quarterly-review

Quarterly Review

Purpose

Re-runs the fluency assessment, compares to the previous scorecard, identifies what moved and what didn't, and produces the next board update. This is the ongoing cadence skill — run it every quarter to keep adoption on track and the board informed.

Core principle: Measure the delta, not just the current state. The board wants to see trajectory, not a snapshot.

Flow

digraph review {
    "Retrieve previous artifacts" [shape=box];
    "Re-run fluency-assessment" [shape=box];
    "Compare scorecards" [shape=box];
    "Identify what moved, what didn't" [shape=box];
    "Update blocker report" [shape=box];
    "Produce comparison report" [shape=box];
    "Draft board update" [shape=box];
    "Route to next action" [shape=doublecircle];

    "Retrieve previous artifacts" -> "Re-run fluency-assessment";
    "Re-run fluency-assessment" -> "Compare scorecards";
    "Compare scorecards" -> "Identify what moved, what didn't";
    "Identify what moved, what didn't" -> "Update blocker report";
    "Update blocker report" -> "Produce comparison report";
    "Produce comparison report" -> "Draft board update";
    "Draft board update" -> "Route to next action";
}

Process

1. Previous scorecard is REQUIRED. If none exists, run `full-adoption-cycle` instead. 2. Ask ONE question at a time during the re-assessment. 3. Always show the comparison — previous score vs. current score, with what changed. 4. Do NOT inflate progress. If a score stayed flat, say so. 5. The board update must reference the delta, not just current numbers.

Step 1: Retrieve Previous Artifacts

Ask the founder for their previous scorecard, blocker report, and 90-day plan results. Summarize what you're comparing against:

"Let's see how things have moved since last quarter. I'll re-run the fluency assessment, compare it to your previous scores, and we'll draft your next board update. First — do you have your previous scorecard handy?"

Step 2: Re-Run Fluency Assessment

Run fluency-assessment as normal, but with two additions:

• For each pillar, ask "Last time you scored [X/5] here. Has anything changed?" before the detailed questions
• Keep the re-assessment tighter — focus on what's different, not re-establishing baseline

Step 3: Compare Scorecards

Produce a side-by-side comparison:

Pillar	Previous	Current	Change
Psychological Barriers	X/5	X/5	+X / -X / =
Integration Failures	X/5	X/5	+X / -X / =
Ownership Gaps	X/5	X/5	+X / -X / =
Overall	X/5	X/5	+X / -X / =

Step 4: Identify What Moved and What Didn't

For each pillar, explain the change (or lack of change) in one sentence:

• Improved: What specific action caused the improvement
• Flat: Why it didn't move — was the action taken but ineffective, or was nothing done?
• Declined: What went wrong — this is the most important finding

Step 5: Update Blocker Report

Review the previous blocker report:

• Resolved blockers: Mark as resolved, note what fixed them
• Persistent blockers: Still there — escalate severity if no action was taken
• New blockers: Anything that emerged since last quarter

Step 6: Produce Comparison Report

Use the Output format below.

Step 7: Draft Board Update

Use board-ai-update template to draft the quarterly update. The "What Happened" section must reference the delta — "Psychological barrier score improved from 2/5 to 3/5 after we..." not just "Our psychological barrier score is 3/5."

Anti-Patterns

Celebrating Flat Scores

Symptom: "Scores stayed the same — that's stable!" Consequence: Flat scores after a quarter of effort mean the effort didn't work. The board will notice. Fix: Flat = something didn't work. Diagnose why. Was the action taken? Was it the wrong action? Was there no action?

Ignoring New Blockers

Symptom: Only reviewing previous blockers without looking for new ones. Consequence: Adoption surfaces new problems as it progresses. Scaling creates friction that didn't exist at pilot scale. Fix: Always ask "What new challenges have come up since last quarter?" New blockers are expected — missing them is not.

Comparing to Ideal Instead of Previous

Symptom: "You're at 3/5, you should be at 5/5." Consequence: Demoralizing. Ignores real progress. Fix: Compare to the previous scorecard. "You went from 2/5 to 3/5 — here's what moved it, here's what to do next to get to 4."

Missing Adoption Trade-offs

Symptom: Scores improved but the founder doesn't mention that pushing adoption created new friction — code quality concerns, team morale dips, or engineers feeling surveilled. Consequence: The cure causes a new disease. Next quarter's blockers are side effects of this quarter's push. Fix: Always ask: "Did pushing adoption create any new tension — with quality, morale, or team autonomy?" If yes, log it as a new blocker. A score that went up by forcing behavior isn't real progress.

Output

Produce the quarterly review in this exact format:

## Quarterly AI Adoption Review
**Company:** [name] | **Period:** [Q and year] | **Date:** [date]
**Previous review:** [date or "initial assessment"]

### Score Comparison

| Pillar | Previous | Current | Change | What Happened |
|--------|:--------:|:-------:|:------:|---------------|
| Psychological Barriers | X/5 | X/5 | [+/-/=] | [One sentence] |
| Integration Failures | X/5 | X/5 | [+/-/=] | [One sentence] |
| Ownership Gaps | X/5 | X/5 | [+/-/=] | [One sentence] |
| **Overall** | **X/5** | **X/5** | **[+/-/=]** | |
| Team composition | [X] people | [Y] people | [+/-X] | [If changed: e.g., "3 roles eliminated, 1 AI-focused role created. Score changes partly reflect who left, not just behavior change." If stable: "No change." If restructuring status transitioned (e.g., in-progress → completed): note the transition.] |

### Blocker Status

| Blocker | Previous Status | Current Status | Action Taken |
|---------|:--------------:|:--------------:|-------------|
| [Blocker 1] | [severity] | Resolved / Persistent / Escalated | [What was done] |
| [Blocker 2] | [severity] | Resolved / Persistent / Escalated | [What was done] |
| [New blocker] | New | [severity] | [Identified this quarter] |

### What Worked
[2-3 bullets: specific actions that produced measurable results]

### What Didn't Work
[2-3 bullets: actions that were taken but didn't move scores, or actions that weren't taken]

### Next Quarter Focus
[The single highest-leverage action for next quarter, with named owner and target metric]

### Board Update Draft
[Use board-ai-update template — include the delta narrative]
[If team composition changed: add this line to the board update:]
Note: team composition changed this quarter. Adoption metrics reflect both behavior change and structural change.

Next Skill

Situation	Recommended next skill
New blockers emerged	blocker-diagnosis on the new blockers
Need a new use case for next quarter	first-use-case-picker
Need to rebuild the plan	90-day-plan-builder
Need to rehearse for the board	board-narrative-coach
Default	Founder is prepared — review complete

References

• fluency-assessment — re-run as part of the quarterly review
• blocker-diagnosis — previous blocker report is compared against
• adoption-scorecard — provides current adoption data for the comparison
• board-ai-update — template for the board update section
• full-adoption-cycle — first-time process; quarterly-review is the ongoing cadence