For reporters and editors
CiteVantage press kit: sourced findings on how AI search names real estate agents
By Abdul Subkhan Updated 12 October 2026 CC BY 4.0, free to quote with a link back
This CiteVantage press kit gives a reporter what a story needs without a first email to us. Each finding is one sentence made to be quoted whole, with its figure, its study, its date and a link to the exact page. Methods, data files, chart images, citation text, logo files and a founder bio follow it.
Findings on this page
7
findings, each one sentence written to be quoted whole
Quotable findings from CiteVantage's real estate AI search research
Six of the seven come from our own studies and one is a third-party figure, credited to its publisher. The ecommerce result is labelled as ecommerce. Each Copy button puts the sentence and its source line on your clipboard together.
-
94.0%
In 2,000 AI answers to 500 Houston real estate queries, 94.0% of the Houston Association of Realtors' 50,000-plus members were never named.
Hand-checked on 394 names drawn at random, 95% interval 93.6 to 94.5. The study's own headline rounds this to 94%.
- Date
- Published 23 September 2026
- Fieldwork
- 18 to 22 September 2026
-
83%
83% of the websites AI engines cite on real estate questions are cited by only one engine: across 150 questions and 747 measurements, 1,077 of 1,297 cited domains reached a single engine and just 3 reached all five.
The questions are the ones a real estate professional asks about marketing and AI visibility, not consumer property search.
- Date
- Published 13 September 2026, updated 23 September 2026
- Fieldwork
- 8 to 12 September 2026
-
87%
87% of the brands we tested were invisible to AI: 158 of 181 United States ecommerce brands were neither named nor cited by ChatGPT, Perplexity, Gemini or Google AI Overviews when each engine was asked a buying question in that brand's own product category.
Every brand was paying for Meta ads, so the sample is not random. An earlier single-engine pass found a lower share, and the four-engine result is the one to quote.
- Date
- Published 8 August 2026, updated 11 September 2026
- Fieldwork
- June 2026
-
82.6%
In 2,000 AI answers to 500 Houston queries, 82.6% of the agents, teams and brokerages named appeared in just one of 40 Houston areas.
Hand-checked, 95% interval 79.7 to 85.6. Local Falcon's 82.6% for agents named by one platform counts platforms, a different measure.
- Date
- Published 23 September 2026
- Fieldwork
- 18 to 22 September 2026
-
88%
Across 747 measured AI answers to real estate questions, 88% of all citations went to ordinary company websites rather than to portals, trade press or directories, and all four major property portals combined drew 35 citations against 55 for YouTube alone.
Site type is our own rule-based classification, published with the list. Same questions as the 150-query study.
- Date
- Published 13 September 2026, updated 23 September 2026
- Fieldwork
- 8 to 12 September 2026
-
36.6%
When AI platforms were asked for the best real estate agent near me in the 100 largest US cities, Google Business Profile supplied 36.6% of their citations, and 72.5% of Google AI Mode's.
Vendor measurement: 37,500 queries, 3 phrases, 5 platforms, 100 largest US cities. The figure belongs to Local Falcon, so cite Local Falcon first.
- Date
- Published 18 August 2026
- Fieldwork
- August 2026
-
8 of 8
On a site hosted on Hostinger, the host's CDN refused OpenAI's GPTBot on 8 of 8 cache-miss requests with HTTP 429, while 56 of 56 requests from seven other identities returned 200.
The test is a few minutes of requests on one host. It shows a rule on one platform, not how common the rule is.
- Date
- Published 14 September 2026
- Fieldwork
- measured 14 September 2026
How each study was run
Each study is summarised here by its numbers, its dates and what limits it. The full method, the limits and the row-level data sit on the study page, and the figures below are the ones that page states. Read the limits before quoting, because several of these studies rest on one run in one window.
-
The 500-Query Houston Real Estate AI Visibility Study
CiteVantage
Full method and limits for Houston 500-query study- 500queries
- 4AI surfaces
- 2,000answers, 1,917 counted
- Engines
- ChatGPT, Gemini (API-based), Gemini (web), Perplexity
- Dates
- Fieldwork 18 to 22 September 2026. Published 23 September 2026.
- Limits
- 60 answers about the wrong place were excluded, and 23 further answers were excluded because they named businesses from outside the Houston metro. Names were extracted by a heuristic, and the agent figures come from 394 names labelled by hand. One run in one window, and CiteVantage sells AI visibility work to real estate firms.
-
- 150queries
- 5AI engines
- 747measurements, 3,038 citations
- Engines
- ChatGPT, Perplexity, Gemini, Google AI Overviews, Copilot
- Dates
- Fieldwork 8 to 12 September 2026. Published 13 September 2026, citation unit corrected 23 September 2026.
- Limits
- Grok was excluded because only 19 of 150 queries completed. Gemini returned a source list on a minority of answers, so its count is a floor. One run in one week, and CiteVantage's own domain appears in the data.
-
Which Domains AI Cites on Real Estate Questions
CiteVantage
Full method and limits for Open list of cited domains- 1,297domains recorded
- 747measurements
- 5AI engines
- Engines
- The same five engines and the same questions as the 150-query study
- Dates
- Fieldwork 8 to 12 September 2026. Published 13 September 2026, updated 23 September 2026.
- Limits
- Site type is a rule-based classification, with the rules published. The data cannot support a per-city cut, so there is none.
-
- 181US ecommerce brands
- 4AI engines
- 1unbranded question per brand and engine
- Engines
- ChatGPT, Perplexity, Google AI Overviews, Gemini
- Dates
- Sample built 15 June 2026 to 17 June 2026. Engines queried 18 June 2026 to 25 June 2026. Published 8 August 2026.
- Limits
- Not a random sample: every brand was advertising on Meta. One question and one run per engine, and Gemini was queried on a rate-limited free tier. These are ecommerce brands, not real estate firms.
-
Hostinger CDN Blocks GPTBot: an HTTP 429 Case Study
CiteVantage
Full method and limits for GPTBot case study- 64cache-busted requests
- 8user-agent identities
- 45 of 77pages readable to GPTBot on 21 September 2026
- Engines
- Requests sent as GPTBot, OAI-SearchBot, ChatGPT-User, Googlebot, ClaudeBot, PerplexityBot and an ordinary browser
- Dates
- Measured 14 September 2026, retested 21 September 2026.
- Limits
- One host and our own site. Many small agent sites run on shared hosting, and this page cannot say how many are affected.
-
The Realtor AI Visibility Index
Third party: Local Falcon
Read the source for Local Falcon, Realtor AI Visibility Index- 37,500queries
- 100largest US cities
- 5AI platforms
- Engines
- Google AI Overviews, Google AI Mode, Gemini, ChatGPT, Grok
- Dates
- Fieldwork August 2026. Published 18 August 2026.
- Limits
- A vendor measurement with its method published, covering "near me" agent queries only. We quote it with Local Falcon named and have not re-run it.
Data files and charts to download
All 17 files are served from this site, and the build checks that each one exists. Our data and charts are free to reuse under CC BY 4.0 with a link back, with no permission request and no embargo. Each row shows the file type, the size and, for images, the pixel dimensions.
Houston 500-query study
- Houston, row-level data All 2,000 answers: query, area, surface, every firm named and every domain cited. CSV, 532 KB Download
- Houston, named-firm appendix Every one of the 4,782 canonical names with mention count, areas and surfaces. JSON, 780 KB Download
- Houston, the query set The 500 queries as asked, with area, intent and shape. JSON, 79 KB Download
- Houston, hand-labelled sample 394 extracted names with the label each was given by hand. CSV, 45 KB Download
-
Houston, names by number of areas How many Houston areas each named agent or firm appears in. PNG, 215 KB 2400 x 1520 px Download -
Houston, own-website citations How often an answer cites the named firm's own domain, by tier. PNG, 191 KB 2400 x 1440 px Download
150-query study
- 150-query study, row-level data All 747 measurements: query, engine and every domain cited. CSV, 107 KB Download
- 150-query study, Google AI Overviews links Every source link Google AI Overviews returned: query, position, URL and domain. CSV, 99 KB Download
-
150-query study, reach How many of the five engines cited each domain. PNG, 291 KB 2400 x 1624 px Download -
150-query study, top source per engine The most-cited source for each of the five engines. PNG, 229 KB 2400 x 1264 px Download -
150-query study, engine overlap Shared sources between each pair of engines. PNG, 247 KB 2400 x 1744 px Download
181-brand study
- 181-brand study, row-level data All 181 rows, anonymised: category, platform and each engine's verdict. CSV, 13 KB Download
-
181-brand study, headline chart The 87% finding, with the source line and the licence inside the image. PNG, 297 KB 2400 x 1412 px Download -
181-brand study, absence by engine Absent, mentioned and cited for each of the four engines. PNG, 294 KB 2400 x 1504 px Download
How to cite the CiteVantage press kit
Cite our own figures as CiteVantage (2026) with the link to the study. For the Local Falcon figure, name Local Falcon and link its page, because the figure is theirs. Each card below gives a reference line and a one-sentence form, and each has its own Copy button.
-
01The 500-Query Houston Real Estate AI Visibility Study
Reference
CiteVantage (2026). The 500-Query Houston Real Estate AI Visibility Study. https://citevantage.com/research/houston-real-estate-ai-visibility-2026/
In a sentence
A 2,000-answer study by CiteVantage found that 94.0% of the Houston Association of Realtors' 50,000-plus members were never named by four AI surfaces (CiteVantage, 2026).
-
02The 150-Query Real Estate AI Search Study
Reference
CiteVantage (2026). The 150-Query Real Estate AI Search Study. https://citevantage.com/research/ai-search-real-estate-2026/
In a sentence
A 747-measurement study by CiteVantage found that 83% of the domains AI engines cite on real estate questions are cited by only one of five engines (CiteVantage, 2026).
-
03Which Domains AI Cites on Real Estate Questions
Reference
CiteVantage (2026). Which Domains AI Cites on Real Estate Questions. https://citevantage.com/research/domains-ai-cites-real-estate/
In a sentence
CiteVantage's open list of 1,297 cited domains found that 88% of AI citations on real estate questions went to ordinary company websites (CiteVantage, 2026).
-
04The 181-Brand AI Visibility Study
Reference
CiteVantage (2026). The 181-Brand AI Visibility Study. https://citevantage.com/research/181-brand-ai-visibility-study/
In a sentence
A 181-brand audit by CiteVantage found 87% of US ecommerce brands were absent from all four major AI engines (CiteVantage, 2026).
-
05Hostinger CDN Blocks GPTBot: an HTTP 429 Case Study
Reference
CiteVantage (2026). Hostinger CDN Blocks GPTBot: an HTTP 429 Case Study. https://citevantage.com/research/hostinger-cdn-blocks-gptbot/
In a sentence
CiteVantage measured one Hostinger-hosted site and found its CDN refused GPTBot on 8 of 8 cache-miss requests while other crawlers passed (CiteVantage, 2026).
-
06This page
Reference
CiteVantage (2026). CiteVantage press kit: sourced findings on how AI search names real estate agents. https://citevantage.com/press/
In a sentence
CiteVantage's press kit collects 7 sourced findings from its original AI search research for real estate (CiteVantage, 2026).
CiteVantage logo and brand files
The mark below is the file you download, shown unchanged on a light and a dark ground. The name is written CiteVantage, one word, and is set in ordinary text beside the mark. The mark comes as vector files, with the app icon also supplied as two PNG sizes for places that need a raster image.
- CiteVantage mark, SVG The mark with its constellation, as vector. SVG, 1 KB vector Download
- CiteVantage mark, dots variant, SVG The same V with three typing dots in place of the tail. SVG, 1 KB vector Download
- CiteVantage app icon, SVG The mark in white on a violet tile. SVG, 969 B vector Download
-
CiteVantage app icon, 512 px PNG The app icon as a raster file. PNG, 117 KB 512 x 512 px Download -
CiteVantage app icon, 192 px PNG The same icon at a smaller size. PNG, 14 KB 192 x 192 px Download
Colours read from the mark file #6C5CE7#4B32C3#8B7BFF
CiteVantage boilerplate and founder bio
Both texts reuse wording from the About page and the site footer, so they match what a reader finds when they click through. Copy either one as written, or lift a single sentence from it and credit it to CiteVantage. The legal facts, such as the Wyoming registration, come from the same record the About page uses.
About CiteVantage
CiteVantage is an AI SEO agency for real estate. It helps agents, teams and brokerages get named by ChatGPT, Perplexity, Gemini and Copilot, and it publishes original studies of what those engines say, with the row-level data under CC BY 4.0. CiteVantage LLC is registered in Wyoming and led by its founder, Abdul Subkhan.
Founder bio
Abdul Subkhan is the founder of CiteVantage. He taught himself web development, marketing and sales from 2020, left university at the end of 2023 to work in search full time, and went on to lead an SEO team as Senior SEO Team Lead at a Top Rated Plus Upwork agency. In 2026 he left that role to work on AI citation visibility, registered CiteVantage LLC in Wyoming and began publishing the research on this page.
Contact for comment and data questions
Abdul Subkhan is available for comment, data questions and methodology detail at abdul@citevantage.com. Email reaches Abdul directly rather than a shared inbox or a queue. For a data question, name the study and the figure you are checking, and say your deadline.