$ man scoring

Scoring

Assigning a numeric value to a company or contact based on how well they match your criteria. Simple version: 0-10 scale. Advanced version: 200-point composite with weighted dimensions.

Binary yes/no qualification is too rigid. Scoring adds nuance. A company might be a 7 — not perfect, but close. Maybe they're missing one signal. Maybe they're in a gray-zone industry. Scoring lets you tier leads by strength instead of just pass/fail. It also lets you route by priority — 9-10 gets the premium sequence, 6-7 gets standard. and when you move from a simple 0-10 scale to a multi-dimensional composite score, you can weight what matters most per partner. persona fit might matter more than revenue size. vertical match might matter more than tech stack. the weights tell the model what to prioritize.

I use two scoring models depending on partner complexity. model 1 — simple 0-10: the qualification prompt checks firmographics against ICP and outputs a score, confidence level, and reasoning. 8-10 = qualified, 6-7 = needs research, <6 = skip. this works for most campaigns. model 2 — 200-point composite: I built this for a partner with complex qualification needs. 100 points for Fit Score (Persona Match 30pts, Revenue Band 25pts, Vertical Match 25pts, Tech Stack 20pts) and 100 points for Engagement Score (Email Opens 5pts, Link Clicks 15pts, Website Visit 10pts, Demo Page Visit 25pts, Reply 25pts, Meeting Booked 20pts). the composite score lets you rank leads by both fit AND engagement — a perfect-fit company with zero engagement scores differently than a good-fit company that's been clicking every email. the 200-point model feeds into Clay routing logic: 150+ = hot lead (route to AE immediately), 100-149 = warm (high-priority sequence), 50-99 = nurture, <50 = skip. scoring models should match the complexity of the sales motion. don't build a 200-point model for a 2-email campaign.