AI Visibility for Real Estate Agencies

Property decisions start with questions, and more of those questions now go to an AI assistant before a portal or a search box. Here are realistic examples across the journey:

• "Who is the best real estate agent in [neighborhood] for first-time buyers?"
• "What commission do real estate agents charge in [city], and is it negotiable?"
• "Which agencies in [county] specialize in selling older homes?"
• "Is [agency name] a good agency to list my house with?"
• "Is [agent name] a licensed realtor, and which brokerage are they with?"
• "What are the best neighborhoods in [city] for families with a commute downtown?"
• "Should I use a local agent or a discount brokerage to sell my condo?"

The recommendation prompts matter most commercially. The neighborhood and fee prompts matter because the answer often names agencies as sources or next steps.

What you type into a tracker is not what the engine searches. OpenAI's help page on ChatGPT search says it typically rewrites your query into one or more targeted queries. So any prompt set, including yours, is a sample of the real demand. Read whether tools can see the prompts people actually type before treating a prompt list as ground truth.

Each engine reads the web through a different door. As of October 2026, here is how they differ for an agency site:

• ChatGPT answers some questions from its training snapshot and searches the web for others, at the model's discretion. Its search index is fed by OAI-SearchBot, not GPTBot.
• Google AI Overviews and AI Mode draw on Google's index. Google says a page must be indexed and eligible to show with a snippet to appear as a supporting link, with no extra technical requirements. Googlebot renders JavaScript.
• Gemini is a separate Google product, governed in part by the Google-Extended control.
• Perplexity uses its own crawler, PerplexityBot, for its index.
• Microsoft Copilot grounds on the Bing index.

These indexes diverge. In OtterlyAI's 2026 experiment with two AI-generated blog sites, Google deindexed both, Bing kept them, and Copilot kept citing them while Google's AI surfaces almost never did. The lesson for an agency: check Google Search Console and Bing Webmaster Tools separately, and measure each engine on its own.

AI platforms get specific facts wrong about specific businesses. For agencies, the risky facts are the ones that change or that sit in hard-to-read places.

Service areas

Agencies often show coverage as an interactive map or a widget. If the town names appear only after JavaScript runs, a non-rendering crawler sees nothing. The model then guesses from old office addresses or a franchise page that lists a different territory.

Agent licensing and brokerage affiliation

Agents move between brokerages, hold licenses in some states and not others, and change designations. A model's training snapshot freezes whatever was true at its cutoff. OpenAI's model docs list a February 16, 2026 knowledge cutoff for the GPT-5.6 family and August 31, 2025 for ChatGPT's Instant path. An agent who joined you in spring may still be filed under the old brokerage.

Fees and commission

If you don't state how you charge, the engine fills the gap with a generic figure or a competitor's page. A plain sentence on your own site gives it something accurate to quote.

Team, offices and specialties

Departed agents, closed offices and outdated specialties linger on third-party profiles long after you update your site.

The practical move: write each of these facts once, in plain HTML, on a page crawlers can fetch, and keep it consistent with every profile you control.

Many agency sites pull listings from a feed embedded through a JavaScript widget or iframe. Vercel's crawler study found that ChatGPT and Claude crawlers fetch JavaScript files but don't execute them, so they can't read client-side rendered content. Googlebot is the exception because it renders.

Even a perfectly rendered feed has a second problem. A listing can go under contract this afternoon. A model trained months ago, or an index refreshed last week, can't keep up. Expect AI answers to be wrong about specific active listings, and don't build your plan around getting them right.

What to do instead:

• Server-render the stable facts: who you are, where you work, who your agents are, how you charge, how to reach each office.
• Let property portals carry live inventory, and check what those portals say about your agency and agents.
• Treat any AI answer that quotes a price or status for a specific home as unverified until it matches the live listing.

When an engine cites other sites instead of yours, being named accurately on those sites is the route into the answer. That is an association in the evidence, not a guaranteed effect. For agencies, check these:

• Property portals: your agency profile, each agent profile, office addresses and sold history.
• Review profiles: your agency and individual agents, including old entries for agents who left.
• Industry and local directories: see whether directory listings help AI answers before paying for more of them.
• Franchise or brokerage network pages: territory and roster pages are a frequent source of wrong service areas.
• Your licensing regulator's public lookup: confirm each agent's record matches what your site says.
• Local "best agents" roundups and news pages: if an engine keeps citing one, follow getting listed on pages AI engines already cite.

Agents without their own site still appear through these pages, which is why businesses can show up in AI answers without a website. Fix the errors you find at the source: AI answers repeat what these pages say.

Access is where the evidence is strongest, and every check here removes a block rather than adding a trick.

robots.txt, per bot

Rules are per crawler. OpenAI's crawler docs say each setting is independent: GPTBot governs training use, while OAI-SearchBot controls whether you appear in ChatGPT search. A blanket AI block can silently remove search eligibility. A minimal pattern that allows the search crawlers looks like this:

User-agent: OAI-SearchBot
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: Claude-SearchBot
Allow: /

Firewall and CDN reachability

A firewall or CDN can block a crawler that robots.txt allows, often with a 403. From September 15, 2026, Cloudflare blocks Training and Agent crawlers by default on ad-displaying pages for new domains and existing free-tier customers, while Search stays allowed. Check your settings even if you changed nothing. The AI Crawler Checker requests a page as each AI crawler and shows which ones are turned away. To confirm hits in your own logs are genuine, see verifying real AI crawlers vs spoofed user agents.

Index and snippet controls

noindex, nosnippet, max-snippet:0 and data-nosnippet keep content out of AI answers. Templates often apply them by accident, for example to a footer that holds your office addresses.

Raw HTML

View your service area, team and fees pages with JavaScript off. Anything missing there is missing for most non-Google AI crawlers.

Say you run a hypothetical four-office agency with twenty agents and an embedded listings widget. Here is a month of work that stays inside what you control.

1. Days 1 to 7: access. Audit robots.txt per bot, test firewall and CDN responses for each AI crawler, scan for stray noindex and nosnippet tags, and confirm indexing in both Google Search Console and Bing Webmaster Tools.
2. Days 8 to 14: facts in plain HTML. Publish service areas as town and neighborhood names, one agent page each with license state and brokerage, office addresses, and a clear sentence on how you charge. Check each page renders without JavaScript.
3. Days 15 to 21: third-party cleanup. Update portal, review, directory and franchise profiles. Remove departed agents and closed offices. Match every fact to your own pages.
4. Days 22 to 30: baseline. Ask your key buyer and seller prompts on each engine separately and record what is named and what is wrong. The free checker on this site gives a snapshot for three buyer questions. Repeat monthly, because results move with model updates and retrieval choices you can't see.

Skip the tactics without evidence. Lab studies of rewriting content for citation were run on pages already in the model's context, as lab GEO studies vs live engines explains.

Does adding schema markup to my listings help AI engines cite them? There's no credible evidence that it does. Google says structured data isn't required for its generative AI features. Ahrefs' matched-control test found AI Overview citations fell 4.6% after pages added schema, although every page in that study was already heavily cited. Existing markup does no harm, so there's no need to remove it. It just isn't a lever for AI answers, and your time is better spent on access and plain-HTML facts.

Should a real estate agency block GPTBot? That's a training decision, not a visibility one. OpenAI documents GPTBot as the crawler for training its models, and OAI-SearchBot as the crawler for ChatGPT search. The settings are independent, so you can block GPTBot and still allow OAI-SearchBot. Blocking GPTBot does not remove you from ChatGPT search, and allowing it does not get you cited there.

Can I get ChatGPT to call my agency the best in my area? No one can promise that. Engines decide per query whether to search, which pages to retrieve and what to say, and recommendation answers lean toward brands that are already widely mentioned. What you can control is whether engines can reach your pages, whether your facts are accurate there, and whether the third-party pages they cite describe you correctly.

Why does AI list the wrong brokerage for one of my agents? Usually because the answer comes from an old source. That might be a model's training snapshot from before the agent moved, or a portal, franchise or review profile that was never updated. Fix the agent's page on your site in plain HTML, update every third-party profile you control, and confirm the licensing regulator's record matches. Then recheck the answer on each engine over the following weeks.