Engagement Scoring Model
3D Engagement Scores (Interest × Intent × ICP Fit)
3D Engagement Scores modernize lead/contact scoring by combining Interest, Intent, and ICP Fit into a single prioritization framework—built on observed signals, explainable drivers, and operational refresh cycles.
!The Problem
Legacy scoring approaches can miss signal interactions, lack transparency, and aren’t optimized for modern multi-signal GTM data—especially during platform migrations.
What It Does
Calculates three scores to answer three different questions:
- 1Interest score (short window):“Are they engaged?” (modeled over ~90 days)
- 2Intent score (longer window):“Are they in-market / ready to act?” (modeled over ~270 days)
- 3ICP / ideal-fit score:“Are they a strong fit?” (using firmographics + historical wins/opportunities logic)
Then combines them into a final confidence score (e.g., 40% interest, 40% intent, 20% ICP).
Key Capabilities
Multi-Source Learning
Learns from multiple GTM data tables (contacts, opportunities, leads, responses)
Gradient Boosting
Uses gradient boosting approaches (LightGBM) to capture feature interactions
Explainability
Provides explainability via SHAP-style feature contribution explanations
Consumable Output
Produces a consumable output file (CSV) and a lightweight viewer app
Outputs
- Contact-level interest/intent/fit scores + final confidence score
- Explainability artifacts for BDR/Sales enablement
- Repeatable pipeline that can refresh frequently (e.g., daily)
Primary Users
- SDR/BDR teams
- Marketing Ops
- RevOps
- Campaign owners
Success Metrics
- Higher meeting conversion rate from prioritized outreach
- Better routing and follow-up SLAs
- Reduced “false positives” vs legacy scoring
- Increased transparency and trust in scoring