diagnostic-analyzer

Skill: Diagnostic Analyzer

A single diagnostic response tells you about one person. Ten responses tell you about a pattern. The analyzer surfaces what the pattern reveals — about your ICP, your messaging, your offer, and your pipeline.

When to Use

• After collecting 5+ responses on a single diagnostic
• During weekly review (via diagnostic-monitor-agent)
• Before a product build decision (product validation diagnostics)
• When pipeline quality feels off and you want data to explain why

---

Analysis Components

Component 1: Score Distribution

For each diagnostic, show:

## Score Distribution: [Diagnostic Name]
**Total responses:** [N] | **Period:** [date range]

| Band | Respondents | % | Avg score |
|------|-------------|---|-----------|
| 🟢 High (66–100) | [N] | [X%] | [avg] |
| 🟡 Medium (36–65) | [N] | [X%] | [avg] |
| 🔴 Low (0–35) | [N] | [X%] | [avg] |

**Overall average:** [X] / 100
**Median:** [X]
**Range:** [min] – [max]

What to flag:

• If > 60% are Low → your diagnostic is reaching the wrong people, or the problem isn't real enough
• If > 70% are High → your diagnostic may be too easy, or you're only reaching pre-qualified people
• If the distribution is trimodal (peaks in all three bands) → you likely have 3 distinct sub-segments

---

Component 2: Dimension Breakdown

For each dimension, show the average score across all respondents:

## Dimension Averages

| Dimension | Avg score | % of max | Interpretation |
|-----------|-----------|----------|---------------|
| [Dim 1] | [X]/[max] | [X%] | 🟢 Healthy across respondents |
| [Dim 2] | [X]/[max] | [X%] | 🟡 Mixed — worth watching |
| [Dim 3] | [X]/[max] | [X%] | 🔴 Consistently weak |
| [Dim 4] | [X]/[max] | [X%] | |

What consistently low dimensions reveal:

Diagnostic type	Low dimension	Likely interpretation
Lead	Problem Clarity	Your marketing is attracting too-early buyers — they haven't defined the problem yet
Lead	Budget Reality	Wrong ICP segment — targeting too small or too early-stage
Client	Goal Progress	Delivery gap — something about how you work needs examining
Client	Value Perception	Communication gap — results aren't being attributed to your work
Product	Willingness to Pay	Problem exists but isn't painful enough to pay for
Self	Differentiation	Positioning work needed before growing

---

Component 3: Segment Analysis

Group respondents by band and look for patterns within each group:

## High-Band Respondents ([N] people)

Common characteristics:
- [Dimension X] average: [X%] (highest across all segments)
- [Dimension Y] average: [X%]
- Most common answer on [key question]: "[Option]"

What this tells you: [2-3 sentence interpretation]

→ ICP match: [High / Medium / Low] — [brief reason]

## Low-Band Respondents ([N] people)

Common characteristics:
- [Dimension X] average: [X%] (lowest)
- Most common answer on [key question]: "[Option]"

What this tells you: [2-3 sentence interpretation]

→ These respondents need: [education / different service / different timing]

---

Component 4: Question-Level Heatmap

Show which specific questions are getting the strongest/weakest answers:

## Question Heatmap

| Question | Avg score | % picking score 0 | % picking score 4 |
|----------|-----------|-------------------|-------------------|
| [Q1 text abbreviated] | [X]/4 | [X%] | [X%] |
| [Q2] | | | |
[...]

**Most revealing questions (highest variance):**
[List questions where answers are most spread — these are the real discriminators]

**Least revealing questions (everyone scores the same):**
[Questions where 80%+ of respondents picked the same option — consider replacing]

---

Component 5: Pipeline Impact (Lead Diagnostics)

For lead diagnostics, connect score to pipeline outcomes:

## Pipeline Impact

| Cohort | Respondents | Proposals sent | Converted | Conversion rate |
|--------|-------------|----------------|-----------|----------------|
| 🟢 High band | [N] | [N] | [N] | [X%] |
| 🟡 Medium band | [N] | [N] | [N] | [X%] |
| 🔴 Low band | [N] | [N] | [N] | [X%] |

**Key finding:** [What the conversion data reveals about which band is worth pursuing]

**Recommendation:** [Adjust pipeline stage routing / change CTA for medium band / etc.]

Note: Requires cross-referencing response files with pipeline.md. Run automatically by diagnostic-monitor-agent weekly.

---

Component 6: Trend Analysis (Over Time)

When 3+ weeks of data exist, show trend:

## Score Trend Over Time

| Week | Responses | Avg score | Band distribution |
|------|-----------|-----------|-------------------|
| [W1] | [N] | [X] | 🔴[N] 🟡[N] 🟢[N] |
| [W2] | [N] | [X] | |
| [W3] | [N] | [X] | |

**Trend:** [Improving / Declining / Stable] — [interpretation]

If improving: Your diagnostic is reaching a better-fit audience over time,
or respondents' actual situations are improving.

If declining: Traffic source quality may be degrading, or seasonal effects.

---

Component 7: Synthesis and Recommendations

Always end with 2–3 concrete actions:

## What This Data Suggests

**Finding 1:** [The most important pattern]
→ Action: [Specific thing to change or do]

**Finding 2:** [Second pattern]
→ Action: [Specific action]

**Finding 3:** [Third pattern, if applicable]
→ Action: [Specific action]

## Recommended Diagnostic Updates

Based on [N] responses, consider updating:
- [ ] Replace [Q] — [X%] of respondents picked the same option (not discriminating)
- [ ] Adjust [band] threshold — most [high/medium] scorers are [behaving differently than expected]
- [ ] Add a [dimension] — several respondents mentioned [theme] in the notes that isn't captured

---

Analysis by Diagnostic Type

Lead Diagnostics → Pipeline Quality Analysis

Primary question: Are high-scoring leads converting better than low-scoring ones?

If yes → the diagnostic is working. Consider tightening the CTA for Medium to filter further. If no → either the questions aren't measuring fit, or the score threshold is wrong.

Secondary question: Which dimension best predicts conversion?

The dimension with the highest correlation to Closed Won is the one to weight more heavily in the next iteration.

Client Diagnostics → Retention Risk Analysis

Primary question: What do low-scoring clients have in common?

Look for patterns: industry, project type, engagement length, contract size, how they came in. The pattern reveals the clients to be more selective about accepting.

Secondary question: Are you running health checks early enough?

If churn is happening before the diagnostic reveals it, the cadence is too infrequent or the questions aren't detecting the right early signals.

Product Validation Diagnostics → Build/No-Build Signal

Primary question: Is the average score above 55?

• 60 across 30+ respondents → strong signal to build

• 45–60 → real problem, but not urgent enough — validate the urgency dimension specifically
• < 45 → problem exists but isn't actionable — pivot the audience or problem framing

Secondary question: Which dimension scores are lowest — and why?

If Willingness to Pay is low even when Pain Acuity is high → price sensitivity issue, not demand issue. If Current Solution Dissatisfaction is low → alternatives are working well enough. These are different problems with different solutions.

---

Save

Analysis report: data/1-Projets/diagnostics/[slug]/analysis-[date].md

Integration Points

• diagnostic-monitor-agent: runs this skill weekly for all active diagnostics
• monday-morning-agent: receives diagnostic summary in Monday briefing if configured
• sales-pipeline: cross-reference lead scores with pipeline outcomes
• business-health-advisor: client health diagnostic data feeds into overall health scan
• diagnostic-builder: analysis recommendations feed into next diagnostic iteration