playbooks

home / skills / gtmagents / gtm-agents / signal-scoring

signal-scoring skill

/plugins/intent-signal-orchestration/skills/signal-scoring

This skill helps design and audit multi-source signal scoring models with decay, weighting, and governance for RevOps.

npx playbooks add skill gtmagents/gtm-agents --skill signal-scoring

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

Files (1)
SKILL.md
1.3 KB
---
name: signal-scoring
description: Use to design composite intent scoring models with decay, weighting,
  and governance.
---

# Signal Scoring Blueprint Skill

## When to Use
- Designing or revising multi-source signal scoring models.
- Auditing why certain accounts jump tiers or stay suppressed.
- Communicating scoring logic to RevOps, sales, and compliance stakeholders.

## Framework
1. **Source Inventory** – list each signal type, freshness cadence, coverage %, and reliability.
2. **Weighting Model** – assign base weights, adjust for persona relevance, and define topic multipliers.
3. **Decay Curves** – set time-based decay (hours/days/weeks) per signal category with thresholds for expiration.
4. **Thresholds & Tiers** – document what scores map to activation tiers, nurture status, or executive escalation.
5. **Governance** – outline ownership, review cadence, and change management triggers.

## Templates
- Scoring worksheet (source, weight, decay, threshold, notes).
- Tier definition matrix with recommended plays.
- Change log template capturing rationale, data evidence, and approvals.

## Tips
- Blend at least two independent signals before green-lighting high-effort plays.
- Version control your scoring logic and share diffs in RevOps channels.
- Pair with `suppression-logic` to avoid conflicting triggers.

---

Overview

This skill helps design and maintain composite intent scoring models that combine multiple behavioral and firmographic signals with decay, weighting, and governance. It provides a structured blueprint, templates, and practical rules to turn raw signals into actionable tiers and plays. Use it to make scoring transparent, defensible, and repeatable across RevOps, sales, and CS teams.

How this skill works

The skill guides you through cataloging signal sources, assigning base weights and persona/topic multipliers, and defining time-based decay curves per signal category. It includes threshold-to-tier mappings, governance processes for change control, and templates for worksheets, tier matrices, and change logs. The result is a version-controlled, auditable scoring model that supports activation, suppression, and escalation decisions.

When to use it

  • Designing or revising a multi-source intent scoring model
  • Auditing why accounts moved tiers or remained suppressed
  • Aligning RevOps, sales, and compliance on activation logic
  • Defining decay and expiration rules for time-sensitive signals
  • Preparing playbooks tied to defined score thresholds

Best practices

  • Inventory every signal: capture freshness cadence, coverage, and reliability
  • Require at least two independent signals before authorizing high-effort plays
  • Apply time-based decay per signal category and set clear expiration thresholds
  • Version control scoring logic and publish diffs to RevOps channels
  • Document ownership, review cadence, and change-management triggers

Example use cases

  • Create a scoring worksheet to combine web, email engagement, and intent data with decay rules
  • Audit a sudden account tier change and trace which signals and weights caused it
  • Define tier thresholds and map each tier to specific sales or marketing plays
  • Implement a governance process with a change log capturing rationale and approvals
  • Pair scoring with suppression logic to avoid duplicate or conflicting triggers

FAQ

How many signals should I require before triggering a high-effort sales play?

Require at least two independent signal types to reduce false positives; prefer combinations across behavioral and firmographic sources.

How do I handle stale signals?

Apply decay curves that reflect signal lifetime (hours/days/weeks) and set explicit expiration thresholds so stale signals stop contributing to scores.