---
name: deal-qualification-scan
description: Score a deal from a call's Evidence Ledger against MEDDPICC (or SPICED) with the 0-2 rubric, a stakeholder grid, Green/Yellow/Red readiness, and the exact questions to ask next. Use when an AE or sales leader asks "is this deal real", "MEDDICC this call", "qualify this opportunity", "what are we missing on this deal", "who is the economic buyer", or "map the stakeholders".
---

# Deal Qualification Scan

For AEs and sales leaders. Answers one question: what do we actually know about this deal, and what do we need to find out next?

Rebuilt from the Momentum prompt library (MEDDICC Call Review, MEDDPICC Call Review, Framework Gap Analysis Coach, Multi-Threaded Deal Map), with evidence rules from the GTM AI OS.

## Inputs

- `ledger.md` from `transcript-denoise`. If it does not exist, run that first. Never score from a raw transcript.
- `rep-read.md` if present (rep confidence, doubts, stakeholders).
- `gtm-context.md` for personas (who should be the champion and economic buyer) and competitors.
- Optional: earlier ledgers for the same account in `calls/`. Read them in date order. Qualification builds across calls.
- Framework: MEDDPICC by default. Use SPICED if the user names it or `gtm-context.md` says so.

## Step 1: Score each element, 0 to 2

Score the **evidence on the deal**, using only ledger rows. Cite row numbers and timestamps.

| Element | 0 | 1 | 2 |
|---|---|---|---|
| Metrics | no metric discussed | a metric named, impact not quantified | metric named and the buyer quantified the impact |
| Economic Buyer | not identified | named, priorities or authority unclear | met on a call, authority confirmed, priorities stated in their own words |
| Decision Criteria | not uncovered | some criteria, no ranking | criteria stated with what matters most |
| Decision Process | not discussed | some steps, missing stakeholders or dates | steps, people, and dates mapped |
| Paper Process | not discussed | aware of legal, security, or procurement steps | steps, owners, and lead times stated |
| Identify Pain | no pain | pain stated, no impact or urgency | pain quoted, impact quantified, urgency stated |
| Champion | none | a likely champion, not tested, or the rep's notes question it ("champion?") | champion is selling internally (evidence: they booked the next meeting, shared internal info, or named the risk) |
| Competition | not discussed | named, no differentiation | alternatives named, including do-nothing, and why they would lose |

Total out of 16. Bands: 0 to 5 Poor, 6 to 9 Fair, 10 to 13 Good, 14 to 16 Strong.

SPICED alternative (0 to 2 each, total 12): Situation, Pain, Impact, Critical Event, Decision (process and criteria), Evaluation.

Scoring rules:
- PAIN scores 2 only with a quoted buyer statement tagged Verified or Supported.
- Inferred rows can support a 1, never a 2.
- An element the rep claims in the Rep Read but the ledger does not support scores on the ledger and gets a note: "rep believes X, transcript does not show it".

## Step 2: Stakeholder grid

One row per buyer-side person in the ledger or Rep Read:

| Name and title | Role (Economic Buyer, Technical Buyer, User Buyer, Champion, Influencer, Blocker, Unknown) | Influence (H/M/L) | Stance (Champion, Neutral, Skeptical, Blocker) | Priorities in their words | Concerns | Evidence rows |

Then list **missing stakeholders**: personas from `gtm-context.md` with no one in the grid, and any approver named but never met.

## Step 3: Readiness

| Status | Rule |
|---|---|
| Green | total 10 or more, Pain is 2, Economic Buyer is 2, a dated COMMITMENT exists |
| Yellow | total 6 to 9, or Green with one of those missing |
| Red | total 5 or less, or no Pain, or no next step |

Print the Rep Confidence (C1 to C5) beside the status. Map Green = C4 to C5, Yellow = C3, Red = C1 to C2. On any mismatch, say so in one line and list what the rep may know that the transcript does not show. The rep's read is not overwritten.

## Step 4: Next questions

For each element scored 0 or 1, write one question for the next call. Rules for the questions:
- Open-ended: start with How, What, Walk me through, or Tell me about.
- Built on something the buyer already said; quote it.
- Neutral, not leading.
- Quantify where possible.

Example: Pain scored 1 on "month-end takes forever". Next question: "You said month-end takes forever. How many days does it take today, and what does that delay cost the team?"

## Output: `deal-scan.md`

```markdown
# Deal Qualification: <account>, <meeting>, <date>
Framework: MEDDPICC | Total: <n>/16 (<band>) | Readiness: <G/Y/R> | Rep Confidence: C<n>

## Scorecard
| Element | Score | Evidence (quote, time) | Gap |

## Stakeholders
(grid, then missing stakeholders)

## Readiness
(status, rule met, rep vs evidence note)

## Next questions
| Element | Question | Built on |

## Top risk
One sentence. The single thing most likely to kill this deal, with its evidence.
```

## Rules

- Score the deal from buyer-side evidence. Do not grade the buyer.
- Every score cites ledger rows, or says "no evidence".
- No invented names, titles, or dates.
- Keep it to what a manager can read in 2 minutes before a deal review.