Identifying Listings Before They Hit the Market: An 11x Advantage in LA
How Swarmalytics helped agents target homes 11 times more likely to list — with implications for national real estate teams
In early 2025, a senior investment consultant at one of the world’s largest real estate franchise brands approached Swarmalytics with a critical question: could AI be used to predict which homes are most likely to list before they hit the market?
They wanted to test our model’s accuracy in one of the most competitive housing markets in the U.S.—Los Angeles. The goal was to prioritize outreach and marketing efforts toward the right homeowners at the right time so that agents could improve listing conversion, prioritize outreach to the right potential sellers, and predict more potential revenue opportunities.
The Work
To answer the question, we designed a focused pilot in three high-volume zip codes in LA: 90016, 90043, and 90047. Here’s how it worked:
We applied Swarmalytics’ national AI model, which is built from 36 localized submodels trained on over 4,000 variables per property
These variables include everything from historical sales trends to neighborhood dynamics
We flagged all homes in the area scoring 900 or above on our internal “Sellability Index,” our threshold for high-likelihood listings
In total, 247 homes were flagged as likely to list within 90 days.
The Outcome
Three months later, 23 of those 247 homes were listed (a listing rate of 9.3%). To compare, only 295 of the 35,135 other homes in the same area were listed in that same time period, a baseline of 0.8%. Meaning that the homes identified by Swarmalytics were over 11 times more likely to list than the market average.
We also looked at the dollar impact:
Median list price of flagged homes: $799,900
Total value of listed properties from our list: $18.4M
That means our model surfaced $18+ million in potential inventory that agents and investors could have reached first
The Opportunity
Swarmalytics enables national real estate organizations to predict seller intent with precision. Rather than broad marketing across entire zip codes, teams can focus their outreach on properties with the highest likelihood of converting to listings. Our AI model makes this possible by identifying seller signals early, enabling more precise targeting, more productive outbound activity, and revealing increased revenue opportunities.