How AI Search Chooses Which Realtors to Recommend: ChatGPT, Gemini, Google AI Mode & AI Overviews

When a homebuyer asks ChatGPT, Gemini or Google:

“Who is the best Realtor near me?”

the answer is not simply a copy of Google's traditional search rankings.

AI systems can search across real estate portals, Realtor websites, Google Business Profiles, directories, reviews and other third-party sources before producing a much shorter list of agents.

And different AI platforms often recommend completely different Realtors.

A 2026 Local Falcon study ran 37,500 Realtor-related AI searches across the 100 largest U.S. cities using ChatGPT, Gemini, Google AI Mode, Google AI Overviews and Grok.

Among the established agents and teams benchmarked in the study, 91.5% were never cited by any of the AI platforms tested. Even more importantly, 82.6% of the Realtors who appeared were named by only one platform and no others.

That means there is no single universal “AI ranking.”

A Realtor can be highly visible in ChatGPT and almost invisible in Google AI Mode—or the opposite.

So how do these systems decide who to recommend?

The answer appears to depend on a combination of:

• what the buyer asks,
• where the buyer is located,
• what information exists about the Realtor online,
• which sources each AI platform relies on,
• how clearly the Realtor's expertise is documented,
• and how each platform retrieves and synthesizes information.

Understanding those differences is becoming increasingly important for real estate professionals.

The Short Answer: How Does AI Choose Realtors to Recommend?

AI search systems appear to build Realtor recommendations by retrieving information from multiple online sources and using that information to determine which agents are relevant to the user's question.

For a query such as:

“Who is a good Realtor for buying a condo in Richmond?”

the system may need to understand several things:

• Which agents work in Richmond?
• Which agents have evidence of condo expertise?
• Which agents have relevant reviews or third-party mentions?
• Which sources support those claims?
• Which agents appear consistently across trusted websites?
• Which agents are geographically relevant to the user?

The AI then reduces those candidates into an answer.

This is very different from simply sorting Realtors by sales volume.

Local Falcon compared AI recommendations with RealTrends' production rankings and found that only 21.2% of the highest-volume individual agents in the markets studied were cited by any AI platform.

In other words:

Being a top-producing Realtor does not automatically make you visible to AI.

AI can only work with information it can retrieve and interpret.

1. ChatGPT: Heavy Reliance on Real Estate Portals and Directories

ChatGPT can use web search to retrieve current information and provide answers with links to supporting sources.

OpenAI explains that ChatGPT search combines conversational answers with current information from the web and can use third-party search providers and content partners to find relevant information.

For Realtor recommendations, independent research suggests ChatGPT currently relies heavily on established real estate portals and agent directories.

In Local Falcon's 37,500-query Realtor study, ChatGPT's most-cited sources included:

• Zillow — 39.8%
• Realtor.com — 18.0%
• HomeLight — 7.3%
• Expertise.com — 5.5%
• RateMyAgent — 2.3%
• Effective Agents — 2.2%
• FastExpert — 1.6%

This is a significant finding.

A Realtor may assume:

“If my own website is good, ChatGPT should recommend me.”

But ChatGPT may encounter the Realtor first through an external source.

That could be:

• Zillow,
• Realtor.com,
• a review platform,
• an agent directory,
• a brokerage page,
• or another website discussing that Realtor.

What this means for Realtors

ChatGPT visibility is not only about your website.

It is also about your digital footprint across the web.

AI needs enough external evidence to connect:

Your name → your location → your specialty → your reputation.

If those connections are weak or inconsistent, ChatGPT may have less reason to include you.

2. Gemini: Realtor Websites Appear to Matter More

Gemini does not behave exactly like ChatGPT.

In Local Falcon's Realtor study, Gemini relied on a wider variety of sources and appeared to draw more heavily from Realtors' own websites than some competing platforms.

Its most-cited sources included:

• FastExpert — 24.3%
• Zillow — 17.7%
• Realtor.com — 5.6%
• List With Clever — 3.2%
• Expertise.com — 3.0%
• RealTrends — 3.0%
• Homes.com — 2.4%
• Compass — 2.0%.

Local Falcon concluded that Gemini pulled from agents' own websites more than the other platforms in its test.

This makes website clarity particularly important.

Your website should make it easy for a system to answer:

Who are you?

Where do you work?

What property types do you specialize in?

Who do you help?

What evidence supports your expertise?

A vague website such as:

“Helping you achieve your real estate dreams.”

does not communicate very much.

A much clearer statement would be:

“Jane Smith is a Richmond, BC Realtor specializing in condos, first-time buyers and residential resale across Richmond and Vancouver.”

The second statement creates explicit relationships between:

person + profession + location + specialties.

That type of clarity is useful for both humans and machines.

3. Google AI Mode: Google Business Profile Is Extremely Important

Google AI Mode behaves differently again.

Google describes AI Mode as particularly useful for questions requiring deeper exploration, comparison and reasoning.

Instead of relying on one search, Google says AI Mode can use a technique called query fan-out, where the system runs multiple related searches across different subtopics and data sources before constructing its response.

For example, a buyer might ask:

“Who is a good Realtor in Burnaby for a first-time condo buyer with an $800,000 budget?”

AI Mode could theoretically break that into several related information needs:

• Burnaby Realtors
• condo specialists
• first-time buyer expertise
• local reviews
• brokerage information
• location relevance

Google then retrieves supporting information from its Search index.

But local business data appears to play an unusually important role.

In Local Falcon's Realtor research, 72.5% of Google AI Mode citations came from Google Business Profile.

Its other major sources included:

• Yelp — 4.8%
• Zillow — 4.4%
• FastExpert — 2.3%
• U.S. News — 1.2%
• Realtor.com — 1.0%.

This is dramatically different from ChatGPT.

What this means for Realtors

If you care about Google AI Mode, your Google Business Profile is no longer only a traditional local SEO asset.

It may also influence your visibility in generative search experiences.

That makes basic information especially important:

• Realtor name
• business category
• location
• phone number
• website
• reviews
• business description
• service areas
• up-to-date business information

Google's own documentation also advises businesses to keep their Business Profile information current when trying to appear across Google's AI search experiences.

4. Google AI Overviews: A Smaller and More Selective Surface

Google AI Overviews and Google AI Mode are often grouped together.

They should not be.

Google explicitly states that AI Overviews and AI Mode can use different models and techniques, which means they may produce different responses and different supporting links.

AI Overviews also do not appear for every query.

Google says they are shown only when its systems determine that an AI Overview adds value beyond the standard search experience.

That difference is clearly visible in Realtor recommendation research.

Local Falcon found that AI Overviews often did not trigger at all for queries such as:

“best Realtor near me.”

The report noted that roughly seven times out of ten, the tested AI Overview query returned no AI-generated Realtor answer.

When AI Overviews did provide sources, they included:

• Zillow
• FastExpert
• Google
• Reddit
• U.S. News
• Yelp
• RealTrends
• Realtor.com.

That makes AI Overview visibility fundamentally different from simply ranking in traditional Google search.

You cannot assume:

Page-one Google ranking = AI Overview inclusion.

Different Platforms Recommend Different Realtors

One of the biggest misconceptions about AI visibility is that there is one “AI ranking.”

There isn't.

Local Falcon found that among the Realtors mentioned in its study:

• 82.6% appeared on only one AI platform
• 11.1% appeared on two
• only a small minority appeared across three or more.

BrightLocal found the same general pattern outside real estate.

Its September 2026 study analyzed 200,875 localized AI searches and approximately 1.9 million AI citations across:

• ChatGPT
• Google AI Mode
• Google AI Overviews.

The overlap between platforms was surprisingly low.

Google AI Mode and Google AI Overviews shared only about 29% of recommended businesses.

ChatGPT overlapped only around:

• 20% with AI Overviews
• 19% with AI Mode.

That means a Realtor's visibility should be measured platform by platform.

Being visible in ChatGPT says very little about whether you will appear in Gemini or Google AI Mode.

AI Recommendations Are Much Shorter Than Traditional Search Results

Traditional Google search can give a user dozens or hundreds of potential results.

AI normally gives far fewer.

BrightLocal's 200,000-search study found that the average local AI answer mentioned approximately:

• 4.1 businesses in ChatGPT
• 3.5 businesses in Google AI Mode
• 2.5 businesses in Google AI Overviews.

This changes the competitive environment.

Traditional SEO often asks:

“Can I reach page one?”

AI search increasingly asks:

“Can I make the shortlist?”

If AI returns only three or four names, the difference between being included and excluded becomes significant.

This is one of the fundamental reasons AI visibility matters for Realtors.

Location Still Matters

AI search is conversational, but local proximity has not disappeared.

BrightLocal tested identical local prompts across different physical points within the same city.

The businesses recommended changed substantially depending on where the search originated.

Overlap across different points in the same city was approximately:

• ChatGPT — 36.2%
• Google AI Mode — 46.9%
• Google AI Overviews — 47.4%.

Most recommendations also remained geographically close to the searcher.

Across the platforms tested, roughly 53%–59% of recommendations were within 5 kilometres, and approximately 72%–78% were within 10 kilometres.

For Realtors, this means location signals still matter.

A Vancouver Realtor trying to appear for Richmond searches should make their Richmond expertise explicit.

A generic statement such as:

“Serving the Lower Mainland.”

may be less useful than clearly documented market expertise across:

• Richmond
• Burnaby
• Vancouver
• Coquitlam
• Surrey

provided those claims accurately reflect the Realtor's business.

The Exact Question Matters Too

AI recommendations can change when the wording changes.

Compare these prompts:

“Best Realtor in Vancouver”

versus:

“Best Realtor for a first-time condo buyer in Vancouver”

versus:

“Mandarin-speaking Realtor who specializes in Richmond condos”

versus:

“Realtor experienced in helping families sell a condo and upgrade to a detached home in Coquitlam.”

These are different intents.

AI is not simply trying to identify the most famous Realtor.

It is trying to answer the specific question.

That means specialization can become a visibility advantage.

A Realtor who clearly publishes useful information about:

• first-time buyers,
• condos,
• luxury homes,
• relocation,
• downsizing,
• investment properties,
• specific neighborhoods,

gives AI more evidence for specialized queries.

This is one reason GEO content should focus on actual customer problems, not simply generic market commentary.

Your Website Still Matters

The rise of AI search has led some businesses to wonder whether websites are becoming irrelevant.

Current research suggests the opposite.

BrightLocal analyzed approximately 1.9 million AI citations and found that:

93% of the unique domains used as AI sources were business websites.

Business websites accounted for approximately 42% of all citations in the study.

Google's official guidance reinforces this.

To appear as a supporting link in AI Overviews or AI Mode, a webpage must be indexed and eligible to appear in regular Google Search.

Google also recommends:

• allowing pages to be crawled,
• using clear internal links,
• making important content available in text,
• keeping structured data consistent with visible content,
• maintaining accurate business information,
• and creating useful, reliable, people-first content.

Google specifically states that there is no special AI schema or secret GEO markup required to appear in AI Mode or AI Overviews.

That is an important distinction.

GEO is not about inserting magic keywords into a website.

It is about making your expertise easier to retrieve, understand and verify.

Third-Party Sources Matter Just as Much

For Realtors, your own claims are only one part of your digital identity.

AI systems frequently rely on external sources.

Across Local Falcon's Realtor study, the most-cited sources included:

• Google Business Profile
• Zillow
• FastExpert
• Realtor.com
• Yelp
• U.S. News
• HomeLight
• Expertise.com
• RealTrends
• Reddit.

Together, the top ten sources accounted for approximately 78% of all citations in the study.

This illustrates a fundamental GEO concept:

AI visibility is built across the web, not on one website.

Your website may say you specialize in Richmond condos.

Your brokerage profile may confirm your Richmond market presence.

Your reviews may mention condo purchases.

A local article may quote your Richmond market expertise.

Your Google Business Profile may establish your location.

Together, those sources form a much stronger machine-readable picture than any single claim.

Reviews Matter, But Reviews Alone Are Not Enough

Reviews clearly contribute to the information ecosystem AI uses.

Google Business Profile, Yelp and RateMyAgent all appeared as sources in current local AI research.

But there is little evidence that simply having the largest number of reviews guarantees recommendation.

The same is true for production.

Local Falcon found that four out of five top-producing agents in its benchmark still failed to appear in AI results.

So the better way to think about reviews is:

Reviews are evidence, not a guaranteed ranking factor.

They help establish:

• reputation,
• customer experience,
• specialization,
• location context,
• and independent validation.

Those signals can become particularly useful when reviews contain specific information.

For example:

“Sarah helped us buy our first condo in Richmond.”

contains much more contextual information than:

“Great Realtor. Highly recommended.”

AI Visibility Is Not Stable

Another important difference between classic rankings and AI recommendations is variability.

BrightLocal repeated identical local prompts multiple times and found only:

• 23% average overlap in ChatGPT results
• 21% in AI Mode
• 33% in AI Overviews.

A business that appeared once was likely to appear again about half the time, but not every time.

So asking ChatGPT once:

“Who is the best Realtor in Richmond?”

and checking whether your name appears is not a reliable AI visibility test.

A meaningful test should use:

• multiple prompts,
• multiple platforms,
• repeated runs,
• different customer intents,
• and, for local searches, multiple geographic points.

AI visibility is better understood as probability of inclusion than as a fixed ranking position.

How Should Realtors Improve Their AI Visibility?

There is no guaranteed formula for getting recommended by AI.

Anyone promising a guaranteed ChatGPT ranking should be treated carefully.

However, the available evidence points toward several practical priorities.

1. Make Your Realtor Identity Explicit

Clearly state:

• your full professional name,
• brokerage,
• service areas,
• specialties,
• property types,
• customer segments,
• and professional role.

Do not make AI infer these details from vague marketing language.

2. Build Strong Local Pages

If Richmond is important to your business, create useful Richmond-specific content.

If you specialize in condos, create useful condo-specific content.

If you specialize in first-time buyers, demonstrate that expertise.

The objective is not to create hundreds of thin SEO pages.

It is to create credible evidence of expertise.

3. Keep Your Google Business Profile Accurate

This appears especially important for Google AI Mode and AI Overviews.

Make sure your:

• category,
• address,
• service area,
• website,
• phone number,
• reviews,
• and business details

remain accurate and consistent.

4. Strengthen Third-Party Profiles

Your presence on trusted third-party platforms may influence whether AI can verify who you are.

Relevant platforms can include:

• brokerage websites,
• Realtor.com,
• Zillow,
• RateMyAgent,
• RealTrends,
• local media,
• local business directories,
• professional associations.

The exact importance of each source varies by platform.

5. Create Content Around Real Customer Questions

Instead of only publishing:

“Vancouver Market Update — September 2026”

also answer questions like:

“Should I sell my condo before buying a detached home in Coquitlam?”

“How much down payment does a first-time condo buyer need in BC?”

“What should a Richmond condo seller do before listing?”

These questions help establish the relationship between your expertise and specific consumer needs.

6. Track Multiple AI Platforms

Do not optimize for ChatGPT alone.

The evidence shows substantial disagreement between:

• ChatGPT
• Gemini
• Google AI Mode
• Google AI Overviews.

A useful AI visibility strategy needs to measure them separately.

ChatGPT vs Gemini vs Google AI Mode vs AI Overviews

Based on current research, the platforms can be summarized like this:

These patterns should not be treated as permanent algorithms.

AI search products change rapidly.

But they demonstrate why Realtors should stop thinking about AI as a single channel.

Frequently Asked Questions

How does ChatGPT choose which Realtor to recommend?

ChatGPT can search the web and synthesize information from multiple sources. Current Realtor-specific research suggests it frequently relies on real estate portals and directories such as Zillow, Realtor.com, HomeLight and agent review platforms when answering local Realtor recommendation queries.

Does Gemini recommend the same Realtors as ChatGPT?

Usually not consistently. Local Falcon found that 82.6% of agents appearing in its 37,500-query study appeared on only one AI platform.

Does Google AI Mode use Google Business Profile?

Current independent research indicates that Google Business Profile is extremely important for local AI Mode results. In Local Falcon's Realtor study, Google Business Profile represented 72.5% of AI Mode citations.

Are AI Overviews the same as AI Mode?

No. Google states that AI Overviews and AI Mode can use different models and techniques and therefore may show different responses and supporting links. AI Overviews also appear only when Google's systems determine that they add value to the traditional search results.

Does ranking first on Google mean AI will recommend me?

No. BrightLocal found substantial differences between traditional Google Maps visibility and AI recommendation visibility. A tracked business appeared in approximately 66% of Google Maps searches but only 32%–38% of AI searches in its study.

How many Realtors does AI normally recommend?

There is no fixed number. In BrightLocal's broader local search study, ChatGPT mentioned an average of 4.1 businesses, Google AI Mode 3.5 and AI Overviews 2.5.

Can I guarantee that ChatGPT or Gemini recommends me?

No. AI recommendations vary by platform, prompt, location and even repeated runs of the same query. Optimization can improve the clarity and availability of information about a Realtor, but no legitimate GEO strategy can guarantee inclusion.

The Bottom Line

AI does not appear to choose Realtors using one simple ranking factor.

It builds answers from an information ecosystem.

ChatGPT may lean on real estate portals and directories.

Gemini may rely more heavily on websites and broader web sources.

Google AI Mode appears strongly connected to Google Business Profile and Google's local information ecosystem.

AI Overviews are more selective and may not appear at all for many Realtor recommendation queries.

And none of them consistently agree.

That creates a new challenge for real estate marketing.

For years, Realtors asked:

“Where do I rank on Google?”

The next question is:

“When buyers ask AI who they should work with, how often am I part of the answer?”

That is the foundation of AI visibility for real estate.

And because AI recommendations can change by platform, prompt and location, the only way to understand your actual visibility is to measure it.

Maahes currently runs your selected local questions through the Gemini API each week. You can see whether you appear, how you are described, and which sources are cited when available. This does not test the consumer Gemini app or other platforms discussed here.

Start with an AI Visibility Test.