# SalesCoach Meeting Analysis Rubric

## Purpose

Central rubric for analyzing uploaded meeting transcripts. It combines objective meeting intelligence with sales coaching frameworks documented in `docs/skills/`.

## Output principles

1. Do not invent facts.
2. Separate fact, inference and recommendation.
3. Cite short transcript evidence where possible.
4. Be constructive, specific and commercially useful.
5. Prefer next-action coaching over generic criticism.

## Section 1 — Metadata extraction

Extract:

- Meeting title/topic.
- Date/time, if present.
- Duration, if present or inferable from timestamps.
- Participants and likely roles.
- Companies mentioned.
- Speaker turns.
- Talk-time or word-share distribution.
- Language.
- Meeting type: discovery, demo, negotiation, follow-up, onboarding, support, internal, other.
- Confidence level for each field.

## Section 2 — Conversation intelligence

Extract executive summary, timeline/key moments, customer pains, current state, desired future state, business impact, objections, buying signals, competitors/status quo, decision process clues, next steps/owners and open questions.

## Section 3 — Sales coaching score

Overall score: 0–100.

Suggested weighting:

- Discovery depth: 20
- Buyer understanding/tactical empathy: 15
- Business impact clarity: 15
- Conversation control/process: 10
- Objection handling/negotiation: 10
- Value articulation/reframe: 10
- Next-step quality: 10
- Ethical persuasion/framing: 10

## Section 4 — Framework mapping

Apply only relevant frameworks:

- **SPIN Selling** — classify seller questions and discovery progression.
- **Gap Selling** — map current state, future state, gap, root cause and impact.
- **Never Split the Difference** — evaluate tactical empathy, labels, mirrors, calibrated questions and real commitment.
- **Challenger Sale** — evaluate insight, tailoring, constructive tension and process control.
- **MEDDICC** — qualify enterprise deal evidence and forecast risk.
- **Pre-Suasion / Influence** — evaluate opening frame, salience, proof, authority and ethical urgency.
- **Mom Test** — detect hypotheticals, compliments and false validation.
- **Sandler** — upfront contract, pain/budget/decision qualification and disqualification discipline.

## Section 5 — Red flags

Flag if present:

- Seller talk time above 65% in discovery.
- Few or no buyer questions answered deeply.
- Pitch before problem/impact discovery.
- No quantified pain.
- No clear next step with owner/date.
- Vague enthusiasm but no commitment.
- Decision-maker absent and no path to them.
- Objection dismissed quickly.
- No discussion of status quo or competition.

## Section 6 — Coaching output

Always include:

1. What went well.
2. What was missed.
3. Top 3 coaching priorities.
4. Better questions the seller could have asked.
5. Recommended follow-up email bullets.
6. Next meeting plan.
7. Deal risk assessment.

## Conceptual JSON schema

```json
{
  "metadata": {},
  "participants": [],
  "talk_time": [],
  "conversation_intelligence": {},
  "sales_analysis": {
    "overall_score": 0,
    "subscores": {},
    "framework_findings": [],
    "strengths": [],
    "improvements": [],
    "recommended_questions": [],
    "follow_up_recommendations": []
  }
}
```