Highlights
- The dataset: The analysis behind this study covers over 160,000 AI source citations from prompts specific to the luxury real estate industry, with AI responses from four AI models run in 11 countries during Q2 2026.
- Google.com surged to #3 from nowhere: While Yelp fell out of the top 25 entirely between the two halves of the window.
- 18% of Grok's citations are from user-generated domains: Yelp, Facebook, and Instagram, roughly double any other model's share.
- 65% local sourcing in Dutch: The most localized market, with French lowest at 26%.
- 86% of top sources differ: Any two AI models share only about 14% of their top-20 cited luxury-real-estate domains.
- 96% of Microsoft Copilot's citations are from commercial domains: Agency and listing sites, making Microsoft Copilot the most commercial model by far.
Which sources should you target to get cited as a luxury real estate brand?
Start with the shared spine: luxuryestate.com is the only domain in all four models' top-20 lists, and engelvoelkers.com, sothebysrealty.com, idealista.com, and realestate.com.au each appear in three, so a strong listing-portal presence pays off across every model. Beyond that spine each model rewards a different footprint: ChatGPT anchors on reddit.com, Grok on yelp.com and social platforms, Microsoft Copilot on agency brand sites like Engel & Völkers, and Google AI Overview on its own properties.
The content mix sharpens the priorities. Microsoft Copilot sends 96.4% of its citations to commercial agency and listing sites, while Grok leans hardest on user-generated content, so Yelp, Facebook, and Instagram profiles matter most there. Editorial coverage barely registers for any model, so a strong agency profile on a listing portal moves the needle far more than a luxury-magazine placement, and a brand chasing Microsoft Copilot in a market like the Netherlands needs local-domain content while one chasing Grok benefits more from an international .com presence.
How have AI source rankings changed over time in the luxury real estate industry?
We split the ~110-day window at its midpoint (early May 2026) and compared each top domain's rank in the first half versus the second, reading relative rank shifts rather than absolute counts.
The directional signal is consistent: as the window progressed, AI models leaned relatively more on first-party agency domains and Google's own surfaces, and relatively less on social platforms and crowd-sourced review sites. Yelp fell out of the top 25 entirely (from #3), Facebook slid from #6 to #16, while Fine & Country, Savills, idealista.com, and JamesEdition all climbed and Engel & Völkers overtook Sotheby's Realty for #1.
What type of content do AI models cite for luxury real estate?
Commercial content, agency sites, listing portals, brokerage pages, is the backbone of every model's citations, but the dependence varies sharply. We classified each cited domain by category.
The two extremes are Microsoft Copilot and Grok. Microsoft Copilot is almost purely commercial: 96 of every 100 categorized citations point to an agency or listing site. Grok is the outlier in the other direction, one in six of its citations points to user-generated content like Yelp reviews, Facebook pages, and Instagram profiles, roughly double any other model. Editorial coverage barely registers anywhere (0.4%–6.5%), so a strong agency profile on a listing portal moves the needle far more than a luxury-magazine placement.
Do AI models cite local-language content for luxury real estate?
In a non-English market we treat a citation as "local" when the domain ends in that market's country code. Across all non-English prompts, 36.1% of citations are local by this measure, global, mostly English-language .com sources carry the majority.
The Dutch market is the most localized, two of every three citations stay on .nl domains. French and Spanish are the most globalized: barely a quarter of citations carry a local country code, partly because luxury real estate in those languages sits on international .com portals this measure cannot credit. Compared with most industries, luxury real estate skews global, these are international buyers shopping international portals.
Which AI model relies most on local sources for luxury real estate?
Localization behavior splits by model as cleanly as content type does. ChatGPT and Microsoft Copilot rely most on local sources, while Grok is the most globalized, defaulting to international .com portals even in non-English markets.
Google AI Overview returned almost no citations in non-English prompts in this window (German n=46, Italian n=31, none in French or Dutch), so those cells are indicative only. For an agency operating across borders, the lesson is that visibility strategy must be model-aware and market-aware at once: a Dutch agency chasing Microsoft Copilot needs strong .nl content (80% local there), while the same brand chasing Grok benefits more from a strong international .com presence.
Should luxury real estate agencies optimize for each AI model separately?
Almost no overlap, so yes. We built each model's 20 most-cited domains and measured how much every pair shares. The average across all six model pairs is 14.0%, the typical pair agrees on fewer than three of every twenty top sources.
The few domains that surface widely are telling. luxuryestate.com, a global luxury listing portal, is the only domain to appear in all four models' top-20, and engelvoelkers.com, sothebysrealty.com, idealista.com, and realestate.com.au each appear in three. Beyond that shared spine, the models scatter: ChatGPT anchors on reddit.com, Grok on yelp.com and social platforms, Microsoft Copilot on agency brands like Engel & Völkers, and Google AI Overview on its own properties. A luxury brand optimized for one AI model is absent from roughly 86% of the domains the others prefer.
How many sources does each AI model cite per answer?
Citation breadth is where the models diverge most starkly. Grok cites 28.8 sources per answer on average, more than five times ChatGPT's 5.6. Google AI Overview (9.2) and Microsoft Copilot (7.6) sit in between.
A luxury agency has roughly five times as many chances to be cited inside a single Grok answer as inside a ChatGPT answer, but those Grok slots are diluted across far more competing sources.
Context
This report draws on AI model responses to luxury-real-estate-agency prompts, questions a high-end buyer, seller, or investor might ask an AI assistant about the best or leading luxury real estate agencies, collected between 10 March and 29 June 2026 across 11 country markets and six prompt languages (English, Spanish, German, Italian, Dutch, French). For each response we recorded every source the model cited, the domain it belonged to, and the domain's content category.
It measures what AI models cite, not why a model picks a source or the quality of the underlying page. The localization measure relies on a domain's country code rather than verified language detection, so true local-language rates are likely somewhat higher than reported. Findings describe how these AI models behaved in this window, not a guarantee of future behavior.
Methodology
How we measured this
We tracked four AI models responding to luxury-real-estate-agency prompts in local languages across 11 country markets and six languages over roughly 110 days. Each response was parsed to extract source citations: the URLs, domains, and metadata referenced. Domain categories (commercial, user-generated, editorial, reference, institutional) were assigned from a global domain registry. Cross-model overlap was measured by comparing each pair of models' top-20 most-cited domains. Temporal analysis split the observation period at its midpoint (early May 2026) and compared domain rankings in each half. Model coverage varied across the observation window, so temporal movements are reported as relative ranks within each half.
The analysis covers over 160,000 cited sources (of 287,962 total source records) from 16,657 AI responses. Sample sizes exceeded recommended thresholds for every reported finding. Localization is measured by a country-code-domain proxy, the source records carry no detected content language, so those figures are directional, exclude English-language prompts, and are especially conservative for luxury real estate, where international .com portals carry much of the local-language content.
Frequently asked questions
Do different AI models cite the same sources for luxury real estate?
No, barely. Any two models share only about 14% of their top-20 cited domains on average, so roughly 86% of one model's top sources are missing from another's. There is no single set of "trusted" sources across AI.
Which AI model cites the most sources per answer?
Grok, by a wide margin, about 29 sources per answer versus roughly 6 for ChatGPT, a 5.1x gap. Google AI Overview (9.2) and Microsoft Copilot (7.6) fall in between.
What kind of websites do AI models cite for luxury real estate?
Overwhelmingly commercial sites, agency pages and listing portals, at 74–96% of citations depending on the model. Editorial coverage is minor (under 7%). The exception is Grok, where about one in six citations (17.5%) is user-generated content like Yelp, Facebook, and Instagram.
Is Microsoft Copilot really that commercial?
Yes, almost exclusively so. 96.4% of Microsoft Copilot's categorized citations point to commercial domains, agency and listing sites, with user-generated and editorial content together under 1.5%. For Microsoft Copilot visibility, a strong agency website and listing-portal presence is essentially the entire game.
Do AI models cite local-language content outside English markets?
About a third of the time. Across non-English prompts, 36.1% of citations sit on a local-country domain. It ranges from 65% in the Dutch market down to 26% in French, and global English .com portals remain a path into every market, luxury real estate is an unusually international category.
Which model is best for reaching non-English luxury audiences?
Microsoft Copilot and ChatGPT rely on local sources most consistently, Microsoft Copilot reaches 80% local sourcing in Dutch and 61% in German. Grok is the most likely to default to international English portals in non-English markets, as low as 22% local in Spanish.
Did the most-cited sources change over the period?
Yes. Relative to the first half, social and review sites lost ground fastest, Yelp fell out of the top 25 and Facebook from #6 to #16, while agency domains (Engel & Völkers, Fine & Country, Savills) and Google's own properties rose. We report rank shifts rather than absolute counts.
What should a luxury real estate agency do with this?
Treat AI visibility as four separate channels, not one. Maintain a strong, well-structured agency website and a presence on key international portals (luxuryestate.com, Sotheby's Realty, JamesEdition) for all models; cultivate reviews and social presence specifically for Grok; and invest in local-domain content for non-English markets, especially for Microsoft Copilot and ChatGPT, which reward it most.
