home / skills / gtmagents / gtm-agents / deal-quality-model

This skill helps prioritize deal hygiene and win likelihood by scoring signals, enabling targeted inspections and alerts across CRM and dashboards.

npx playbooks add skill gtmagents/gtm-agents --skill deal-quality-model

Review the files below or copy the command above to add this skill to your agents.

Files (1)
SKILL.md
1.2 KB
---
name: deal-quality-model
description: Scoring system for opportunity hygiene, win likelihood, and inspection
  prioritization.
---

# Deal Quality Model Skill

## When to Use
- Prioritizing inspection focus during forecast and pipeline reviews.
- Flagging opportunities with incomplete data or low win signals.
- Creating alerts when critical hygiene steps (MEDDIC, next steps) are missing.

## Framework
1. **Input Signals** – discovery notes, exec sponsor, mutual plan, activity recency, budget proof.
2. **Weighting** – assign weights per signal (required, strong, nice-to-have) per segment.
3. **Score Bands** – define thresholds for healthy, watch, and critical opportunities.
4. **Automation Hooks** – integrate with CRM validation rules, Slack alerts, and dashboards.
5. **Governance** – review scoring monthly, capture overrides, and document rationale.

## Templates
- Signal scoring matrix with thresholds + automation IDs.
- Inspection checklist for reps/managers.
- Alert message template for Slack/email.

## Tips
- Align scoring with enablement frameworks (MEDDIC, SPICED) to drive adoption.
- Expose scores in forecasts so execs can challenge risky commits.
- Pair with `guardrail-scorecard` or pipeline commands to operationalize.

---

Overview

This skill scores opportunity hygiene, estimates win likelihood, and prioritizes inspection work across the pipeline. It produces a composite deal-quality rating driven by configurable signals and segment-specific weights. The output helps sales, RevOps, and leadership focus reviews and automation on risky or high-potential deals.

How this skill works

The model ingests structured and unstructured signals such as discovery notes, executive sponsor presence, mutual plans, activity recency, and budget evidence. Each signal is assigned a weight (required, strong, nice-to-have) per segment and combined into a numeric score. Scores map to bands (healthy, watch, critical) and trigger automation hooks like CRM validation, Slack alerts, and dashboard badges. Governance controls capture monthly reviews, documented overrides, and score rationale.

When to use it

  • Prioritizing inspection focus during forecast and pipeline reviews
  • Flagging opportunities with incomplete data or low win indicators
  • Creating alerts when key hygiene steps (MEDDIC/SPICED elements, next steps) are missing
  • Routing deals for manager review or gating commit inclusion in forecasts
  • Driving enablement by exposing common hygiene gaps to reps and coaches

Best practices

  • Align scoring signals with your enablement framework (MEDDIC, SPICED) to improve adoption
  • Calibrate weights by segment and review performance monthly to prevent model drift
  • Expose scores in forecasts and dashboards so executives can challenge risky commits
  • Document all manual overrides and the rationale to maintain trust and auditability
  • Automate lightweight CRM validation rules first, then expand to alerts and remediations

Example use cases

  • Auto-flag mid-market deals missing an exec sponsor and route to AE for action
  • Downgrade commit status for deals in the critical band ahead of executive forecast call
  • Trigger a Slack alert when discovery notes lack budget or timeline evidence
  • Run a weekly hygiene report for managers showing top watch/critical deals to inspect
  • Integrate with pipeline guardrails to prevent low-quality deals from counting toward targets

FAQ

Can weights be customized by segment?

Yes. Assign required/strong/nice-to-have weights per segment and tune them based on historical win rates.

How do automation hooks work?

Score band transitions can call CRM validation rules, send Slack/email alerts, or update dashboard fields through configured automation IDs.