AI Search for Real Estate Agents: How to Get Found in ChatGPT

AI search for real estate agents is a specific case of answer engine optimization: shaping what AI tools can find and trust about you, so they put your name in the answer when someone asks about buying or selling in your market.

That sounds abstract, so let’s follow one business through this guide. Call it Lakeside Home Group: a made-up four-agent team in Austin, Texas, that works mostly with first-time buyers and with sellers moving up to larger homes. The team ranks reasonably well on Google for a few neighborhood searches and gets most of its business from referrals.

A young couple relocating for work opens ChatGPT and types: “We’re moving to Austin with a $450,000 budget and a 30-minute commute to the Domain. Which neighborhoods should we look at, and who are good buyer’s agents for first-time buyers?”

ChatGPT searches the web, reads a mix of portal pages, review sites, Google Business Profile data and a few agent websites, and writes back an answer: three or four neighborhoods with a line on each, then three or four agents or teams with a short reason for each, and a Sources button underneath. It may also suggest opening the Zillow or Redfin app in the chat to browse homes on a map.

Lakeside wants three things from that moment: to be named, to be described accurately, and to have its own first-time buyer page among the cited sources.

It overlaps heavily with classic local SEO. Google’s own documentation says there are no special requirements for its AI features beyond being indexed and eligible for a snippet in normal Search (Google Search Central). What changes is the target. You’re no longer only trying to rank a page. You’re trying to be the agent a model feels safe repeating. If you’re new to the idea, our guide to what AEO is and how answer engines pick sources covers the basics, and our step-by-step AI search optimization guide walks through the platform-by-platform playbook.

Home buyers mostly use ChatGPT for the research they used to do alone before calling anyone: working out what they can afford, learning how the process works, and comparing neighborhoods.

The cleanest recent numbers come from Bank of America’s 2026 Homebuyer Insights Report. Sparks Research surveyed 2,000 US adults (1,000 homeowners and 1,000 renters) online between 13 April and 10 May 2026. One in five prospective buyers and current homeowners (20%) said they had used AI tools or chatbots for homebuying research in the past year, rising to 28% of millennials and 32% of Gen Z (Mann Report).

What they used it for is the useful part for agents. Among prospective buyers who used AI, 57% used it to estimate affordability, mortgage payments or closing costs, 55% for general education about the buying process, and 52% to research neighborhoods, market trends or property values. The same survey found that buyers still prefer a human for touring homes (55%) and for legal or contract advice (54%).

Realtor.com’s own survey gives a wider, more enthusiastic number. In August 2025 it asked 1,000 US adults who were interested in buying or selling a home, or had done so in the past two years, where they get housing information. 82% said they use AI for real estate insights, with ChatGPT used by 67% and Gemini by 54% (Realtor.com via PR Newswire). The gap between 20% and 82% comes from different questions and samples: Realtor.com asked people already in the market about any real estate insight.

The Realtor.com survey also asked which sources were a good use of time. Agents still came first, but AI was close behind, ahead of friends and family, TV, social media and the news.

None of this replaces agents. NAR’s 2025 Profile of Home Buyers and Sellers, which covers transactions from July 2024 to June 2025, found that 88% of buyers and 91% of sellers used an agent or broker (NAR), and 43% of buyers found their agent through a referral (Virginia Realtors). The same report put the median first-time buyer age at a record 40.

What AI changes is who gets considered before that call. NAR’s own June 2026 coverage described AI as an “early step” in the journey (NAR). For Lakeside, the takeaway is simple: the questions buyers ask AI first (affordability, process, neighborhoods) are the questions its website should answer best.

They are apps that the big US portals built with OpenAI’s Apps SDK, so a buyer can search listings and use portal tools without leaving ChatGPT.

OpenAI introduced apps in ChatGPT on 6 October 2025 at its DevDay event, with Booking.com, Canva, Coursera, Expedia, Figma, Spotify and Zillow as launch partners. You can call an app by starting a message with its name, and ChatGPT can also suggest an app when it fits the conversation. OpenAI’s own example is exactly our topic: a conversation about buying a home can surface Zillow so the user can browse listings on an interactive map (OpenAI). At launch, apps were available to logged-in users on Free, Go, Plus and Pro plans outside the European Economic Area, Switzerland and the UK.

Zillow. Zillow’s app went live the same day for logged-in US users on Free, Plus and Pro plans. A user types something like “Zillow, show me homes for sale in Kansas City” and gets for-sale homes and rentals with photos, maps, prices, core facts, broker and MLS attribution and required disclosures. Tours, agent contact and financing happen on Zillow.com, one click away (Zillow).

Redfin. Redfin followed on 6 February 2026. Its app surfaces listings that Redfin says are “updated directly from the MLS,” lets users refine the search by chatting (its example is “Can you show me homes within 10 miles of a hospital?”), and hands off to Redfin for tours, a Redfin agent or financing (Redfin).

Realtor.com. Realtor.com launched on 30 March 2026 with a different angle. Its app focuses on “pre-search” questions: an affordability calculator, conversational budgeting, neighborhood discovery and rent-versus-buy, then it routes serious buyers back to Realtor.com to see full listings, contact an agent and book tours. Listings in ChatGPT show only a preview, and the company says using MLS data to train the model is “strictly prohibited” (Real Estate News).

It isn’t only a US story. By February 2026, REA Group had launched a realestate.com.au app in ChatGPT for Australian buyers, and France’s Leboncoin had opened its classifieds to ChatGPT search (Online Marketplaces). Portals are also building the same experience on their own sites: Zillow put a conversational “AI mode” into limited beta on 25 March 2026, with a Fair Housing Classifier that checks both user questions and AI responses (Zillow).

If you want to see what the Zillow experience looks like in practice, OpenAI’s DevDay keynote includes a live demo: ChatGPT suggests a city, the Zillow app opens a map of homes for sale, and the presenter narrows it to three-bedroom homes with a yard by simply asking.

OpenAI · YouTube

OpenAI · OpenAI DevDay 2025: Opening Keynote with Sam Altman · watch on YouTube

The key detail for agents is the handoff: buyers browse in ChatGPT but contact an agent on the portal, so your listings and portal profile become one of the first things they see.

Google is changing home search in two ways at once: its AI features answer agent and neighborhood questions, and Google itself has started showing home listings in search results.

AI Mode and AI Overviews. These sit on top of Google’s normal index. Google says they may run “multiple related searches across subtopics and data sources” to build an answer, and that a page only needs to be indexed and eligible for a snippet to be shown as a supporting link (Google Search Central). For agent questions, though, Google’s AI leans on its own local data. In Local Falcon’s study, 72.5% of Google AI Mode’s citations for agent searches came from Google Business Profile listings, with Yelp at 4.8% and Zillow at 4.4%. AI Overviews spread its sources more widely, including Zillow (13.2%), FastExpert (10.5%), Reddit (9.6%) and US News (9.0%) (Local Falcon).

Google’s own listings. In December 2025, Google began testing a carousel of homes for sale in mobile results for searches like “homes for sale chicago” and “homes for sale austin.” Google described it as “a curated selection of properties brought to you in paid partnership between Google and ComeHome” (Search Engine Roundtable). In June 2026, Google said the program, run through HouseCanary’s ComeHome under agreements with each MLS, would reach consumers in all 50 states that summer. Listing brokers get “prominent attribution” and agents get click-to-contact on listing pages at no extra cost, while “promoted” buyer’s agents can appear through Local Services Ads (Real Estate News). By July, four MLSs had signed on: Bright MLS, California Regional MLS, San Diego MLS and My State MLS (RISMedia).

So for Google, the agent’s checklist looks like this: a complete Google Business Profile for AI Mode, indexable pages for AI Overviews, and, if your MLS participates, accurate listings with your attribution in Google’s own carousel. We cover AI Mode specifically in how to optimize for Google AI Mode.

For a team like Lakeside that has spent years on its website and almost nothing on its Google Business Profile, that is exactly backwards for Google’s AI.

ChatGPT decides by searching and reading. When a question needs current or local information, it runs web searches, reads what comes back, and names the agents that the sources it trusts actually support.

Here’s what we know from OpenAI’s documentation. ChatGPT searches the web automatically when a question might benefit from it. Search queries can be sent to third-party search providers, with Bing named as an example, and ChatGPT uses general location information based on your IP address to make results more relevant (OpenAI Help Center). OpenAI also runs its own search crawler, OAI-SearchBot, which is “used to surface websites in search results in ChatGPT’s search features” (OpenAI).

Two newer features make local answers sharper. Since late March 2026, users can opt in to share their device location, including precise location, so “near me” questions return nearby results (Search Engine Roundtable). And since December 2025, clicking a business name in ChatGPT can open a local knowledge panel; one business owner reported that his panel mirrored his Google Business Profile and updated shortly after he changed it (Search Engine Roundtable).

What OpenAI doesn’t publish is how it ranks sources. So we have to look at what it actually cites. In Local Falcon’s agent study, ChatGPT took 39.8% of its citations from Zillow, 18.0% from Realtor.com, 7.3% from HomeLight and 5.5% from Expertise.com. It leaned harder on portals and directories than any other platform. And when it cited Zillow, the link usually pointed to a city-wide agent directory page (86.9% of Zillow citations overall), not an individual agent profile.

So a recommendation gets built from the city’s agent directories, a few “best agents” roundups, review sites and some agent websites, and ChatGPT names the agents who show up repeatedly with consistent signals. A half-empty Zillow profile gives it little reason to pick Lakeside.

They name different agents because they read different sources, and because AI answers aren’t fixed rankings in the first place.

The per-platform split in Local Falcon’s study is the clearest explanation we’ve seen. ChatGPT is portal-heavy. Google AI Mode is almost entirely Google Business Profile. Gemini draws on FastExpert, Zillow and, more than any other platform, agents’ own websites. AI Overviews mixes portals, review sites, Reddit and US News. Grok leans on Zillow, US News, FastExpert and Yelp.

On top of that, answers move. BrightLocal ran the same local searches repeatedly over 60 days and found that only 20% to 33% of businesses overlapped between repeat runs on average (ChatGPT 23%, AI Mode 21%, AI Overviews 33%). Answers also changed with location inside the same town: overlap across grid points was 36.2% for ChatGPT and about 47% for Google’s two AI surfaces. ChatGPT named an average of 4.1 businesses per answer, AI Mode 3.5 and AI Overviews 2.5 (BrightLocal).

One more BrightLocal detail matters for agents. Storefront businesses were mentioned slightly more often than service-area businesses (33.6% versus 29.4%). Many agents run as service-area businesses with a hidden address, which is often the right setup under Google’s rules, but it means the rest of your signals need to work harder.

So one screenshot of one answer tells you almost nothing. Our guide on how to measure AI visibility covers that method in depth, and we’ll set up a simple version for Lakeside below.

They mean that every page, profile and FAQ you write to win AI answers follows the same rules as any other real estate advertising. AI search creates no loopholes; it multiplies the places a careless sentence gets repeated.

The Fair Housing Act. In the US, 42 U.S.C. 3604(c) makes it unlawful to publish any notice, statement or advertisement about the sale or rental of a dwelling that indicates a preference, limitation or discrimination based on race, color, religion, sex, handicap, familial status or national origin (Cornell LII). That applies to a neighborhood guide written for ChatGPT exactly as it applies to a print ad. “Perfect for young professionals,” “great for families” or descriptions of who lives in an area can cross the line. Describe the home and the area (square footage, the park two blocks away, the commute), not the people you imagine living there.

The federal guidance changed recently, but the law didn’t. HUD had issued guidance on applying the Fair Housing Act to advertising through digital platforms on 29 April 2024. In a Federal Register notice published on 6 April 2026, HUD listed it among eight fair housing guidance documents it withdrew, effective 17 September 2025 (Orrick InfoBytes). Withdrawing guidance doesn’t change the text of the statute above, and state and local fair housing laws can go further.

The NAR Code of Ethics. If you’re a Realtor, Article 12 requires you to “present a true picture” in advertising and marketing, and Standard of Practice 12-5 requires your firm’s name to be disclosed “in a reasonable and readily apparent manner.” NAR’s guidance notes that calling something marketing rather than advertising doesn’t avoid this, that a team name alone isn’t the company name, and that online ads can meet the rule with a link to a page showing the required disclosures, though many states require the company name in the ad itself (NAR; the article is from NAR’s archive, so check the current Code).

The FTC’s review rule. The FTC’s final rule, announced on 14 August 2024, bans fake reviews (including AI-generated ones), buying reviews that depend on a positive or negative sentiment, undisclosed reviews from insiders, and review suppression through threats or intimidation (FTC). Reviews are one of the strongest signals AI tools read, which makes them tempting to game. Don’t.

You make your site readable by letting the right crawlers in and putting your important content where they can see it. This is the fastest part of the whole job, and it’s often where agent websites quietly fail.

Allow the search crawlers. ChatGPT search relies on OAI-SearchBot. Blocking GPTBot only signals that your content shouldn’t be used to train OpenAI’s models; OpenAI says each setting is independent, so you can block training and still appear in search. After a robots.txt change, OpenAI says it can take about 24 hours for its search systems to adjust (OpenAI). Make sure Googlebot and Bingbot are allowed too, since Google’s AI features use Google’s index and ChatGPT can send queries to providers such as Bing. If you’re weighing what to block, our guide to blocking AI crawlers walks through the trade-offs.

Check the firewall, not just robots.txt. Many real estate sites run on platforms with a CDN or bot protection in front. Those can block AI crawlers even when robots.txt allows them. Our free AI crawler checker tests whether the main AI bots can reach a page.

Watch your IDX setup. Many agent websites show listings through an IDX widget or iframe that loads after the page. That’s fine for listings, which the portals already cover, but don’t let your whole site depend on it. Your bio, your area pages and your FAQs should be plain HTML text.

Keep key pages indexed. Check Google Search Console and Bing Webmaster Tools for your homepage, about page, team bios and area pages. A page that isn’t indexed can’t be a supporting link in AI Overviews or AI Mode.

Don’t hide facts in images. Badges as images, sales numbers in PDFs and testimonials in script-loaded sliders are invisible to most crawlers. Put each claim, with its source and year, in text.

In our made-up example, Lakeside’s bot protection blocks several AI crawlers and its neighborhood guides are image-heavy PDFs. Both problems are common on agent sites, and both take days, not months, to fix.

Set it up exactly the way Google’s guidelines describe, then make it the most complete, most accurate version of you anywhere online. For Google’s AI, it’s close to the whole game.

Know which profile you should have. Google’s guidelines define an individual practitioner as “a public-facing professional, typically with their own customer base.” Practitioners in public-facing roles can have their own profile, but they must be contactable directly at the verified location during stated hours, and support staff shouldn’t create profiles. Where several practitioners work at one location, the brokerage gets its own profile and each practitioner’s profile carries only their name. If you are the only public-facing practitioner at a branded location, Google suggests sharing one profile named “[brand/company]: [practitioner name]” (Google Business Profile Help).

Use your real name only. Your profile name should match your real-world name. Adding service or location keywords (“Jane Doe Realtor Best Austin Homes”) isn’t permitted and can get the profile suspended.

Get the location right. Virtual offices and mailboxes aren’t eligible. If you serve clients at their homes or at listings rather than at an office, set it up as a service-area business, hide the address, and keep the service area within roughly two hours’ drive of your base, as Google’s guidelines describe.

Fill every field that’s true. Primary category, the areas you serve, hours, phone, website link to the right page (your team or bio page, not just the homepage), services such as buyer representation and listing services, languages spoken, photos of you and your team, and regular posts about real market activity.

Keep it consistent. The same name, brokerage, phone and website on Google, Zillow, Realtor.com, Yelp and your own site. AI tools cross-check sources, and conflicting details give them a reason to pick someone else.

We go deeper on the local side, including how Google Business Profile feeds AI answers for every kind of local business, in our guide to AEO for local businesses.

Publish pages that answer the exact questions buyers and sellers ask AI, in your market, with specifics a model can repeat. That’s what turns your website from a brochure into a source.

Go back to the Bank of America numbers. Buyers who use AI ask about affordability, the buying process and neighborhoods. Those three topics are your content plan.

Area pages. One page per neighborhood you really work in. Open with a two-sentence answer to the most common question, then add dated median prices with a source, home types, commute times, and named parks, schools and transit. Describe the place, not the people.

Process pages for your market. Buyer closing costs in your state, typical time from offer to close, how buyer agreements work locally. Answer first, then explain.

Who-you-help pages. “First-time buyers in Austin” or “selling an inherited home in Travis County.” These match how people phrase questions to ChatGPT and give AI a clear reason to name you for a situation.

Verified proof. Your production, in text, with the source and year (“RealTrends Verified, 2026”), your license number and brokerage, and the neighborhoods where you’ve closed deals.

Now, about the “best agents in [city]” pages. Local Falcon found that 18 of the 25 most-cited agent websites had one, usually listing their own owner first. It plainly works on today’s AI tools. But think about Article 12 before copying it. A page that ranks yourself as the top agent in your city, dressed up as a neutral guide, is hard to square with presenting “a true picture.” Our recommendation is a transparent alternative: a guide to choosing an agent in your city that explains what to look for, says clearly that you’re one of the options and why, and links to independent sources such as RealTrends or your portal reviews. It answers the same question without misleading anyone.

Here’s how that looks for Lakeside’s first-time buyer page.

The “after” version is still a made-up example, and a real page would date and source every price. But it answers a question with specific places and numbers, and it makes sense on its own. Our AEO content checker scores a page for that answer-first structure in your browser.

Forums matter in some markets too: Reddit made Local Falcon’s top-ten source list, and our guide on whether Reddit helps AI visibility explains how to take part without breaking community rules.

Use the schema.org RealEstateAgent type on your main business page and keep every value identical to what visitors see and to your Google Business Profile.

RealEstateAgent sits under LocalBusiness, which itself inherits from both Organization and Place (schema.org). That means you can use the usual local business properties: name, address, telephone, email, openingHours, areaServed, priceRange, employee for each agent on the team, and sameAs links to your Google Business Profile, Zillow and Realtor.com profiles. For team pages, a Person entry for each agent, connected to the brokerage with worksFor, makes the relationship explicit.

Keep expectations realistic. Google says there’s no special schema needed to appear in AI Overviews or AI Mode, and that structured data should match the visible text on the page (Google Search Central). Schema doesn’t earn recommendations on its own. What it does is remove ambiguity: it tells machines that “Lakeside Home Group” is a real estate agent, where it operates, who works there and which profiles belong to it.

Two cautions. Don’t add review markup for reviews collected on other sites; that isn’t what review snippets are for. And don’t put listing data in schema in ways your MLS rules don’t allow. Our free schema generator builds valid LocalBusiness-style markup, and our guide to schema markup for AEO explains which types actually help.

Measure the things AI search actually changes, and start with the fixes that are fast and cheap before the work that takes months.

What to measure. Track four numbers every month: mention rate (the share of your prompts where you’re named), accuracy (whether the description is right), citation rate (how often your own pages or profiles are cited) and AI referral traffic to your site. Add the leads that mention ChatGPT or “AI” in your intake form; ask every new client how they found you and record the answer. Then compare by platform and by area, because the earlier data shows they behave differently.

How long it takes. Crawler changes are the quickest: OpenAI says about 24 hours for ChatGPT search. Profile and page changes need to be recrawled and reread, so think in weeks. Reviews, list inclusions and mentions on trusted sites take months, and that’s the work that moves you from occasionally named to regularly named.

For Lakeside Home Group, week one means unblocking the AI crawlers and running a 25-question baseline. The next few weeks go to its Google Business Profile, its Zillow and Realtor.com profiles, and a review request for every closed client. Over two months it rewrites five pages: first-time buying in Austin, Texas closing costs and three neighborhood guides. Then it keeps going, because agents who get named consistently are the ones whose evidence keeps growing.

None of this is a trick, and none of it is guaranteed. AI tools change their sources, portals change their apps, and answers vary from run to run. But the work itself (being findable, consistent, reviewed and genuinely useful) is the same work that wins referrals. If you’d like help with it, our AEO service starts with exactly this audit.

1. Zillow: Zillow debuts the only real estate app in ChatGPT (6 Oct 2025)
2. Zillow: How Zillow's app in ChatGPT expands listing reach and protects industry rules (9 Oct 2025)
3. OpenAI: Introducing apps in ChatGPT and the new Apps SDK (6 Oct 2025)
4. OpenAI on X: You can now chat with apps in ChatGPT (6 Oct 2025)
5. OpenAI on YouTube: OpenAI DevDay 2025 Opening Keynote with Sam Altman
6. Redfin: Redfin debuts its home search app in ChatGPT (6 Feb 2026)
7. Real Estate News: Realtor.com, the latest portal to launch a search app in ChatGPT (30 Mar 2026)
8. Online Marketplaces: ChatGPT integrations for Redfin, REA Group and Leboncoin (13 Feb 2026)
9. Zillow: Zillow debuts AI mode (25 Mar 2026)
10. Realtor.com survey: 82% of Americans use AI for housing market information (9 Oct 2025, 1,000 US adults)
11. Mann Report: Bank of America 2026 Homebuyer Insights Report (23 Jun 2026, 2,000 US adults)
12. NAR: AI becomes early step in homebuying journey (30 Jun 2026)
13. NAR: Top 10 takeaways from NAR's 2025 Profile of Home Buyers and Sellers (3 Nov 2025)
14. Virginia Realtors: Key takeaways from NAR's 2025 Profile of Home Buyers and Sellers (8 Dec 2025)
15. Local Falcon: The Realtor AI Visibility Index (18 Aug 2026, 37,500 AI searches)
16. Inman: 91.5% of established agents don't show up in AI search (20 Aug 2026)
17. InboundREM: AI search real estate winners, April 2026 dataset (27 May 2026)
18. BrightLocal: What 200k local AI searches tell us about which businesses get recommended (16 Sep 2026)
19. BrightLocal: Uncovering ChatGPT search sources for local businesses (12 Dec 2024, 800 searches)
20. HousingWire: Does Zillow's ChatGPT integration violate MLS rules? (8 Oct 2025)
21. HousingWire: Redfin launches a ChatGPT app, raising MLS data licensing questions (6 Feb 2026)
22. Search Engine Roundtable: Google tests property listings in search results (15 Dec 2025)
23. Real Estate News: Google to serve up real estate listings nationwide (11 Jun 2026)
24. RISMedia: Google enters the portal wars (14 Jul 2026)
25. OpenAI: Overview of OpenAI crawlers
26. OpenAI Help Center: ChatGPT search
27. Search Engine Roundtable: ChatGPT enables location sharing for more localized near me results (31 Mar 2026)
28. Search Engine Roundtable: ChatGPT gains local knowledge panels (22 Dec 2025)
29. Google Search Central: AI features and your website (updated 10 Dec 2025)
30. Google Business Profile Help: Guidelines for representing your business on Google
31. Schema.org: RealEstateAgent
32. Cornell LII: 42 U.S. Code 3604, discrimination in the sale or rental of housing
33. Orrick InfoBytes: HUD withdraws eight fair housing guidance documents (10 Apr 2026)
34. NAR: Advertising, no team is an island (Article 12, archive, 2017)
35. FTC: Final rule banning fake reviews and testimonials (14 Aug 2024)
36. Malay Mail / Media OutReach: LINE OpenChat reaches 24 million monthly active users in Thailand (5 Nov 2026)
37. The Nation: DDproperty 2025 platform data on where Thais search for homes (17 Jan 2026)

Keep reading

• AEO basicsWhat Is AEO? Answer Engine Optimization Explained
• Measurement & localAEO for Local Business: Get Cited by AI Near You
• IndustriesHow to Get Your Law Firm Recommended by ChatGPT
• IndustriesHow to Get Your Hotel Recommended by ChatGPT and AI Search