Score · Grow with AI
How the audit scores each area
Six areas, scored zero to four from your own evidence. AI scores each area against this rubric and names the rung and the evidence it relied on. A score stays fixed until you change that area's answers.
Buy ScoreThe six areas
Each area maps to one question in the audit intake: your top objective, your highest-friction workflow, how far you trust current measurement, your approved AI tools, your review path, and your team size.
| Area | What it looks at |
|---|---|
| Objectives | Your top objective and its success metric |
| Workflow clarity | Your highest-friction marketing workflow and its rework signals |
| Measurement trust | How far you trust your current measurement |
| AI readiness | Approved AI tools and current use |
| Governance | Your review path before expanding AI use |
| Execution capacity | Your team size and constraints |
Score, zero to four
Each score comes from the length of your own answer for that area, after trimming whitespace. A longer, more specific answer scores higher. An empty or "unknown" answer scores zero rather than a middling default.
| Your answer | Score |
|---|---|
| Empty, or the word "unknown" | 0 |
| 1 to 19 characters | 1 |
| 20 to 59 characters | 2 |
| 60 to 119 characters | 3 |
| 120 characters or more | 4 |
Evidence status
Evidence status is computed from the same answer, independently of the score. A short, confident answer can score low but still read as "Evidenced." A long answer that hedges still reads as "Assumed" even at a high score.
| Your answer contains | Evidence status |
|---|---|
| Nothing, or is "unknown" | Not evidenced |
| A hedge word, such as "assumed" or "undocumented" | Assumed |
| A partial-confidence word, such as "manual" or "varies" | Partly evidenced |
| None of the above | Evidenced |
Review flag
The report sets a review flag when either of these holds, regardless of the other four areas:
- Governance scores exactly 1.
- Measurement trust scores exactly 0.
Both point to the same underlying risk: expanding AI use on top of an ungoverned or unmeasured workflow. The flag adds one line to your report. It never blocks the report or the PDF export.
Priorities, at most three
The report drops every area already scoring 4, since there is nothing to prioritize there. It sorts what remains by score, lowest first, and keeps the fixed area order above to break ties. It takes the first three. If every area already scores 4, the report states that no priority is required this cycle.
What this rubric does not do
It does not verify your claims or inspect your systems. The AI judges only what you write in your answers and notes. The rules in this rubric also produce a score for each area. If the AI score differs from it by more than one point, the area is flagged for your review, and without AI credits you see the rule score marked provisional. It is a decision aid, not an audit in the assurance sense. A competent person must review every score and priority before anyone acts on it. See Responsible AI and use of outputs in the Terms.