AI names real estate agents almost every time, and it names a lot of them. Claude with live web search named someone in 46 of our 50 real estate questions (92%), with 6.1 names per answer, second only to restaurants in our US study. But one name in three (32.7%) was an individual agent, not a brokerage. Counting firms only, 86% of answers named one, with 4.4 per answer.
The answers leaned on referral sites and agent finders: HomeLight was the most cited site, 45 times in 19 answers. In 10 answers most of the sources were spam-style pages on unrelated sites, the most of any industry we measured. And the lists were wide open: 91.9% of the 246 names appeared in only one answer.
What we asked
50 questions written the way US clients ask them. The 40 English questions were set in Chicago, Seattle, Austin, Tampa, Charlotte, San Antonio, Houston and Denver (5 each), covering 10 needs from first-time buyers and condos to property managers and commercial brokers: 16 compare, 16 specific, 4 price and 4 trust. The 10 Spanish questions were set in San Diego and Las Vegas (5 each), tagged "mixed". Six of them, word for word:
| Question | Language | Intent |
|---|---|---|
| Best real estate agent in Chicago for first-time home buyers? | English | compare |
| Who is the best listing agent to sell my house fast in Denver? | English | specific |
| Which real estate agents in Austin charge a lower commission? | English | price |
| How do I find an honest real estate agent in Charlotte? | English | trust |
| Mejor agente de bienes raíces que hable español en San Diego (best Spanish-speaking real estate agent in San Diego) | Spanish | mixed |
| Agente de bienes raíces de confianza en Las Vegas para vender mi casa (a trustworthy agent in Las Vegas to sell my house) | Spanish | mixed |
How often AI named an agent
Claude named at least one agent or company in 92% of real estate answers, a little under the study average of 93.8%, and hedged on 4 answers (8%) by naming no one.
| Question type | Questions | Named anyone | Names per answer |
|---|---|---|---|
| Compare | 16 | 100% | 7.3 |
| Specific | 16 | 100% | 6.3 |
| Price (lower commission) | 4 | 50% | 2.0 |
| Trust (honest agent) | 4 | 50% | 1.5 |
| Spanish (all types) | 10 | 100% | 5.8 |
The split is sharp. Every compare and specific question named someone, often many: the Houston apartment locator answer named 17 businesses and the Denver investor answer 14. But the two questions a seller cares about most, who charges less and who is honest, named someone only half the time. All 4 hedges were those questions: lower commission in Austin and San Antonio, honest agents in Chicago and Charlotte. Where commission answers did name someone, they named platforms and discount brokerages such as Redfin and Clever, not local agents. The Seattle answer said "The results don't name individual Seattle agents, so I can't say which specific person charges the least."
Real estate answers name people: 32.7% of all names were individual agents, second only to law firms (32.9%).
Where the answers came from
Claude cited 458 sources for the 50 answers, 9.2 per answer, from 135 websites, sorted into the same nine types as the main study.
| Type of source | Real estate (458) | All 10 industries (4,587) |
|---|---|---|
| Directories, agent finders and "best of" list sites | 40.4% | 37.3% |
| Brokerages' and agents' own websites and other sites | 31.4% | 29.0% |
| Off-topic hosts (spam-style pages on unrelated sites) | 20.3% | 15.9% |
| News and media | 3.9% | 11.0% |
| Review platforms | 2.4% | 3.2% |
| Press releases | 0.9% | 1.2% |
| Social media and forums | 0.7% | 1.6% |
| Government sites and Wikipedia | 0.0% | 1.0% |
Directories appeared in 44 of the 50 answers, and each kind of question had its own. The sites cited most:
| Site | Citations | Answers | What it is, and where it showed up |
|---|---|---|---|
| homelight.com | 45 | 19 | Agent referral site; most used for "sell my house fast" |
| biggerpockets.com | 34 | 6 | Investor network; its agent finder fed 5 of the 6 "agent for investors" answers |
| simplifyem.com | 30 | 5 | Lists of property managers by city, cited in all 4 property management answers |
| redfin.com | 28 | 12 | Brokerage and listings site; its condo listings fed the condo answers |
| remax.com | 14 | 5 | A brokerage's own pages, mostly in "most trusted brokerage" answers |
| listwithclever.com | 13 | 5 | Discount agent marketplace, almost all in lower commission answers |
| experience.com | 11 | 9 | Review platform with agent profiles |
The spam layer was heavy: 93 sources from 31 unrelated websites, in 28 answers. Described without naming the sites, which are often victims: a hotel software company's blog with posts like "The Top Tampa Real Estate Agents Every Homebuyer Should Know", cited 26 times; forum profiles on a fertility patient education site, cited 22 times, which the Chicago answer called "spammy"; and a university repository in Indonesia with "realtor in chicago il advice for selling your home faster" pages. In Charlotte, 41.3% of real estate sources were pages like these.
Review sites were talked about, not read: 8 answers mentioned Yelp and 10 Google reviews, and none cited either. In at least 33 answers Claude warned its sources were marketing, sponsored or low quality.
The agents and brokerages named most often
The 50 answers named 246 different agents and companies in 281 mentions, and 91.9% were named only once. The top five took 7.5% of mentions. Named most often in our answers:
| Name, as given | Answers (of 50) | Where |
|---|---|---|
| Redfin | 7 | Seattle (2), Chicago, Tampa, and Las Vegas and San Diego in Spanish (3) |
| Compass | 4 | Denver (2), Chicago, Charlotte |
| Chris Nevada (an individual agent) | 4 | Las Vegas only: 4 of the 5 Spanish Las Vegas answers |
| RE/MAX | 3 | Seattle, Austin, Houston: 3 of the 4 "most trusted brokerage" questions |
| AptAmigo | 3 | Austin, San Antonio, Houston: 3 of the 4 apartment locator questions |
Each repeat name owned one kind of question through its own pages. RE/MAX was cited for its own "Voted the Most Trusted Agents in Real Estate" claim, which the Seattle answer called "a company's own marketing claim". AptAmigo was cited through its own city guides for apartment hunting in Texas. Chris Nevada was found through his brokerage's blog and subdomains, and the answers noted the source was self-promotional. A Denver agent, Rick Janson, was named in 3 of the 5 Denver answers, each time with his own website among the sources.
Not a ranking. Counts from one assistant on one day, 21 October 2026, not our ranking or an endorsement. An agent missing here may be excellent, and the lists change from day to day.
City by city
Each city had 5 questions: notes, not rankings.
| City | Named anyone | Names per answer | Off-topic sources | Directory sources |
|---|---|---|---|---|
| Houston | 5 of 5 | 8.0 | 2.2% | 45.7% |
| Denver | 5 of 5 | 6.6 | 4.3% | 60.9% |
| Seattle | 5 of 5 | 5.0 | 4.3% | 45.7% |
| Tampa | 5 of 5 | 4.4 | 17.4% | 69.6% |
| Chicago | 4 of 5 | 7.8 | 34.0% | 40.4% |
| Charlotte | 4 of 5 | 3.5 | 41.3% | 21.7% |
| Austin and San Antonio | 4 of 5 each | 6.5 and 8.0 | 6.7% and 19.6% | 35.6% and 41.3% |
| Las Vegas and San Diego (Spanish) | 5 of 5 each | 5.6 and 6.0 | 37.8% and 35.6% | 22.2% and 20.0% |
- Tampa: the most directory-driven city, at 69.6%. The condo answer was built from Redfin listing pages and named just one brokerage, from its testimonials page.
- Denver: a same-name slip. The "sell my house fast" answer's sources included HomeLight's page for Denver, North Carolina and an agent directory page for Denver, Indiana; the answer itself stayed with Colorado.
What the Spanish questions showed
| English (40) | Spanish (10) | |
|---|---|---|
| Answers naming at least one agent or company | 90% | 100% |
| Names per answer, when any | 6.2 | 5.8 |
| Hedged | 10% | 0% |
| Sources with a "best" or "top" style title | 19.0% | 25.6% |
Spanish answers named someone every time, but from English pages. Only 3 of their 90 sources (3.3%) were Spanish pages, the lowest of our 10 industries (study average 23.8%), cited in 3 of the 10 answers. Off-topic hosts made up 36.7% of Spanish sources. And the names were mostly people: 53.4% were individual agents, so counting firms only, 80% of Spanish answers named a brokerage. The two "Spanish-speaking agent" answers named 9 agents and brokerages each, found mostly through English profile pages that say the agent is bilingual.
Three real answers
Which real estate agents in Austin charge a lower commission?
I couldn't find a definitive, independently verified list of Austin agents with lower commissions. The results are mostly from low-commission brokers and marketplaces, which have a stake in the answer.
Who is the best listing agent to sell my house fast in Denver?
The sources point to a few ways to narrow the field, and most of the rankings come from HomeLight, which is a referral company with its own interest in the outcome.
Mejor agente de bienes raíces que hable español en San Diego
No hay un ranking confiable que designe a un "mejor" agente. Los resultados son perfiles de agentes que se presentan a sí mismos como bilingües, así que sirven como punto de partida para buscar.
What an agent or brokerage should do
- Complete your HomeLight profile. It was cited in every city, and answers quoted its numbers: days on market, sale-to-list ratio, years licensed.
- Use the finder that matches your niche. Investor agents: the BiggerPockets agent finder. Property managers: the city lists of property managers. Commercial brokers: ClearlyRated, cited in all 4 commercial broker answers. Rentals: an apartment locator guide per city of your own.
- Publish one page per service and city. "Investor agent in Denver, Colorado", with your recent sales, neighborhoods and days on market. RE/MAX and AptAmigo were named for the questions their pages answered.
- Say how you charge. Commission questions hedged half the time and were filled by discount marketplaces. A plain page on your fee, or how you set it, gives the answer a local agent to name.
- Put your license number and state lookup on your page. Several answers told people to check the agent's license with the state (Illinois IDFPR, the Florida DBPR, the Nevada Real Estate Division, the California DRE). Make that check one click.
- Link every agent to the brokerage, both ways, so a mention of the person also counts for the firm. A third of the names were people.
- If you serve Spanish speakers, write Spanish pages. A Spanish profile and one Spanish page per main service, with the city and state. Only 3 of 90 Spanish sources were Spanish pages, so there is almost no competition.
- Measure the same questions monthly on ChatGPT, Gemini and Claude, and keep your Google and Zillow reviews healthy: answers send people there next.
What we'd do for you, and what we don't promise
We do this work for US agents and brokerages, remotely from Dubai, with traditional SEO included and ChatGPT and Gemini (Google AI) tracked on every plan. It starts with a free AI check: 10 real client questions and the agents AI names instead of you. The AI Visibility Audit (30 questions, 4 AI tools, 3 competitors, screenshots and a 90-day plan) is USD 549. Monthly plans are USD 1,249 (Foundation), USD 2,349 (Growth) and USD 4,399 (Authority), month to month, with no lock-in and no setup fee. Details at adjoltz.com/us.
In the first 30 days of any plan, if AI does not name you for at least 3 commercial questions on the list you approved, or you are unhappy for any other reason, we refund 100% of what you paid, in cash, within 7 working days (guarantee rules). We don't promise a top spot on any assistant, rankings, paid places on "best of" lists, leads or closings, or a US office. We do not have one.
Method and limits. 50 real estate questions from our 500-question US AI visibility study 2026, each sent once, word for word, to Claude with Anthropic's live web search tool on 21 October 2026, between 21:31 and 23:12 UTC. Named agents were kept, and firms-only figures remove them. Source types were coded by us with hand-checked lists, so the off-topic share is a floor, and simple name matching may make the "named once" share slightly high. One assistant, one day, one run: Pew Research Center found 44% of US adults use ChatGPT and 6% use Claude. We did not check that every named agent is licensed or active. Every question and its numbers: study data (CSV).