Farming List Prioritization: Seller Likelihood Scoring
A listing agent farms 500 homes in a Nashville neighborhood. Instead of mailing to all 500, they want AI to help build a scoring model that identifies the 50 most likely sellers. This is predictive analytics applied to prospecting — targeting resources where they'll have the highest return.
You are a real estate data strategist who builds seller likelihood scoring models for geographic farming.
Build a seller likelihood scoring model I can apply to my farm area of 500 homes.
Farm area: Sylvan Park, Nashville. Data available: property tax records (owner since date, assessed value), public records (mortgage origination date, divorce filings, probate), MLS history (last sale date, price), and demographic data (estimated age of homeowner).
Create a weighted scoring system with 0-100 scale. List each data point, the scoring criteria, and the weight. Explain how to identify the top 50 from 500.
You are a real estate data strategist who builds seller likelihood scoring models. I farm 500 homes in Sylvan Park, Nashville. I want to identify the 50 most likely sellers in the next 12 months. Data I can access: - Property tax records (owner since date, assessed value, homestead exemption status) - Public records (mortgage origination date, divorce filings, probate filings) - MLS history (last sale date, last sale price) - Demographic estimates (homeowner age range) - Property condition (from street-level observation) Build a weighted scoring model on a 0-100 scale. List each data point, scoring criteria, and weight. Show me how to identify the top 50.
SELLER LIKELIHOOD SCORING MODEL — Sylvan Park Farm DATA POINT 1: Ownership Duration (Weight: 25 points) - Owned 0-3 years: 5 points (unlikely to sell, negative equity risk) - Owned 3-5 years: 10 points (approaching average hold) - Owned 5-7 years: 20 points (at average Nashville tenure) - Owned 7-10 years: 25 points (highest seller probability) - Owned 10+ years: 15 points (stable, but equity-rich and potential downsizers) DATA POINT 2: Life Event Indicators (Weight: 25 points) - Divorce filing in last 12 months: 25 points - Probate filing in last 12 months: 25 points - Homestead exemption removed: 20 points (may indicate move) - Pre-foreclosure notice: 25 points - None detected: 0 points DATA POINT 3: Equity Position (Weight: 20 points) Calculate: Current estimated value minus last purchase price - Equity gain > 40%: 20 points (strong incentive, can afford to move) - Equity gain 20-40%: 15 points - Equity gain 10-20%: 10 points - Equity gain < 10%: 5 points (limited motivation) DATA POINT 4: Mortgage Age (Weight: 15 points) - Original mortgage, 5-8 years old: 15 points (past the settling-in phase) - Refinanced in last 2 years at low rate: 5 points (rate-locked, less likely to move) - Mortgage 10+ years: 10 points (significant paydown, flexible) - Cash purchase / no mortgage: 8 points (no rate lock-in, but also no urgency) DATA POINT 5: Property Condition Signals (Weight: 15 points) - Visible deferred maintenance (peeling paint, overgrown yard): 15 points - For-rent sign or property manager contact: 15 points (investor may be ready to exit) - Recent exterior improvements: 5 points (investing in the property = staying) - No observable signals: 8 points APPLICATION: Score each of the 500 properties. Sort descending. Your top 50 (scores 55+) get personalized outreach: handwritten notes, door knocks, and targeted market updates. The next 100 (scores 35-54) get your standard mailer. The remaining 350 get digital-only touches. EXPECTED OUTCOME: In a typical 500-home farm, you should identify 40-60 properties scoring above 55. Of those, historically 8-12% will list within 12 months. That's 4-7 listings from targeted prospecting versus the 1-2 you'd get from blanket marketing.
Why It Works
The AI built a practical scoring model using publicly available data. The weighted system prioritizes life events and ownership duration — the two strongest predictors of seller motivation. This is predictive analytics without software: a spreadsheet, public records, and the right framework.