We Asked AI to Find Us a Realtor in Miami. It Couldn’t Decide If We Wanted a Person or a Company.

Buying a home in Miami is one of those decisions where the person helping matters as much as the property itself. A good agent knows which buildings have assessment issues nobody talks about, which neighborhoods are about to get loud from construction, and which sellers are actually motivated. A bad one sends Zillow links.

So when someone asks ChatGPT “best realtor in Miami,” the answer should probably be a person. A specific agent with a track record and a phone number.

That’s what Gemini thinks too. Its top recommendations were individual agents. David Siddons. Dora Puig. Dina Goldentayer. Real people with real names.

ChatGPT took a completely different approach. It recommended Compass. LUXE Properties. Douglas Elliman. Companies, not people. Brokerages with thousands of agents, as if saying “go to this building and talk to whoever’s at the front desk.”

Same question. Two fundamentally different interpretations of what a “realtor recommendation” even means.

How AI Recommends 85 Different Realtors for the Same City

Across all four platforms, 85 different realtors and brokerages were named. Gemini alone produced 45, most of them individual agents. ChatGPT listed 24, leaning toward brand names. Perplexity and Grok landed somewhere in between.

Compass came out on top in the combined ranking, but in a strange way. It was never the top recommendation on any single platform. It just showed up often enough across ChatGPT, Perplexity, and Grok to accumulate the highest combined score. Gemini never mentioned it.

David Siddons was Gemini’s clear favorite, appearing in every single response there. On ChatGPT, Perplexity, and Grok, he didn’t come up once.

Douglas Elliman was Grok’s top pick. The other platforms barely mentioned it.

The top-ranked agent and the top-ranked brokerage occupied completely different platforms. A buyer using Gemini would get a person to call. A buyer using ChatGPT would get a brand to Google.

Why ChatGPT Recommends Brokerages and Gemini Recommends Agents

Real estate has a structure that AI clearly hasn’t figured out. An agent works under a brokerage. Dina Goldentayer might work at Douglas Elliman. The Nancy Batchelor Team operates under Compass. But when AI recommends “Compass,” it’s not recommending Nancy Batchelor or any specific agent. It’s recommending a company with thousands of agents across the country.

That’s like recommending “Marriott” when someone asks for a good hotel in Miami Beach. Technically not wrong, but not really helpful either.

ChatGPT leaned into the brokerage side. LUXE Properties, eXp Realty, The Corcoran Group, Coldwell Banker. Big names, little personal context. A home buyer following ChatGPT’s advice would end up on a corporate website trying to figure out which of the fifty agents listed there is actually good.

Gemini went the other direction and recommended specific people. Dora Puig appeared in 75% of Gemini responses with high rankings. Liz Hogan, Stefania Cambarau, Brett Harris, David Vazquez. All individuals, all with personal brands.

Neither approach is wrong. But they’re solving completely different problems, and a buyer asking the question probably expected one or the other, not a random coin flip between the two.

AI Lists the Same Realtor Under Different Names

The name fragmentation was everywhere.

“Nancy Batchelor Team” appeared on ChatGPT. So did “Nancy Batchelor Team (Compass).” Two entries for the same team, one with the brokerage name attached and one without.

“Sep Niakan” appeared as “Sep Niakan” on one platform and “Sep Niakan (Blackbook Properties)” on another. Same person.

“The Opes Group” and “The Opes Group (Compass)” split into two entries.

“ONEPATH | eXp Realty” and “OnePath Realty” were listed separately.

“Douglas Elliman” and “Douglas Elliman, Miami Beach” counted as two different businesses.

“Riley Smith Group” and “The Riley Smith Group” were treated as separate entities because one had “The” in front.

In a business where personal reputation is everything, having that reputation scattered across multiple AI entries instead of concentrated under one name is a quiet but real problem.

Which AI Platform Recommends the Most Miami Realtors

One detail stood out. Gemini recommended nearly three times as many agents as any other platform. Its responses read like someone who actually follows Miami real estate, dropping names of specific agents with their specialties and neighborhoods.

Perplexity and Grok each recommended 16. Their answers felt more like someone reading a “top 10” listicle, sticking to bigger names and established brokerages.

ChatGPT fell in the middle with 24, mixing brokerages and agents without much apparent logic.

For an individual agent, this matters. Getting recommended by Gemini might be easier than getting recommended by ChatGPT, simply because Gemini casts a wider net and seems more willing to surface individual realtors rather than defaulting to brand names.

What a Missed Recommendation Costs a Realtor

A single real estate transaction in Miami generates a commission of $15,000 to $50,000 or more. A luxury property can mean a six-figure commission from one deal. And unlike most service businesses, a realtor often locks in the client through that first moment of contact. Once a buyer signs with an agent, that relationship typically lasts through the entire transaction.

Most agents spend heavily on Zillow leads, Realtor.com profiles, Instagram ads, open house marketing, and networking events. The cost of acquiring a single client can run into thousands of dollars.

Meanwhile, more and more buyers are starting their search by asking ChatGPT or Gemini something casual. “Good realtor in Miami for first-time buyers.” “Luxury real estate agent Brickell.” These aren’t formal leads. They’re the modern equivalent of asking a friend for a name. And the agent whose name comes back gets a warm introduction that no ad spend can replicate.

How to Check If ChatGPT Recommends a Realtor

The five-minute check is the starting point. Type the kind of question a buyer would ask into ChatGPT, then Gemini, and see what comes back. Pay attention to whether the response names the agent personally or just the brokerage.

Reachd.ai tracks how any business or individual appears across ChatGPT, Gemini, Perplexity, and other AI platforms. Which competitors show up, how often, and what’s behind the difference. For a business where one recommendation can be worth a five-figure commission, knowing the answer is the kind of thing that pays for itself on the first deal.

Because right now, someone in Miami is asking their phone to help them find an agent. And whether they get a person or a corporate homepage depends on which app they asked.