We Searched ChatGPT for Realtors in Boston, Philadelphia, and San Diego. Here’s What We Found.

Greater Boston is one of the most competitive real estate markets in the country, and it draws a buyer profile that is almost perfectly matched to AI tool adoption. The biotech and pharmaceutical corridor along Route 128, the concentration of academic and research institutions, and the rapidly expanding tech sector have made Boston a primary relocation destination for professionals arriving from New York, Washington DC, and major Midwest metros. These are research-first buyers — people who have spent years navigating complex hiring processes and due diligence workflows. They approach buying a home the same way. Before they ever set foot in Cambridge, Brookline, or the South End, they have already done extensive online research. For a growing share of them, that research starts with ChatGPT.

When we ran the ChatGPT search for Boston realtors, we got back a strikingly thin result set. Two to three names appeared with any consistency — agents who had received meaningful coverage in The Boston Globe or who had appeared in regional business publications with high indexing authority. One name appeared primarily because of a feature in a national luxury real estate ranking that AI had treated as a credible citation. Everyone else — the specialists working Cambridge, the Brookline agents who have built their entire practice around the Newton-to-Cambridge corridor, the teams dominating Back Bay condo inventory — was absent.

Boston’s low inventory dynamic makes this especially consequential. Buyers in this market routinely move fast. A transplant professional relocating from DC for a Kendall Square biotech role typically doesn’t have the luxury of interviewing five agents and slowly narrowing down. They need to act. The buyers who use AI to identify an agent before they arrive — a cohort that is growing quickly in this market — are showing up with decisions already made. An agent who isn’t in ChatGPT’s answer doesn’t get considered. They don’t get a second chance when the buyer closes six weeks later.

For a full breakdown of which signals drive AI recommendations in the Greater Boston market, including what the visible agents have that everyone else doesn’t, see our Boston-specific analysis.

Philadelphia has become one of the most active transplant markets on the entire East Coast — and the primary driver is straightforward: buyers who can no longer afford Manhattan or Brooklyn are arriving in large numbers. The Fishtown-to-Chestnut Hill appreciation story over the past three years has given Philly a national profile it hadn’t previously had, and that profile is attracting buyers from across the Northeast who arrive having already done substantial research. Buyers moving from New York and New Jersey are, by and large, extremely AI-tool-native. They use ChatGPT the way an earlier generation used Google — as a first-pass research tool for any significant decision.

What we found when we searched ChatGPT for Philadelphia realtors was striking for a different reason than most markets. The national profile is actually reasonably well covered — national real estate publications have run pieces about Philadelphia’s affordability story, and a handful of names associated with that coverage showed up in results. But local expertise was almost entirely absent. The agents who actually know the difference between Fishtown and Manayunk, who understand the Chestnut Hill vs. Germantown tradeoffs that a New York buyer is navigating, who have closed dozens of transactions in South Philly neighborhoods that are still undergoing rapid appreciation — those agents were invisible. ChatGPT had a vague, national-publication understanding of Philadelphia real estate and almost no locally grounded expertise to offer.

This creates a specific problem for buyers. The transplant buyers arriving from New York don’t know Philly’s micro-markets. They’re asking AI precisely because they need guidance on which neighborhoods fit their criteria. The AI is giving them names associated with general Philadelphia real estate coverage, not neighborhood specialists. And the neighborhood specialists — the agents who could actually serve those buyers best — aren’t in the results at all.

For Philadelphia agents, this is the opportunity. The buyer pool is growing, those buyers are AI-native, and local expertise is nearly entirely absent from AI results. See the Philadelphia market overview for a detailed breakdown of what AI is and isn’t surfacing in Philly searches.

San Diego presents a buyer profile that is genuinely unique in the U.S. real estate landscape. The metro is a major military hub — Marine Corps Base Camp Pendleton, Naval Station San Diego, Marine Corps Air Station Miramar, and several other installations create a constant, high-volume flow of Permanent Change of Station (PCS) relocations. Military families moving on PCS orders have a very specific profile: they are operating on short timelines, they typically have no existing agent relationships in the destination market, and they research extensively because they have to. Among all buyer demographics, military families consistently rank among the most systematic online researchers. They approach a relocation like a logistics problem and they use every available tool — including AI — to gather information before they arrive.

When we searched ChatGPT for San Diego realtors, the results reflected the challenge of the metro’s geographic complexity. San Diego has a wide range of distinct submarkets — Carlsbad, Coronado, Pacific Beach, Chula Vista, La Jolla, Escondido — each with its own price profile, buyer pool, and inventory dynamics. The names that appeared in AI results were predominantly agents who had been covered in regional California real estate media or who had national luxury network affiliations. Agents who specialize in military relocation — a meaningful specialty in this market that involves VA loan expertise, proximity to base, and familiarity with PCS timelines — were essentially absent from results.

San Diego also draws substantial internal California migration. LA buyers escaping higher prices and density have been a consistent buyer cohort in the metro, and those buyers are, like military families, likely to research online before making contact. California buyers are among the most AI-tool-native in the country. The agents who serve this market — working Carlsbad for the north county LA migrants, or Chula Vista for military families near Coronado — largely didn’t show up in ChatGPT results regardless of their production volume or market expertise.

For San Diego agents, the military relocation angle is particularly worth noting. VA-loan-specialist agents and military relocation specialists have a concrete, specific expertise that AI should be able to surface — but almost none of them have built the indexed citation presence that would get them into results. See the San Diego market overview for detail on the AI visibility landscape in the metro.

Boston, Philadelphia, and San Diego are three different markets with different price points, different buyer profiles, and different competitive dynamics. But the agents who appeared in ChatGPT results across all three had built the same underlying structure. After running this experiment across more than a dozen metros, the pattern is consistent enough to describe with confidence.

• 1.Indexed third-party citations in credible publications — In Boston, that meant Boston Globe features and national real estate publication rankings. In Philadelphia, national publications covering the city’s affordability story. In San Diego, regional California real estate media and luxury network PR output. The agents who showed up had been mentioned by name in sources AI has indexed and trusts. Their production numbers, their Zillow ratings, their social media following — none of those signals are legible to the model. What is legible is a published citation in a third-party source with indexing authority.
• 2.Consistent, complete profiles on authoritative aggregators — Full bios on Zillow, Realtor.com, and Google Business that described their specialty, market coverage, and experience in specific, parseable language. AI builds confidence in a name by seeing it consistently described across multiple independent sources. The visible agents had complete, keyword-rich profiles that triangulated across platforms. The invisible agents — even those with strong production records — often had incomplete or generic profiles that gave AI nothing to work with.
• 3.Not: social media reach, beautiful websites, or Zillow reviews — We checked. Across all three markets, agents with large Instagram followings, professionally designed websites, and hundreds of five-star Zillow reviews were no more likely to appear in AI results than agents with minimal online presence. AI search is a different game. It rewards a specific kind of documentation — published, indexed, third-party mentions — that most agents haven’t focused on building because it never mattered before. It matters now.

In all three markets, the AI visibility landscape is still early. The agents who are showing up got there largely by accident — press coverage they didn’t strategize, rankings they didn’t optimize for. The field is open. If you’re in Boston, Philadelphia, San Diego — or anywhere else — and you want to know where you actually stand right now, run this 2-minute test to check whether ChatGPT can find you.