Bobby Kerr in Inman told real-estate agents how answer engines pick who to name. The short version is blunt.
“AI assistants build shortlists from public data — mostly your Google presence.”
— Bobby Kerr, Inman
Kerr’s list of what decides the shortlist is three objects: the Google Business Profile, the website, and third-party mentions and reviews. He also quotes the Whitespark line that, in AI search, “mentions are the new link.” Keep that sentence inside Inman. This desk is not reprinting Whitespark’s tables.
Recommendation is still retrieval
A complete Maps profile matters. Reviews matter. A mention on a local news site matters. None of those facts prove who controls the website, which licence sits on the van, or whether the phone on the profile still rings the same firm.
That is the hole agents feel when ChatGPT names “a top realtor near you” and the person who shows up is a different brokerage with a cleaner Google page. The model did not verify the business. It retrieved the easiest public text.
What to add after you fix Maps
Do the work Kerr describes. Fix the primary category. Write the service areas in plain words. Ask for specific reviews. Then publish a third object the model can fetch when those sources disagree: a dated public file with the legal name, the site you own, and the phone.
Put the seal on that site so a person and a crawler land on the same record. Do not buy a promised number-one slot in ChatGPT. Nobody can honestly sell that.
Read the source: How To Get Recommended By AI As A Real Estate Agent (Bobby Kerr, Inman, Aug. 31, 2026).