New Homesnap platform's algorithm creates "Likelihood to List" scores

Homesnap's algorithm sift through big data to assign a "Likelihood to List" score on properties before they show up on the market.

The platform has a new service that crunches data to show which homes in a neighborhood are the most likely to be listed for sale in the next 12 months.

Homesnap uses an algorithm that crunches millions of records, including MLS data, to come up with a “Likelihood to List” score. The service, available only to licensed agents, shows a “heat map” of neighborhoods color-coded to zero in on the homes most likely to go on the market.

Lou Mintzer, Homesnap’s Chief Product Officer, said:

“Instead of sending a postcard to 5,000 homes, real estate agents can just focus in on the people the algorithm has determined are the most likely to list their homes."

For example, a home withdrawn from the market before it sells is a key indicator that it’s likely to list again, Mintzer said.

“The home might have been priced too high, maybe languished on the market 4 or 6 months, and now the MLS status is `withdrawn,’” Mintzer said. “An unsuccessful listing is a signal the home might be listed again.”

Another indicator is the age of the owners. Someone who is 85 years old and living in a home alone might be ready to think about selling, he said.

It’s a far cry from a decade or more ago when agents had to adopt more scattershot approaches to finding clients, like an ad on the side of a bus.

Read more here