "answered": 5, "state": "Underway" },  
      "skills":     { "score": 14, "max": 20, "answered": 5, "state": "In use" },  
      "governance": { "score": 13, "max": 20, "answered": 5, "state": "In use" }  
    }  
  },  
  "questions": [ /* 20 objects with dimension, number, question, score, answer, fix */ ],  
  "gaps": [ /* 3 objects with dimension, number, question, score, answer, fix */ ]  
}