Lead Scoring Model

What It Is

A model that scores each lead on two independent axes:
- Fit (demographic/firmographic): static match to the ideal-customer-profile — size, revenue, industry, title, geography, tech stack. Do they match who we sell to?
- Engagement (behavioral): activity intensity, with high-intent actions weighted heavily (one demo request ≫ a dozen email opens). Are they actively buying?

The quadrant is the insight: high-fit + high-engagement is the only cell worth an immediate sales touch; high-engagement/low-fit wastes reps; high-fit/low-engagement needs nurture.

Model Types

Demographic (fit), behavioral (engagement), predictive (ML learns which attributes correlate with closed-won from CRM history), and hybrid — most teams land on hybrid, balancing automation with human judgment.

How It Applies to Marketing Factory

The scoring model is the qualification layer that attacks the steepest funnel leak (MQL→SQL, per benchmarks-as-priors). It consumes event-tracking-schema and identity data, feeds lead-routing-sla, and complements the product-qualified-lead signal for PLG motions. The score is a prioritization tool, not truth — recalibrate thresholds from real conversion data. Computing and maintaining the model is fully agent-ownable.

Referenced from: lead-scoring-and-routing