Highlights
- The dataset: The analysis behind this study covers over 180,000 AI source citations from prompts specific to the commercial real estate industry, with AI responses from four AI models run in 11 countries during Q2 2026.
- Cushman & Wakefield is the only brokerage in every model's top five: The single shared anchor across all four models.
- 25% of Grok's citations are from user-generated domains: Yelp and Reddit, roughly three times the share of any other model.
- 50% of citations in non-English prompts are local: A near-even split, ranging from 77% in Dutch down to 38% in Spanish.
- 72% of top sources differ between any two models: Any two of the four share only about 28% of their top-20 cited domains.
- 76% of Microsoft Copilot's citations are from commercial domains: Grok is the least commercial, at 60%.
Which sources should you target to get cited as a commercial real estate brand?
The few domains that surface everywhere are telling. Cushman & Wakefield appears in all four models' top five, and clutch.co, a B2B service-provider directory, is a top-five source for three of them. Beyond that shared spine, the models scatter: ChatGPT anchors on reddit.com, Grok on yelp.com, Microsoft Copilot on directory aggregators like The Manifest, and Google AI Overview on its own properties and European brokerage sites. A commercial real estate brand optimized for one AI model is absent from roughly 72% of the domains the other models prefer.
In practice that means four targets, not one. A strong commercial presence, your own site plus directories like Clutch, works across every model. Reddit presence reaches both ChatGPT and Grok, and Yelp reviews are a first-order signal for Grok. In non-English markets, local-domain content pays off most with Microsoft Copilot, the most consistently local model.
How have AI source rankings changed over time in the commercial 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 movement as relative rank shifts, not absolute declines.
The clearest story is the retreat of crowd-sourced and directory sources and the relative rise of the brokerages and Google's own surfaces. Yelp fell from #2 to #20, and its mobile counterpart m.yelp.com from #6 to #24; The Manifest (#4 → #15) and GoodFirms (#15 → #25) also slid. Meanwhile Google.com surged from #25 to #2, and brokerage sites climbed, Knight Frank (#24 → #9), Colliers Germany (#21 → #8), JLL (#16 → #7), and CBRE Canada (#22 → #13), while Cushman & Wakefield overtook Clutch for the #1 spot. Track the current picture in the live AI visibility rankings.
What type of content do AI models cite for commercial real estate?
Commercial domains, brokerage sites, listing platforms, service directories, are the backbone of every model's citations, but the dependence varies sharply. We classified every cited domain into content categories using a global domain registry.
The headline difference is Grok. One in four of its citations points to user-generated content, Yelp reviews, Reddit threads, social profiles, versus fewer than one in ten for every other model. Grok treats the open web's crowd-sourced opinion as a primary signal for "who's a good broker." The other three behave more like trade-press readers, leaning on brokerage and directory sites. Editorial coverage barely registers anywhere, between 1% and 6%, so a placement in a business journal moves the needle far less than a strong profile on a brokerage directory or a well-reviewed Yelp page.
Do AI models cite local-language content for commercial real estate?
Because a page's language cannot be confirmed directly at scale, we treat a citation as "local" when, in a non-English market, the domain ends in that market's country code (.de.fr.it.nl, and .es/.mx/.ar for the Spanish markets). Across all non-English prompts, 49.9% of citations are local by this measure, an almost even split with global, mostly English-language.com sources.
The Dutch and German markets are heavily localized, three of four Dutch citations stay on .nl domains like FundaInBusiness. Spanish is the most globalized: fewer than four in ten citations carry a local country code, partly because Spanish-language commercial real estate sits on many regional .com properties that this measure cannot credit. English-anchored content remains a viable path into every non-English market.
Which AI model relies most on local sources for commercial real estate?
Localization behavior splits by model as cleanly as content type does. Microsoft Copilot is the most local overall, it leads or ties the local-sourcing rate in the Spanish, French, Italian, and Dutch markets, reaching 92% in Dutch, though ChatGPT out-locals it in German (83% vs 68%). Grok is the most globalized, defaulting to .com properties even in non-English markets: just 32% local in Spanish and 39% in French.
For a broker operating across borders, the lesson is that visibility strategy must be model-aware and market-aware at once. A German brokerage that wants to show up in ChatGPT needs strong .de content (83% local there), while the same brand chasing Grok visibility benefits more from a strong global .com presence.
Should commercial real estate brokers optimize for each AI model separately?
Barely any 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 28.4%, meaning the typical pair of models agrees on fewer than three of every ten top sources. Google AI Overview and Grok are the most aligned at 42.9%; ChatGPT and Grok the least, at just 17.6%. No pair clears the halfway mark.
Even the most aligned pair, Google AI Overview and Grok at 42.9%, disagrees on more than half of its top sources.
How many sources does each AI model cite per answer?
Citation breadth is where the models diverge most starkly. Grok cites 29.9 sources per answer on average (95% CI 29.0–30.9), nearly five times ChatGPT's 6.1 (95% CI 5.9–6.2). Google AI Overview (8.4) and Microsoft Copilot (7.8) sit in between.
A brokerage 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 analysis draws from Temso's AI visibility monitoring platform, which tracks how brands appear in AI model responses across ChatGPT, Microsoft Copilot, Grok, and Google AI Overview. It covers commercial-real-estate-broker prompts, questions a buyer, tenant, or investor might ask an AI assistant about the best or leading commercial real estate brokers, collected between 10 March and 29 June 2026 across 11 country markets and six prompt languages (English, Spanish, German, Italian, Dutch, French).
It measures what AI models cite, not why a model picks a source, whether a citation drove a real-world decision, or the quality of the underlying page. These findings reflect citation behavior for commercial-real-estate prompts in the monitored markets in this window, and may not generalize to other prompt types or industries.
Methodology
How we measured this
We tracked four AI models responding to commercial-real-estate-broker prompts across 11 country markets and six languages over roughly 110 days. Each response was parsed to extract its 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 agreement was measured by comparing each model pair's top-20 most-cited domains, and temporal analysis split the observation period at its midpoint and compared domain rankings in each half.
The analysis covers over 180,000 cited sources (of 312,461 total source records) from 17,690 AI responses and 223,707 brand mentions. Sample sizes exceeded recommended thresholds for every reported finding. Localization is measured by a country-code-domain proxy (source records carry no detected content language), so those figures are directional and exclude English-language prompts, which sit overwhelmingly on generic domains, true local-language rates are likely somewhat higher than reported. Model coverage varied across the observation window, so we report rank shifts rather than absolute counts.
Frequently asked questions
Do different AI models cite the same sources for commercial real estate?
No, barely. Any two models share only about 28% of their top-20 cited domains on average, so roughly 72% 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 30 sources per answer versus roughly 6 for ChatGPT, a 4.9x gap. Google AI Overview (8.4) and Microsoft Copilot (7.8) fall in between.
What kind of websites do AI models cite for commercial real estate?
Overwhelmingly commercial sites, brokerage pages, listing platforms, and B2B directories, at 60–76% of citations depending on the model. Editorial coverage is minor (1–6%). The exception is Grok, where about a quarter of citations are user-generated content like Yelp and Reddit.
Does Grok really rely on reviews and Reddit?
Yes, dramatically more than its peers. 24.9% of Grok's citations are user-generated content, roughly three times the share of any other model. For Grok visibility, online reviews and community threads are a first-order signal, not an afterthought.
Do AI models cite local-language content outside English markets?
About half the time. Across non-English prompts, 49.9% of citations sit on a local-country domain. But it ranges from 77% in the Dutch market down to 38% in Spanish, and English-language .com content remains a path into every market.
Which model is best for reaching non-English commercial real estate audiences?
Microsoft Copilot relies on local sources most consistently, it leads or ties local-source rates in every market except German (where ChatGPT leads, 83% vs 68%), and reaches 92% in Dutch. Grok is the most likely to default to global English sources in non-English markets (as low as 32% local in Spanish).
Did the most-cited sources change over the period?
Yes. Relative to the first half, user-review sites lost ground fastest, Yelp fell from #2 to #20, while brokerage domains (Knight Frank, JLL, Colliers) and Google's own properties rose.
What should a commercial real estate brokerage do with this?
Treat AI visibility as four separate channels, not one. Maintain a strong commercial presence (your own site plus key directories like Clutch) for all models; cultivate reviews and community presence specifically for Grok; and invest in local-domain content for non-English markets, especially for Microsoft Copilot and ChatGPT, which reward it most.
