GeoGenius Research · May 2026

We asked 4 AI models for the best realtors in 36 cities.
They agreed on almost nothing.

The largest published study of how ChatGPT, Grok, Google AI, and Claude cite sources for real-estate queries. 720 probes, 4,455 citations, 1,002 unique domains — and only 1.9% of sources are cited by all four engines.

36
metro markets
4
AI engines tested
720
standardized probes
1,002
unique source domains
Share the study:··
geo:Headline findings

What 720 probes revealed

9 patterns repeated cleanly across markets, engines, and regions.

01CROSSOVER

AI engines barely overlap

Only 19 of 1,002 cited domains (1.9%) are used by all 4 engines. The rest is engine-specific. Your "AI visibility plan" must address each engine separately.

02ENGINE PERSONALITY

Google Business Profile is ChatGPT’s #1 source

37.7% of ChatGPT citations are google.com — and 100% of those are Google Maps URLs (Business Profile pages), not Google search. Even more surprising: ChatGPT is the only engine that cites google.com. Grok, Google AI, and Claude all sit at 0%. If your GBP is incomplete, you’re invisible to ChatGPT.

03DENSITY EFFECT

The rural GBP tax

In rural markets ChatGPT’s GBP dependence jumps to 45% vs 33% in urban. When local website content thins, ChatGPT leans harder on Google Business Profile. The fix isn’t SEO — it’s a complete, consistent GBP with real-estate-agent positioning.

04SOURCE MAPS

Each engine has a clear #1

ChatGPT → Google. Grok → Zillow. Gemini → FastExpert. Claude → US News. Three of four are stable across all density tiers.

05CONTRARIAN

Realtor.com isn’t #1 anywhere

Despite being a top-10 source overall (#7), Realtor.com doesn’t lead for any individual AI engine. Brand recall ≠ AI citation.

06PRACTICAL

Only 3 universal aggregators

The ONLY aggregators every AI engine cites are Zillow, Redfin, and Compass. Not Yelp, not Realtor.com, not HomeLight, not US News.

07NAMED AGENTS

ChatGPT names specific agents — 288 of them

Every one of ChatGPT’s 359 Google citations is a Maps lookup for a named person or team, which let us recover exactly who it surfaced: 288 distinct agents and teams across 23 of the 36 markets. It does not answer in generalities — it picks named people, then checks them against a Business Profile before saying so.

08AGENT POWER

Agent-controlled surfaces win

Individual agent websites are the LARGEST single citation source (36.8%). Combined with regional + national brokerage sites, agent-or-broker-controlled surfaces are 48% of all AI citations — bigger than every aggregator combined. Your own site is the highest-leverage AI visibility surface, not your Zillow profile.

09BLIND SPOT

Where the industry publishes, AI isn’t reading

ActiveRain, LinkedIn, Medium, Inman, Nextdoor, X and TikTok were cited ZERO times out of 4,455. Reddit got 79. The engines cite consumer conversation and third-party evaluation — not agents publishing for other agents. And even that reaches only half the field: ChatGPT and Claude cited community sources 0 times.

geo:Interactive Map

Explore the data by market

Click any pin to see the top-cited sources for that market across all four AI engines.

Error: AuthFailure

A problem with your API key prevents the map from rendering correctly. Please make sure the value of the APIProvider.apiKey prop is correct. Check the error-message in the console for further details.

Showing 36 of 36 marketsPin size is uniform · color = density tier · click any pin for source detail
geo:The Source Map

Top 15 domains cited across all AI engines

Aggregated across all 720 probes. Google leads, but only barely — the long tail is enormous.

google.com
359 (8.1%)
zillow.com
281 (6.3%)
fastexpert.com
275 (6.2%)
realestate.usnews.com
173 (3.9%)
listwithclever.com
172 (3.9%)
yelp.com
159 (3.6%)
realtor.com
148 (3.3%)
realtrends.com
121 (2.7%)
homelight.com
117 (2.6%)
reddit.com
79 (1.8%)
redfin.com
53 (1.2%)
ownluxuryhomes.com
53 (1.2%)
remax.com
44 (1.0%)
effectiveagents.com
44 (1.0%)
expertise.com
44 (1.0%)

4,455 total citations across 1,002 unique source domains.

geo:Engine Personalities

Each AI engine has its own citation universe

Pick an engine to see its top 10 sources. Notice how little overlap there is between any two of them.

ChatGPT

180/180 ok · 5.3 cites/probe
google.com
359 (37.7%)
cityoffargo.com
14 (1.5%)
spokanecity.org
9 (0.9%)
compass.com
8 (0.8%)
charmeck.org
6 (0.6%)
thewaughgroup.com
5 (0.5%)
santaferealestateproperty.com
4 (0.4%)
amystockberger.com
4 (0.4%)
heathermurphygroup.com
4 (0.4%)
acropolisrealtygrouptampa.com
4 (0.4%)

Grok

179/180 ok · 8.7 cites/probe
zillow.com
148 (9.5%)
realestate.usnews.com
115 (7.4%)
yelp.com
104 (6.7%)
realtor.com
95 (6.1%)
fastexpert.com
93 (6.0%)
realtrends.com
52 (3.4%)
reddit.com
51 (3.3%)
listwithclever.com
46 (3.0%)
homelight.com
36 (2.3%)
redfin.com
31 (2.0%)

Google AI

175/180 ok · 8.0 cites/probe
fastexpert.com
159 (11.4%)
zillow.com
107 (7.7%)
listwithclever.com
86 (6.2%)
realtrends.com
68 (4.9%)
realtor.com
52 (3.7%)
homelight.com
44 (3.2%)
remax.com
38 (2.7%)
effectiveagents.com
29 (2.1%)
triple.com
28 (2.0%)
reddit.com
28 (2.0%)

Claude

88/180 ok · 6.3 cites/probe
realestate.usnews.com
58 (10.4%)
yelp.com
55 (9.9%)
listwithclever.com
40 (7.2%)
homelight.com
37 (6.7%)
expertise.com
30 (5.4%)
zillow.com
24 (4.3%)
fastexpert.com
23 (4.1%)
ownluxuryhomes.com
8 (1.4%)
hollywoodreporter.com
8 (1.4%)
homeguide.com
7 (1.3%)
geo:The Universal 19

Four platforms. Fifteen individual agents.

Out of 1,002 unique sources, only these 19 appear in every engine's citation pool. Here is the part nobody expects: only four are platforms or media. The other fifteen are individual agents' and teams' own websites.

alliebeth.com
bethdickerson.com
buyvtrealestate.com
chasemizell.com
chicagomag.com
compass.com
davidhatef.com
dorapuig.com
erinkrueger.com
heathermurphygroup.com
ivanandmike.com
joshflagg.com
justinlandisgroup.com
mcferrinrealestate.com
pursuitvt.com
redfin.com
schlichterteam.com
templetonrealestategroup.com
zillow.com

Critical insight: The only aggregator platforms cited by every engine are Zillow, Redfin, and Compass. Realtor.com, FastExpert, Yelp, HomeLight, US News — none of them. Optimize the universal three first; engine-specific platforms second.

The bigger insight: the remaining fifteen aren't platforms you can join — they're agents who earned their way onto this list on a domain they own. Every one of them dominates exactly one market; not a single agent site appeared in two. There is no national winner here. And the most-cited individual agent in the entire study isn't in New York or Los Angeles — they're in Savannah. Burlington, Vermont placed two firms on this list; Boise placed two. National reach didn't decide it. Local density of evidence did.

geo:Recovered from the citations

Who ChatGPT actually named

Every one of ChatGPT's 359 Google citations is a maps/search/ lookup for a specific person — meaning ChatGPT had already chosen an agent and went to Google Maps to verify them. The names survived inside the URLs.

288
agents & teams named
distinct entities
23
markets covered
of 36 probed
359
Maps verifications
100% of its Google cites

Named agents per market — where ChatGPT had the most to say

Scranton, PA
21
Sacramento, CA
19
Detroit, MI
18
Shreveport, LA
17
Des Moines, IA
16
Miami, FL
16
Pittsburgh, PA
16
Portland, OR
16
Phoenix, AZ
14
Hartford, CT
13
Indianapolis, IN
13
Seattle, WA
13

What this means for you: your Google Business Profile is not a general SEO nicety — it is the step where ChatGPT confirms you are a real, locatable business before it will say your name. If your Maps presence is thin or inconsistent, that verification fails quietly and you drop out of the answer.

Smaller markets produced the most named agents (108 in small/rural vs 87 in major urban), because in a thin market ChatGPT has to reach further down the list to assemble an answer.

Method note: this recovery is possible for ChatGPT only. Grok, Gemini and Claude cite the agent's own site or an aggregator, which doesn't carry the name in a parseable position. Entity names are taken verbatim from the citation URLs; brokerage suffixes were collapsed so "Jane Doe" and "Jane Doe – XYZ Realty" count once.

geo:Density Effect

Top sources by market density

The "rural GBP tax" effect: as markets get smaller, ChatGPT leans harder on Google Business Profile, and the other engines lean on a narrower set of authoritative national sources.

Major Urban

google.com
92 (6.1%)
zillow.com
84 (5.5%)
fastexpert.com
82 (5.4%)
realestate.usnews.com
76 (5.0%)
yelp.com
58 (3.8%)
realtrends.com
56 (3.7%)
listwithclever.com
50 (3.3%)
homelight.com
35 (2.3%)

Mid-Size / Suburban

google.com
114 (7.8%)
fastexpert.com
92 (6.3%)
zillow.com
89 (6.1%)
listwithclever.com
63 (4.3%)
realestate.usnews.com
56 (3.8%)
realtrends.com
52 (3.6%)
realtor.com
48 (3.3%)
yelp.com
44 (3.0%)

Small / Rural-Adjacent

google.com
153 (10.3%)
zillow.com
108 (7.3%)
fastexpert.com
101 (6.8%)
realtor.com
67 (4.5%)
listwithclever.com
59 (4.0%)
yelp.com
57 (3.9%)
homelight.com
46 (3.1%)
realestate.usnews.com
41 (2.8%)
geo:Source Categories

Where citations actually come from

We categorized all 1,002 cited domains into 11 buckets. The headline: individual agent websites are the biggest single category — bigger than every aggregator combined.

Green = agent-or-broker-controlled (you own it)Gray = platforms, aggregators, third-party

Across all 4,455 citations

All four AI engines combined.

Individual agent websites36.81% · 1,640
Aggregators (Zillow, Realtor.com, FastExpert, Redfin, HomeLight)24.71% · 1,101
Review/ranking sites (Yelp, Expertise, US News, Angi)11.72% · 522
Google Business Profile (Maps URLs)8.06% · 359
Regional brokerage or team sites7.72% · 344
Other non-aggregator (long-tail directories, niche services)4.26% · 190
National brokerage chains (Compass, KW, RE/MAX, Sotheby’s)3.55% · 158
Community / forum (Reddit, Quora)2.04% · 91
Social networks (LinkedIn, YouTube, IG)0.54% · 24
Government / local (cityof*, *.gov)0.52% · 23
News / media0.07% · 3

Agent-or-broker-controlled surfaces (highlighted in green): 48.08% of all citations. That includes individual agent websites (36.81%), regional brokerage/team sites (7.72%), and national brokerage chains (3.55%). Bigger than aggregators (24.71%).

By market density

As markets get smaller, individual agent sites decline and aggregators + GBP take over.

CategoryMajor urbanMid-size suburbanSmall rural
Individual agent websites45.0%34.5%30.7%
Aggregators20.6%24.6%29.0%
Review/ranking sites13.9%11.7%9.5%
Google Business Profile6.1%7.8%10.3%
Regional brokerage / team6.1%9.0%8.1%
National brokerage2.7%3.4%4.5%
Community / forum1.2%2.5%2.5%
Other non-aggregator3.0%6.0%3.8%

In urban markets, agent sites are 45% of citations. In rural markets that drops to 31% — and aggregators (+8.4 points), GBP (+4.2 points), and national brokerages (+1.8 points) absorb the difference. The fix for rural visibility isn’t just "build a site" — it’s a complete GBP plus a presence on the platforms AI defaults to.

Each engine has its own category preference

Same 4,455 citations, sliced four ways. The differences are dramatic.

Agent sites + GBP

ChatGPT

Almost never cites aggregators. Citations come from agent-owned web presence — direct sites, brokerage pages, and Google Business Profile.

Individual agent sites39.5%
Google Business Profile37.7%
Regional brokerage / team12.9%
Aggregators0.6%
Review/ranking0.0%
Most diversified

Grok

Spreads citations across categories more evenly than any other engine. Reddit + Quora drive forum share.

Individual agent sites33.5%
Aggregators29.6%
Review/ranking17.9%
Regional brokerage / team7.8%
Community / forum3.9%
Aggregator-heavy

Google AI

Leans hardest on aggregators (Zillow, Realtor.com, FastExpert) and surfaces national brokerage chains more than any other engine.

Individual agent sites38.0%
Aggregators35.8%
Review/ranking7.1%
National brokerage6.8%
Regional brokerage / team4.9%
Review-site oriented

Claude

Leans hardest on third-party review and ranking sites (Yelp, Expertise.com, US News). Less agent-direct than ChatGPT but more curated than Google AI.

Individual agent sites38.5%
Review/ranking26.1%
Aggregators24.6%
Regional brokerage / team5.8%
National brokerage1.3%

Green bars are agent-or-broker-controlled surfaces. Across every engine, those categories combined exceed any single non-agent category — which is the underlying signal behind finding #08.

Want to know where YOU rank across all four AI engines?

GeoGenius runs the same probes used in this study — on you. We show you exactly which of the 1,002 source domains cite you, which AI engines mention you by name, and what's missing.

Takes 30 seconds · No credit card · Real citation data

Methodology

Open data. Reproducible. We welcome scrutiny and follow-up research.

Sample: 36 U.S. markets stratified across three density tiers (12 each) and four regions (Northeast, South, Midwest, West).

Queries: 5 standardized natural-language queries per market — generic discovery, seller intent, buyer intent, luxury specialty, and local search.

Engines & models: ChatGPT (OpenAI Responses API, gpt-4o + web_search), Grok (xAI Responses API, grok-4-fast-non-reasoning + web_search), Google AI (Vertex AI, gemini-2.5-flash with Google Search grounding), Claude (Anthropic Messages API, claude-sonnet-4 + web_search).

Total volume: 720 probes (36 × 5 × 4), 4,455 citations, 1,002 unique source domains. Deduplicated to exactly one probe per (market, engine, query) tuple.

Success rates: ChatGPT 100% · Grok 99% · Google AI 97% · Claude 49%. Claude's lower success rate is an artifact of our own API provisioning during the run, not model behavior: 84 of its 92 failed probes returned an account credit-balance error and 7 more hit a rate limit. On the probes Claude did complete, it returned citations 87 of 88 times. We disclose this because it means Claude's citation pool (556) is smaller than the other engines' and its shares carry correspondingly wider error bars.

Aggregation: All citation URLs were normalized to their registrable domain (e.g., www.zillow.com/agents/...zillow.com) before aggregation.

Reproducibility: Every figure in the headline findings, engine, market, density, region and crossover sections is computed directly from the probe-level and citation-level exports — no figure is entered by hand. The named-agent counts are derived from ChatGPT's Maps citation URLs by a script committed alongside the data. The one exception is the 11-category source classification below: grouping 1,002 domains into categories requires judgment calls no script can make on its own, so treat those percentages as a considered editorial grouping of the same citation set rather than a purely mechanical result.

Data availability: Probe-level and citation-level CSV exports are available on request for academic, journalistic, and industry research. Email research@geogenius.ai.

Report generated May 14, 2026. We re-run this study quarterly to track shifts in AI retrieval behavior.

BESbswy