A storm swath containing 13,000 properties doesn't mean 13,000 equal opportunities. A 10-year-old $500,000 owner-occupied home that received 1.75-inch hail is a fundamentally different lead than a 2-year-old $180,000 rental property that received 1.0-inch hail. AI lead scoring assigns a 0 to 100 priority score to each property based on the factors that actually predict conversion — enabling the outreach system to contact the highest-probability leads first and allocate crew capacity efficiently.
The Four Scoring Variables
Hail size at the specific address (from HailTrace property-level data): 1.75 inches or larger scores highest — this is the threshold where asphalt shingle damage is virtually certain and insurance claims are reliably approved. Roof age (calculated from ATTOM year-built data): roofs older than 15 years on asphalt shingles are near end-of-life and more susceptible to damage; older roofs score higher. Property value (from ATTOM estimated value): higher-value properties generate larger insurance claims — a $500,000 home generates a larger roofing job than a $200,000 home. Owner-occupied status (ATTOM boolean): owner-occupied properties score significantly higher than absentee-owner properties because the occupant has a direct personal interest in filing a claim and getting repairs done.
How the AI Scoring Prompt Works
An OpenAI API call processes each property with a prompt structured as: "Score this property from 0 to 100 for storm damage lead quality. Return only the integer score. Data: Hail size at address: 1.75 inches. Roof age: 23 years. Property value: $487,000. Owner occupied: yes. Roof material: asphalt shingle." The model returns a consistent numerical score that routes into the pipeline's priority queue. The full scoring model and lead card display are demonstrated at omnionlinestrategies.com/storm-lead-ai-machine.