Lead Scoring skill

Lead Scoring is an agent skill for AI coding assistants (Claude Code, OpenClaw, Cursor, Codex). Design and implement lead scoring models. Qualify leads based on behavior, demographics, and engagement. Install with: npx skills-ws install lead-scoring.

conversionv1.0.0Updated
copied ✓
openclawclaude-codecursorcodex
0 installsVirusTotal: cleanSource code

Lead Scoring

Scoring Model Design

Two-Axis Model

Score leads on two independent axes:

  1. Fit Score (0-100): How well they match your ICP (demographics)
  2. Engagement Score (0-100): How actively they interact (behavior)

Combine: Total Score = (Fit × 0.4) + (Engagement × 0.6)

Fit Score (Demographics)

SignalPointsExample
Company size matches ICP+2050-500 employees
Industry match+15SaaS, fintech
Job title/seniority+20VP+, Director, C-level
Budget range confirmed+15>$50K ARR potential
Geography match+10Target market
Tech stack match+10Uses compatible tools
Revenue range match+10$5M-$50M ARR

Engagement Score (Behavior)

SignalPointsDecay
Pricing page visit+20-5/week
Demo request+30None
Free trial signup+25-5/week inactive
Case study download+10-3/week
Blog post read+2-1/week
Email open+1-1/week
Email click+5-2/week
Webinar attended+15-3/week
Multiple sessions (3+)+10-2/week
Returned after 30d absence+15-5/week

Score Decay

Apply weekly decay to prevent stale high scores. A lead who visited pricing 3 months ago isn't hot anymore.

Thresholds

ScoreClassificationAction
0-30Cold leadNurture sequence
31-50Warm leadTargeted content
51-70MQLMarketing-qualified, alert SDR
71-85SQLSales-qualified, direct outreach
86-100HotImmediate sales attention

Qualification Frameworks

Details: references/scoring-models.md

References

  • references/scoring-models.md — BANT, CHAMP, MEDDIC frameworks with implementation guides
  • references/signal-weights.md — Calibrating signal weights with historical data