How to Make Your Real Estate Business Visible in ChatGPT

Buying a house, renting an apartment, finding a real estate agent. People increasingly use ChatGPT for all of these. Instead of visiting several websites and comparing the options themselves, they describe what they're looking for and ask AI to recommend suitable homes or agents.

For real estate businesses, that creates another way for clients to find you. Someone might ask ChatGPT which agents in Denver specialize in historic homes. Or which brokerage could help sell a small multifamily property in Chicago.

The question is: how do you get ChatGPT to recommend your brokerage, cite your website as a source, or show your listings?

I've researched how ChatGPT and other AI platforms choose their sources. Among other things, I analyzed 1,680 AI answers containing 4,616 citations from ChatGPT, Claude, Gemini and Google AI Mode, and compared those sources with Google's regular search results.

In this article, I explain what I would do to improve a brokerage's visibility based on those findings. I look at SEO and GEO, but also at making your own listings directly searchable inside ChatGPT.

A note on my research

My research was done in the Netherlands, with questions asked in Dutch and English. The patterns it found, such as the importance of answering the exact question, are about how AI platforms select pages, and I'd expect them to hold in the US too. The exact percentages may differ in your market, so treat them as indications and test the questions that matter to your business.

How do people search for real estate using ChatGPT?

When someone searches for a home on Google, they usually get a list of portals and brokerage websites. They then have to visit those sites and compare the listings themselves.

ChatGPT works differently. You can describe exactly what you want and refine the search with follow-up questions. For example:

• I'm looking for a condo in Austin with at least two bedrooms and a balcony, for no more than $450,000.
• Which real estate agents in Seattle have experience selling historic homes?
• How do buyer agent commissions work when I sell my house in Phoenix?
• Which brokerage in Miami specializes in commercial real estate?
• Which neighborhoods in Raleigh are good for buying a rental property?

These questions are quite different. In the first, someone is looking for a home. In the second and fourth, they're looking for an agent or brokerage. The others are mainly about gathering information before a decision.

That distinction matters. A listing appearing in ChatGPT doesn't mean ChatGPT will also recommend the agent or brokerage behind it.

The big portals are already inside ChatGPT

Zillow launched an app inside ChatGPT in October 2025, Redfin followed in February 2026, and Realtor.com in March 2026. People can search those portals' listings in a conversation, refine the search as they go, and click through to the portal to contact an agent or schedule a tour.

The same is happening elsewhere: in the Netherlands, for example, rental platform Pararius launched its ChatGPT app in July 2026.

This means ChatGPT doesn't have to rely on individual brokerage websites for every property search. It can also use data that a portal makes available through its app.

Can a brokerage offer its own property search in ChatGPT?

Yes. You don't need to be Zillow to build a property search app for ChatGPT.

OpenAI lets developers create apps and integrations that give ChatGPT access to external data. For a brokerage, that could be a database of available homes, condos, commercial space or new developments.

Say your brokerage has 200 active listings. You could have a ChatGPT app built that lets people search by location, price, beds and baths, square footage and other features. Someone could ask:

Show me the available condos listed by Brokerage X in Denver, with at least two bedrooms and a maximum price of $400,000.

The app would search your current listings and show matching properties, with photos, prices and links to the listing pages. Buyers could explore your inventory without working through all the filters on your website.

Check your MLS rules first. Most listing data comes through the MLS, and IDX and data licensing agreements set where that data may be displayed. The portals' ChatGPT apps have already raised questions about MLS data licensing. Before you send listing data to ChatGPT, confirm with your MLS which data you're allowed to use there.

There's also an important difference between having your own ChatGPT app and being discovered by ChatGPT automatically. People need to find and open your app, and publishing it requires a review by OpenAI. There's no guarantee ChatGPT will suggest your app when someone asks a general question about homes in your area.

That's why I would mainly consider a dedicated ChatGPT app for larger brokerages, franchise networks, developers and specialized property platforms with substantial inventory. For a smaller brokerage, I would first make the existing website and listings easier to discover.

But if you already have a well-structured listings database and the right to use it, a ChatGPT integration is worth exploring. It could become another way for clients to find your properties.

More information: OpenAI, apps and MCP integrations for ChatGPT.

How does ChatGPT decide which agents to recommend?

Take an example. Someone asks ChatGPT:

Which real estate agent in Seattle has experience selling condos on Capitol Hill?

ChatGPT can search the web for this. It might consult brokerage websites, agent profiles, property portals and other publications, and then judge which information is useful for the answer.

There's no fixed checklist that automatically earns your brokerage a recommendation. The sources used vary by question and by AI platform. I saw that in my own research as well.

Ranking well in Google helps, but it isn't enough

In one of my studies, I compared the sources AI platforms cited with Google's top 10 results. For 71 Dutch search queries where I could make an exact match, I found this:

So 60% of the AI citations in this part of my research came from websites that weren't in Google's top 10 for that query.

Ranking well in Google is no guarantee that ChatGPT will cite you, and a website that isn't on page one can still be used as a source. At the same time, websites that ranked higher were cited more often. Traditional SEO still matters.

What I found particularly interesting was how much the platforms differed. Google AI Mode used sources from Google's top 10 far more often than Claude did. ChatGPT was somewhere in between.

For a brokerage, this means you shouldn't only watch your Google rankings. Also find out which agents, brokerages and websites ChatGPT and other AI platforms actually mention.

1. Research the questions you want to be found for

If I were helping a brokerage become more visible in ChatGPT, I would start with the questions potential clients ask. I wouldn't stop at general terms like real estate agent Austin or realtor Denver. I would focus on questions where the brokerage's expertise, market area and specialization really matter.

Take a brokerage in Chicago that works with small investors and commercial property. Relevant questions might include:

• Which real estate agent in Chicago specializes in two-flats and small multifamily buildings?
• Who can help me sell a rental property with tenants in place?
• How do I estimate the value of a leased retail building?
• Which brokerage has experience with industrial space on the South Side?
• What does it cost to sell an investment property through a brokerage?

For a residential brokerage, the questions would look quite different.

Look at the questions people ask before buying or selling

Not every potential client asks for an agent right away. Many start with an informational question:

• Should I sell my house before buying a new one?
• How much is my condo in Denver worth?
• What closing costs does a seller pay?
• How do buyer agent commissions work now?

If your website answers these questions well, ChatGPT may use it as a source. You may not be recommended directly as an agent, but potential clients can still come across your business.

My advice: list around 20 to 30 questions that fit your specialization and market. Then check how ChatGPT answers them, which businesses it mentions and which websites it cites. That shows your current visibility and where the opportunities are. I explain the process in more detail in my guide to researching prompts for AEO.

2. Create pages that answer specific questions well

One of the clearest findings from my AI citation research was the importance of relevance.

I wanted to understand why AI platforms cited some pages and not others. To find out, I compared cited pages with pages from the same websites that weren't cited. That helped me isolate which characteristics of the content itself made a difference.

The strongest relationship was how closely a page matched the question being asked. Pages that answered a question directly were cited more often than pages that only touched on the same general topic.

It may sound obvious, but it has real consequences for a brokerage's content strategy.

Don't rely on one generic services page

Imagine your website has a long page called Our Real Estate Services. It explains that you list homes, represent buyers, help with pricing and work with investors. That's useful, but it probably doesn't answer this question in much detail:

What does it cost to sell a rental property with tenants in place in Chicago?

If that's an important question for your clients, I would create a dedicated page or article about it. You could explain:

• how selling a tenant-occupied property differs from selling a vacant one;
• which factors affect the price;
• which costs the seller should expect;
• how the process works, from listing to closing;
• when working with a specialized agent makes sense.

That genuinely helps the reader. It also makes it easier for search engines and AI platforms to see that your page fits this specific question.

You don't need a separate page for every possible question. Related questions can share a page, as long as each one is answered clearly and completely.

Use your own experience

I would focus on subjects where you can add something from your own experience. Think of how co-op board approvals affect closing timelines in Manhattan, how HOA special assessments come up in condo deals, or what changes when you sell a property with tenants in place.

General explanations of these topics are everywhere. But if you can use your experience with local transactions to explain the problems buyers and sellers actually run into, you add information that isn't on every other website. That makes the page more useful, and gives AI platforms a reason to use your information.

3. Make your specialization clear

When someone asks for an agent in a particular city, several businesses usually qualify. So make it clear what sets yours apart.

Do you specialize in historic homes? Mainly sell luxury condos? Work with investors? Focus on one neighborhood? Then don't leave that only on your homepage. Make it visible on the relevant service and location pages too.

Show your local expertise with specifics

Many real estate websites say something like:

We know the local market better than anyone and guide you personally through every step of buying or selling.

That doesn't tell a client much: almost any agent could say it. I'd rather see specifics:

• Which neighborhoods does the brokerage serve?
• Which property types does it specialize in?
• Which transactions has it handled?
• How does the market differ between neighborhoods or property types?
• Which unusual or complex situations has it dealt with?

Say you regularly sell condos on Capitol Hill in Seattle. I would create a page about buying and selling condos there. You could cover the types of buildings, what to check in older buildings, HOA fees and rules, and local demand. Where possible, use data from your own transactions or reliable public sources.

That gives you a page that doesn't just claim you know Capitol Hill. It shows it.

Keep neighborhood descriptions factual: amenities, housing types, commute options, prices. Statements about who a neighborhood is "right for" can conflict with Fair Housing advertising rules.

Important: don't create dozens of near-identical location pages where only the city name changes. Write about the areas where you actually work and have useful knowledge.

4. Build visibility beyond your own website

Something else stood out in my research: the website a page is published on seems to play an important role in which sources AI platforms choose.

When I compared pages from the same website, several advantages that showed up in the full dataset disappeared. That suggests some characteristics don't explain why a particular page was chosen, but rather which websites get chosen as sources in the first place.

For brokerages, that's an important clue. You can publish excellent articles and still not be used by ChatGPT. So I wouldn't invest in your own content alone. I'd also work on your visibility on other relevant websites.

Where would I start?

First, your existing profiles and listings: your Google Business Profile, your agent profiles on Zillow and Realtor.com, and relevant directories and industry sites. Are your name, office address, specializations and contact details correct and consistent everywhere?

Next, I'd look for ways to get your expertise into external publications. You have data that local news outlets and real estate publications may find interesting. For example, you could research:

• how sale prices have developed in a specific neighborhood;
• differences in days on market between condos and single-family homes;
• how price reductions have changed over the past year;
• changes in demand for different types of commercial space;
• developments in the local rental market.

Publish the findings on your own website and approach relevant media if the results are genuinely newsworthy. That can strengthen your SEO and make your business more visible on websites AI platforms use as sources.

That doesn't mean every mention elsewhere improves your chances of a ChatGPT recommendation. There isn't enough evidence for that. So I'd focus on relevant, trustworthy publications rather than collecting as many mentions as possible.

5. Make your listings easier to discover

So far, this has been about your brokerage. But you also want buyers and renters to find the properties you're marketing. That's why I'd review the quality and accessibility of your listing pages separately.

Provide complete property information

Buyers can be very specific. For example:

I'm looking for a three-bedroom townhouse in Denver with a garage and a yard, for no more than $650,000.

To tell whether your listing matches, a search platform needs enough information. I'd check that every listing page clearly includes:

• the location and property type;
• the price or monthly rent;
• beds, baths and square footage;
• HOA fees and key features such as parking or outdoor space;
• the current status: active, pending or sold;
• anything that makes the property stand out.

Don't put this information only in photos, PDFs or interactive widgets. Make sure the key details are also readable text on the page.

Write a unique description of what makes the property interesting, rather than repeating the MLS remarks everyone else syndicates. For a condo, the HOA and building may matter most. For commercial space, buyers and tenants care more about access, zoning and permitted uses, and available facilities.

Don't forget the portals

Now that Zillow, Redfin and Realtor.com have apps in ChatGPT, accurate and up-to-date listings on those portals matter even more. Most of that data flows from your MLS, so errors there travel everywhere.

A listing on a portal won't automatically appear whenever someone asks ChatGPT a relevant question, though. Which listings appear depends on how each app works. I'd check both your own listing pages and your listings on the portals.

For larger brokerages, I'd also look at whether your own ChatGPT app makes commercial sense.

6. Make sure ChatGPT can access your website

The technical side matters too. To find information online, ChatGPT can use a crawler called OAI-SearchBot, which OpenAI uses to discover websites for ChatGPT search.

Your robots.txt file tells crawlers what they may access. Check that OAI-SearchBot isn't blocked by accident. Also check that important pages work without logging in, that your navigation makes sense, and that listing pages can be reached through internal links. If your listings come through an IDX plugin, check that each listing is a real page on your own domain that a crawler can open.

Don't stop at robots.txt. Since July 2025, Cloudflare blocks AI crawlers by default on new domains, and other firewalls and bot protection can do the same. In my test of 300 websites in the Netherlands, 17% of the random websites that let a normal browser in turned away at least one AI bot, almost always while robots.txt allowed that bot.

A website that's technically hard to reach can hold excellent information that search systems never properly retrieve. You can read more in OpenAI's crawler documentation.

What about schema markup, FAQs and other GEO techniques?

Search for ways to show up in ChatGPT and you'll quickly find advice about schema markup, FAQ sections, bullet points and other techniques that supposedly make content more attractive to AI.

In my research, I found no convincing evidence that pages with schema markup or FAQ sections were consistently cited more often than other pages on the same website. Tables showed a small positive relationship, but that doesn't mean adding a table will get you cited.

So I'd use schema markup for what it's meant for: helping search engines understand your business and website. For a brokerage, that could be RealEstateAgent or another fitting LocalBusiness type. FAQs and tables are useful too, but only when they make your information clearer.

My advice is to prioritize relevant, reliable and complete information. That matters more than adding technical elements that haven't been shown to increase AI citations.

7. Check whether ChatGPT recommends your brokerage

If you don't measure your visibility, you won't know whether your work is paying off. I'd start with a baseline. Take the questions you collected earlier and look at the answers from ChatGPT, Gemini and Google AI Mode. Watch for different kinds of visibility:

I wouldn't only check whether you're mentioned, but also for which questions and how you're described. Say your brokerage shows up for general questions about agents in Seattle, but not for questions about historic homes, while that's exactly your specialty. Then you know where an opportunity lies.

Repeat your measurements

AI answers change. ChatGPT won't necessarily recommend the same businesses every time, and results differ between platforms. So test your most important questions several times and compare the results periodically.

Use the same questions, settings and method each time, and record when you tested and which sources were used. That gives you a better basis for judging whether your visibility is really changing.

You can also use your analytics to see how many visitors arrive from ChatGPT and other AI platforms. Not all AI visibility turns into measurable visits, though: someone may see your brokerage in ChatGPT and call you later. So also watch tour requests, home valuation requests and other inquiries.

Where would I start with a brokerage's visibility in ChatGPT?

If I started on a brokerage's AI visibility tomorrow, I wouldn't begin with dozens of new articles or complicated technical changes. I'd first want to understand where the biggest opportunities are. My approach would look roughly like this:

Which areas need the most attention differs per brokerage. An established brokerage with a large website and strong Google rankings has different opportunities than a new one with little content. And for a commercial brokerage, I'd research different questions and sources than for a residential team serving homeowners.

The technical possibilities differ too. A brokerage that already has clean listing data and the rights to use it is in a better position for a ChatGPT integration than one whose data sits only inside a closed system.

That's why I don't believe in one standard GEO checklist for every real estate website.

Is traditional SEO still enough to get found in ChatGPT?

My research shows that SEO is still an important foundation for visibility in AI answers. Websites that perform well in Google are cited relatively often. At the same time, many AI citations come from websites outside Google's top 10 for the same query.

Traditional SEO gives you a good starting point, but it doesn't make you visible in ChatGPT automatically. For a brokerage, I wouldn't replace an existing SEO strategy with a separate GEO strategy. I'd combine the two.

Make sure your website works properly, that you can be found for the right services and areas, and that you publish information that genuinely helps clients. Then find out which questions people ask AI platforms and which sources answer them, and use that to refine your content.

For brokerages with larger inventories, there's an extra opportunity: a direct integration that lets people search your listings inside ChatGPT. It doesn't replace SEO. It simply gives buyers and renters another way to find what you have.

In the end, the goal isn't to collect as many ChatGPT mentions as possible, or to use new technology because it exists. You want to be found by people looking for property, or for the expertise your brokerage offers. And you want that visibility to bring in new clients.