Landing Pages for ChatGPT: What Gets Real Estate Agents Named (We Tested 20 Questions)
We asked ChatGPT 20 buyer and seller questions. Specific questions surfaced agents’ own service pages. Here is the one page to build, plus a free course.
Ryan Rockwood · 7 min read

On this page
- How ChatGPT decides who to name
- What we found when we asked ChatGPT 20 questions
- The six steps (the course in brief)
- 1. Know how the answer gets built
- 2. Pick two or three things you can prove
- 3. Write the questions people actually type
- 4. Check who gets named, and why
- 5. Build the page
- 6. Run the 5-check audit, publish, recheck
- How can I check if it’s working?
- Get the free course and workbook
- Sources
The short version: when a buyer asks ChatGPT which agent to call for something specific, it searches the web and names agents whose pages give it a clear reason. In our test of 20 real questions, broad questions mostly surfaced portal reviews and brokerage bios. Specific questions often surfaced the agent’s own page about that exact service. So build one strong page for each of the two or three things you can prove.
- Pick 2–3 service + area pairs you can prove, not 50 town pages.
- Put a labeled quick answer at the top, then checkable facts, your real steps, official local links and the questions people ask.
- Skip “best” and “top”. ChatGPT repeated proof, not adjectives.
- Get the free 30-minute course and workbook.
Watch on YouTube · Watch the Short · Download the video
Agents keep asking us the same three things about AI search. How does ChatGPT decide who to recommend? How can I make sure it’s me? And how can I even check? A course making the rounds promises a “landing page for ChatGPT” fix. We wanted our own answer, from our own test and from what OpenAI, Google and Bing actually publish. Here it is, plus a free course you can finish in half an hour.
How ChatGPT decides who to name
For a question like “who’s a good agent for inherited homes in Charlotte?”, ChatGPT usually searches the web, reads a handful of pages and writes an answer with links to the pages it used. OpenAI says it ranks results using “multiple factors intended to help users find relevant, reliable information” and that “placement is not guaranteed” (OpenAI Help Center). It can also use your approximate location for local results and may rewrite a “near me” question into a search with the city name.
Two practical points from OpenAI’s crawler documentation: your site has to let OAI-SearchBot in to appear in ChatGPT search answers, and blocking GPTBot (the training crawler) doesn’t affect that (OpenAI crawlers). Ads in ChatGPT are labeled and kept separate, and OpenAI says they “do not influence the answers” (OpenAI).
Google is just as plain about its AI Overviews and AI Mode: “There are no additional requirements… nor other special optimizations necessary” (Google Search Central). Its newer guide says optimizing for generative AI search “is… still SEO”, and it uses a real estate example: “7 Tips for First-Time Homebuyers” is commodity content, while a first-hand story about an inspection decision isn’t (Google). Bing says the same crawling and indexing foundation feeds Copilot, and that “well-structured, accurate, and focused content” improves citation accuracy (Bing Webmaster Guidelines).
What we found when we asked ChatGPT 20 questions
On November 16, 2026 we sent 20 buyer and seller questions to OpenAI’s search-enabled model (gpt-5.5 with web search, through OpenAI’s API) for three markets: Pasadena, California; Charlotte, North Carolina; and Scarsdale, New York. Each question ran once, in a fresh session with the location set to that city. Then we logged every agent, team or brokerage named, the reason given, and the page linked beside each name.
| Question type | Names | Portal or review profile | Brokerage bio | Agent’s own site | …of those, a page about that exact service |
|---|---|---|---|---|---|
| Broad (7): “good realtor in…”, “best realtor for…”, “sell my house in…” | 28 | 19 | 9 | 5 | 2 |
| Specific (13): inherited homes, ADUs, relocation, new construction, co-ops, staging… | 60 | 8 | 24 | 24 | 16 |
- All 20 answers named agents, 4.4 on average. None declined.
- The reason given for a name was nearly always a line from the linked page: a stated focus (“focused on ADU-ready homes”), a designation, years and sales, a local role, licensed in two states, or a step-by-step description of the service.
- For specific questions, 16 names came with a link to a page about that exact service, like a relocation page, a new-construction page or an inherited-home page.
Limits, honestly: it’s a small, one-day sample with one run per question, and answers change between runs. The API with web search is close to the ChatGPT app but not identical (the app can add maps, business cards, memory and personalization). A name can be backed by more than one kind of page, so the columns overlap. It shows what got cited, not that any page causes a mention. We don’t name the agents.
That fits outside research too. A 2024 study of generative search (Aggarwal and colleagues, published at KDD) found that adding citations, quotations and statistics raised a source’s visibility in AI-written answers by up to 40% on the authors’ benchmark. It wasn’t tested on ChatGPT, so treat it as direction, not a promise (GEO study).
The six steps (the course in brief)
1. Know how the answer gets built
Names come from pages. Ask ChatGPT one question about a service you offer in your area and note who’s named and which pages are linked. That’s your baseline.
2. Pick two or three things you can prove
Good pairs are specific: “selling inherited homes in [County]”, “buying a home with an ADU in [City]”, “buying before you move to [Metro]”. For each, ask: can I prove it, do people ask about it, and is anyone clearly answering it? Two or three real pages beat fifty thin ones. Google warns that pumping out a page for every search variation, mainly to sway rankings or AI answers, breaks its scaled-content spam policy (Google). Keep it about services and properties, never people, because fair-housing advertising rules apply to your website too (24 CFR 100.75).
3. Write the questions people actually type
People don’t type “probate real estate services”. They type “Who can help me sell my mom’s house in Charlotte? It’s in probate.” Write 8–10 questions per pair: who, how, and what-if. Pull them from your sent email, call notes, open houses and the follow-ups ChatGPT suggests.
4. Check who gets named, and why
Ask your “who” questions in fresh chats. Log who’s named, the page linked beside each name, and the reason given. Then find the reason gap: reasons others were named that you could claim truthfully but your site doesn’t say. That gap is your page’s to-do list. Recheck in 30 days; one answer is a snapshot, a few checks show a pattern.
5. Build the page
- Headline: the service and the area in plain words.
- Quick answer, labeled with its question, in 2–4 sentences near the top: who you are, what you do, where and how.
- Key facts: checkable proof with dates.
- How I do it: your real steps, plus one first-hand story with no private details.
- What to know here: local rules linked to the official city, county or state page.
- When to think twice: one honest paragraph.
- Questions people ask: 6–10, each answered in its first sentence.
- Contact block: name, license and state, brokerage, address and phone exactly as on your Google Business Profile. NAR’s Code asks for your firm name and state of licensure on your website (NAR Code of Ethics).
- Labels for search engines (schema): RealEstateAgent, a Person record with your license, and FAQPage, matching the visible text word for word (Google structured data guidelines). FAQ markup no longer produces FAQ boxes in Google (they stopped on May 7, 2026); it’s there to help search engines and AI tools understand the page (Google).
Our drafting prompt writes a first draft from your real facts and leaves gaps marked instead of making things up. There’s also a finished sample page for an ADU service in Pasadena.
6. Run the 5-check audit, publish, recheck
- Says it plainly: the first 150 words name the service, the area and you.
- Proves it: at least three checkable facts; no unsupported “best”, “top” or “#1”.
- Can be found and read: live text, linked from your menu, in your sitemap, not blocking OAI-SearchBot or Bingbot, fast on a phone.
- Labels match: schema matches the visible text; name, address and phone match your Business Profile.
- Fair and sourced: homes, services and rules, never people; local facts linked to official sources.
How can I check if it’s working?
- Re-run your questions in ChatGPT every month and log them in the workbook. Look for patterns, not one answer.
- Bing Webmaster Tools’ AI Performance report shows when Copilot cites your pages and the searches it used (Bing). Use IndexNow or URL submission when you publish.
- Claim your business in ChatGPT’s Business Hub to see mentions and website clicks. Our step-by-step guide walks through it.
- Google Search Console shows whether the page is indexed. Google says that’s the main requirement for its AI features.
Get the free course and workbook
The Get Named by ChatGPT course has all six lessons, each ending in one action, plus a Google Sheet workbook (questions list, naming-check log, self-scoring 5-check audit, page outline) and the drafting prompt. It takes about 30 minutes.
No page guarantees ChatGPT will name you. What a focused page does is give ChatGPT, Google and real buyers a clear, provable reason to pick you for one thing. Want a second set of eyes? Ask us, we’ll help you check.
Sources
- OpenAI Help Center — Searching the web with ChatGPT — help.openai.com/en/articles/9237897-searching-the-web-with-chatgpt
- OpenAI — Overview of OpenAI crawlers — platform.openai.com/docs/bots
- OpenAI — Our approach to advertising and expanding access (Jan 16, 2026) — openai.com/index/our-approach-to-advertising-and-expanding-access/
- Google Search Central — AI features and your website — developers.google.com/search/docs/appearance/ai-features
- Google Search Central — Optimizing your website for generative AI features on Google Search — developers.google.com/search/docs/fundamentals/ai-optimization-guide
- Bing Webmaster Tools — Bing Webmaster Guidelines — www.bing.com/webmasters/help/webmaster-guidelines-30fba23a
- Bing Webmaster Blog — Introducing AI Performance in Bing Webmaster Tools (Feb 2026) — blogs.bing.com/webmaster/2026/2/Introducing-AI-Performance-in-Bing-Web
- Google Search Central — General structured data guidelines — developers.google.com/search/docs/appearance/structured-data/sd-polici
- Google Search Central — Latest Google Search documentation updates (FAQ rich result removal) — developers.google.com/search/updates
- eCFR — 24 CFR 100.75, Discriminatory advertisements, statements and notices — www.ecfr.gov/current/title-24/subtitle-B/chapter-I/subchapter-A/part-1
- National Association of REALTORS® — 2026 Code of Ethics and Standards of Practice — www.nar.realtor/about-nar/governing-documents/code-of-ethics/2026-code
- Aggarwal et al. — GEO: Generative Engine Optimization (KDD 2024) — arxiv.org/abs/2311.09735
- IndexNow — IndexNow.org / Bing — www.bing.com/indexnow
- Virtual Results — 20-question ChatGPT naming test, November 16, 2026 (method and limits above)


