Key takeaways
- We tried to audit 36 real estate domains. Twenty-two, about 61%, blocked our crawler, which is the most prominent finding. Fourteen could be scanned, and 12 of those are brokerages.
- On those 12 brokerages, average AI visibility was 43.3% across 60 live, grounded buyer queries. The median was 40%, and eight firms were named in two of five queries or fewer. None scored 100%, and the best scored 80%.
- Commercial and residential split sharply. The five commercial firms averaged 56%, the seven residential brands 34%. On category queries such as "best real estate agents for selling a home," residential brands were named 3 times in 14.
- Problem-first questions, where the buyer describes a situation, named a brokerage in only 1 of 12 attempts. Comparison queries did best at 75%.
- Yelp was the top cited source, with 34 of 480 citations (7.1%). A brokerage's own website accounted for 3 of 480, 0.6%.
- This is a small sample, chosen by us and heavily national. Whether AI crawlers are also blocked on the sites that blocked ours is something we cannot say.
Why we audited real estate
Real estate looks like the most local business there is, so it is tempting to think AI will not matter. Buyers still tour homes in person and sign with an agent they met. But some buyers, sellers and occupiers put their question to an AI assistant first. Someone asks who to use for a first home, whether to renovate before selling, or who can lease them 40,000 square feet, and reads the few names that come back.
So we asked a plain question. When a buyer, seller or occupier describes what they need without naming a firm, which brokerages get named? And before we could answer it, we hit a bigger finding: most of the sites would not let us look.
22 of 36
real estate domains we tried blocked our audit crawler, about 61%. The other 14 were scanned, and 12 of those are brokerages
Deligatr Real Estate AI Visibility Audit, October 3, 2026
The blocked majority
We attempted 36 real estate domains across two runs. For 22 of them, our crawler was shown a bot-verification page or an error instead of the homepage, and the tool could not read enough of the site to run. The blocked list includes familiar names: Compass, Cushman & Wakefield, RE/MAX, Keller Williams, Century 21, Redfin, Sotheby's International Realty, Colliers, Newmark, eXp Realty, Corcoran, Howard Hanna and others.
Compare that with 5 of 21 sites that could not be scanned in our insurance audit, or 3 of 25 in our law firm audit. Real estate sites are evidently heavily protected, which makes sense: listing data and agent contact details are valuable to scrapers.
Here is what we cannot conclude. We cannot tell from the outside whether the protection that stopped our crawler also stops the crawlers AI assistants use. Many sites treat those differently. If they do block them, that content is not being read at all, and it is worth a technical check. If they do not, nothing is lost. We also note that several of the blocked brands, including Cushman & Wakefield, RE/MAX and Compass, were among the names AI mentioned most often in other firms' answers, so being blocked from our crawler clearly did not stop AI from naming them.
How the audit works
The tool reads each site's homepage, classifies it in the words a buyer would use, writes five buyer-intent queries (one local-style, two category, one problem-first, one comparison) and runs each as a live, grounded AI search with the company's own name left out. It is one run, five queries per company, on one day. AI answers change between runs, so read every score as a snapshot. The full method is in our audit documentation, and all twelve industry audits are on our AI visibility research page. This is a small sample with a large share of blocked sites, so the 43.3% is indicative only.
A company is either in the answer or it is not. Of the 14 sites scanned, we report 12 brokerages in the averages. Zillow is a portal, so it is reported on its own. One more, harcourts.com, turned out to be a Canadian maker of legal and academic robes, not the real estate network of the same name, so its results are excluded from every number below.
The distribution: only one brokerage reached 80%
Brokerages by AI visibility score
12 brokerages, each tested on 5 live grounded buyer queries. Number of firms at each score.
Deligatr Real Estate AI Visibility Audit, October 3, 2026
JLL reached 80%. Avison Young, CBRE and the network scanned as Berkshire Hathaway HomeServices followed at 60%. Eight firms, including Coldwell Banker, Marcus & Millichap, Douglas Elliman and Weichert, were named in two of five queries or fewer. Every firm was named at least once, unlike in our law firm and insurance audits, but this is a sample of large, widely known names. That alone should temper any comfort.
Commercial and residential are two different results
Average visibility, commercial vs residential
Five commercial firms and seven residential brands, split by the category the tool assigned. Indicative only.
Deligatr Real Estate AI Visibility Audit, October 3, 2026
The five commercial firms (JLL, Avison Young, CBRE, Marcus & Millichap and Savills US) averaged 56%. The seven residential brands averaged 34%. The gap is sharpest on category questions. Commercial firms were named in 8 of 10 category queries, such as "best commercial real estate brokers and advisors." Residential brands were named in 3 of 14, for queries like "best real estate agents for selling a home."
We can only speculate about why, and with 5 and 7 firms we would not push it. One plausible reason is that commercial queries in this audit were firm-level and often national, which favors large named brokers, while residential queries were agent-level, which AI tends to answer with portals, review sites and individual agents. That is a hypothesis, not a result.
What kind of question loses the most
How often a brokerage was named, by query type
Share of queries in which the firm was named. 12 brokerages, 60 queries.
Deligatr Real Estate AI Visibility Audit, October 3, 2026
Problem-first queries were the clearest gap. Questions like "I want to sell my house, who should I contact" named one brokerage in 12 attempts. That is the moment a seller has not chosen anyone yet, and it is where the shortlist forms.
Comparison queries did best, at 75%, because "top brokerages compared" favors recognizable brands. The takeaway is not that brand alone is enough. It is that when a person describes a situation instead of naming a brand, these firms were almost never part of the answer.
One caution on the local-style queries. They were generic, "real estate agent near me" without a city, so they say little about how a firm would perform for a real neighborhood query. That is the most important thing this audit does not measure.
Who got named instead, and where the answers came from
When one of our brokerages was not named, AI named someone else, usually a big name. Across the 60 answers, Cushman & Wakefield appeared as a named alternative in 16, CBRE in 14, and RE/MAX and Compass in 12 each. JLL, Redfin, Keller Williams and eXp Realty followed, with Zillow named in 6.
The sources behind the answers were concentrated. Across 480 citations on 225 domains, Yelp led with 34, then Wikipedia (20), HomeLight (20) and Zillow (17). U.S. News Real Estate, Clever Real Estate, Reonomy, FastExpert and LoopNet followed. Several of these sell leads or placement, and we cannot see how AI weighs that.
0.6%
of citations pointed to the brokerage's own website (3 of 480), measured by exact domain match
Deligatr Real Estate AI Visibility Audit, October 3, 2026
Exact domain matching undercounts a little. avisonyoung.us, which appears to be Avison Young's US site, was cited 10 times and does not match the scanned domain. Even counting it, a firm's own site was a small fraction of what AI relied on. For a business that spends heavily on its own website and listings pages, that is the number worth sitting with. AI built its answers from review sites, portals and reference pages.
Zillow is a portal, and it behaves like one
Zillow was named in all five of its queries, a score of 100%. We do not count it with the brokerages. Our queries for it were portal-style: homes for sale near me, what is my home worth, which listing website has the most homes. Its own site was 7 of its 40 citations, a far higher own-domain share than any brokerage. It shows what it looks like when a business is the default answer to its own category, and it is the competitor every brokerage's content strategy has to work around.
What to do with this, if you run a brokerage or team
- Check that crawlers can read your site. Test what your security layer shows a plain automated visitor, and what it does for the crawlers AI assistants use. Do not assume either is fine.
- Make brokerage, team and agent three clear entities. Consistent names, licenses, offices and links between them, so AI can place a person under a brand.
- Audit your presence where AI looks. Review sites, portals and directories showed up far more than your own pages did. Make sure listings are accurate and reviews are genuine.
- Publish neighborhood and market knowledge with dates and an author. Local market reports and plain answers about pricing, renovating and fees give AI something to quote.
- Treat commercial and residential differently. Commercial is relationship driven and often suits direct outreach to named occupiers and owners alongside visibility. Residential is local-search and review driven.
- Re-test regularly, and by city. AI answers change, and this audit did not test neighborhood queries.
If you want the inbound work done as a program, see our AEO, GEO and SEO service for real estate brokerages and teams. For commercial firms, signal-based outbound to named owners and occupiers runs alongside it in the same CRM; see our outbound guide.
Book a free consultation and we will run your own constructed questions against live assistants and show you who is named instead of your firm. Inbound is $2,500 a month on monthly billing.
Book a consultationEvery company we scanned
Company | Visibility | Appeared on | Top rival named instead |
|---|---|---|---|
| JLL | 80% | Category x2, Problem, Comparison | CBRE |
| Avison Young | 60% | Local, Category x2 | CBRE |
| CBRE | 60% | Local, Category x2 | Cushman & Wakefield |
| Berkshire Hathaway HomeServices (scanned as realliving.com) | 60% | Category x2, Comparison | Keller Williams Realty |
| Coldwell Banker | 40% | Local, Comparison | RE/MAX |
| Marcus & Millichap | 40% | Category, Comparison | CBRE |
| Realty ONE Group | 40% | Category, Comparison | RE/MAX |
| Savills US | 40% | Category, Comparison | CBRE |
| Windermere Real Estate | 40% | Local, Comparison | RE/MAX |
| Douglas Elliman | 20% | Comparison | Sotheby's International Realty |
| NextHome | 20% | Local | Compass |
| Weichert Realtors | 20% | Comparison | Redfin |
| Zillow (portal, not a brokerage, not in averages) | 100% | Local, Category x2, Problem, Comparison | Redfin |
| harcourts.com (a robe maker, not the real estate network, excluded) | Excluded | Not counted | Not counted |
Not scanned (bot-verification page or error, 22 domains): allentate.com, berkshirehathawayhs.com, bhgre.com, century21.com, colliers.com, compass.com, corcoran.com, cushmanwakefield.com, era.com, exprealty.com, fathomrealty.com, halstead.com, houlihanlawrence.com, howardhanna.com, johnlscott.com, kw.com, longandfoster.com, nmrk.com, realogy.com, redfin.com, remax.com and sothebysrealty.com.
Check your own number
Every company above was run through the same tool below. It takes about ninety seconds, and your company's name is never in the query.
Do AI assistants recommend you?
Five buyer-style queries for your category, run through live AI search, show who gets named instead of you. Takes about ninety seconds.
Free. The full report arrives by email so the tab can close.
Methodology and honest limitations
What we did. On October 3, 2026 we ran the live audit tool that powers our free checker against 36 real estate domains in two runs, five queries per company, visibility measurement only. The step-by-step method is in our audit documentation. A company that disputes its result can ask us to re-run it.
The sample. We chose well-known US-facing brokerages, franchise networks and commercial firms, plus one portal. It is not random and not a census, and it is dominated by large national names. Smaller brokerages, independents and teams were not tested, and they may do better or worse than this sample did. Twenty-two of 36 domains could not be scanned, so the 14 we could scan may differ from the 22 we could not, and 12 firms is a small base. Averages are indicative only.
What it does not cover. Only the homepage was read, so the firm's classification depends on how the tool read it. The queries were generic and brokerage-level, not tied to a city or neighborhood, and not run for individual agents. Five queries per firm is a small sample of real buyer language. Results come from one run on one day, and AI answers change over time. For the commercial versus residential split we used the category the tool assigned, with 5 and 7 firms, so treat it as a pattern worth checking, not a finding. We did not test whether any firm pays for directory or portal placement, and we cannot say whether AI crawlers are also blocked on the sites that blocked ours.
On our own data. One scanned domain, harcourts.com, was the wrong company: a Canadian robe maker rather than the real estate network of the same name. We excluded it. Including it would have shifted several averages.
On advertising rules. Nothing here is legal advice. Real estate advertising, fair housing, licensing and messaging rules vary by state and locality.
FAQ
How visible are real estate brokerages in AI answers?
Moderately, with a very small sample. Across the 12 brokerages we could scan, average visibility was 43.3% over 60 buyer queries, and the median was 40%. Eight of the 12 were named in two of five queries or fewer. With only 12 firms, treat the averages as indicative, not definitive.
Why did so many real estate sites fail to scan?
We tried 36 domains and 22 of them, about 61%, including Compass, RE/MAX, Keller Williams and Redfin, showed our audit crawler a bot-verification page or an error. Listing data and agent contact details are valuable to scrapers, so heavy protection is no surprise. We cannot tell from outside whether it also stops the crawlers AI assistants use.
Is Zillow in the results?
Yes, reported separately. Zillow scored 100%, but it is a listing portal, not a brokerage, and our queries for it were portal questions such as what is my home worth. We left it out of the brokerage averages.
Who did AI name instead of the audited firms?
Mostly large national names. Cushman & Wakefield was named as an alternative in 16 of 60 answers, CBRE in 14, and RE/MAX and Compass in 12 each. Zillow and Redfin also recurred. Several of those names were among the sites that blocked our crawler.
Where does AI get its answers about brokerages?
Mostly from third parties. Across 480 citations, Yelp was the most cited domain, followed by Wikipedia, HomeLight and Zillow. A brokerage's own domain accounted for 3 of 480 citations, 0.6%, by exact domain match.
Does this audit tell me how an individual agent or a local team would do?
No. We audited brokerage-level sites, mostly national names, and our queries were not tied to a specific city or neighborhood. Agent-level and neighborhood-level visibility, which is how most real buyers search, was not measured.
