Real Estate · Multifamily · Off-Market Sourcing
Off-market multifamily deal sourcing, on autopilot.
How a Bay Area multifamily investor turned fragmented county records into a ranked, contactable deal pipeline that refreshes overnight.
Nine Bay Area counties scanned every night into one scored, contactable deal list.
The Challenge
The investor competes for multifamily assets that almost never reach a listing. The signals that precede an off-market sale sit buried across county recorder sites, each with its own format, and surfacing them meant analysts copying records into spreadsheets by hand. By the time a property was researched, scored, and matched to an owner, the window to reach out had usually closed. Coverage was shallow, the data went stale within days, and nobody could say which leads were actually worth a call.
Pain Points:
- Off-market signals scattered across nine county recorder sites, each in a different format.
- Manual record entry that went stale within days of being collected.
- No consistent way to score or rank which properties deserved outreach first.
- Owner and comp lookups done one property at a time, so the best deals were reached late or not at all.
Our Approach
We built an off-market sourcing engine that ingests county records, scores them, and hands the investor a ranked call list every morning:
County-Record Scrapers
Automated ingestion across Bay Area recorder sources, with per-source toggles so coverage can expand county by county. Three counties were live at handover, with the architecture built to scale to nine.
Distress & Opportunity Signals
Raw records are normalized and de-duplicated, then scanned for the ownership and financial signals that tend to precede an off-market sale.
Scoring & Ranking Engine
Every property is scored and ranked so the highest-intent opportunities surface first, instead of analysts reading through undifferentiated lists.
Comps Engine & Geocoding
Each lead is enriched with comparable sales and precise geocoding, so a ranked, contactable list is ready the moment the investor opens the dashboard.
The System in Action
Stack: Railway (scrapers) + Vercel + Supabase (Postgres) + geocoding & comps services
Results
County Coverage
manual, ad hoc
9 counties scanned nightly
Lead Pipeline
scattered spreadsheets
1,800+ ranked leads
Pipeline Freshness
stale within days
overnight refresh
Deal Readiness
manual comp + owner lookups
contactable on open