Technology-based brokerage Redfin is offering a new online service that suggests listings based on how buyers use the company’s website.
The twist, says Redfin CEO Glenn Kelman, is that Redfin agents approve or reject homes on the list based on the agent’s personal knowledge of what properties their clients are most likely to be interested in.
By taking into account clients’ entire online profile, and combining it with what Redfin agents learn about buyers while taking them through homes, the “Redfin Matchmaker” tool is able to make “uncanny recommendations about listings” clients haven’t yet asked to see, Kelman said in a statement.
Unlike the suite of tools Redfin makes available for consumers to discover properties themselves — including a Home Value Tool, School Search and “Tour Insights” — Redfin agents use the Matchmaker tool on behalf of their clients.
The tool uses machine learning algorithms to analyze homes viewed by redfin.com users to make recommendations that are reviewed by agents.
In cases where Redfin clients are employing search criteria that are too narrow, the tool can dig up properties that they might not otherwise have considered. Buyers looking for homes in the Georgetown neighborhood of Washington, D.C., might be pointed to similar homes in Cleveland Park, for example.