2.5M renter chats: which operators are winning AI search? | The Unlock
We spent nine months watching ChatGPT, Claude, Gemini and Perplexity recommend apartments, across 2.5M renter conversations. Here's what we learned.
It's all in the third edition of the AI Discovery Black Book, built on 1,701 four-model assessments and 8,442 competing communities.
The owners winning AI search right now
Three owners who weren't in June's top 12 now hold the top three spots. The top 12 run anywhere from 24 to 329 properties.
| # | Owner | Move | Score |
|---|---|---|---|
| 1 | Pantzer Properties | New | 88.5 |
| 2 | Monarch Investment and Management Group | New | 82.1 |
| 3 | MG Properties | New | 81.3 |
Top owners and operators by Market Performance, Jun–Aug 2026: how visible a property is against the most visible property in its own market, scored 0 to 100.
Five of June's top 12 fell off the list, including one of the biggest names in multifamily. Most of the leaders who slid didn't get worse. The best property in each of their markets got better, which raised the bar. One top-12 owner's Austin property showed up in AI answers about as often as last winter and still watched its score drop from 100 to 40.
So what: Being big earns no bonus. Even inside one company, some properties show up in AI all the time and others barely show up at all. Check each property, not just the portfolio.
AI recommends communities, not brands
Ask AI for an apartment and 80% of answers name a specific community. Only 14% name a management company. Corporate homepages get under 0.3% of citations.
So what: A brand-level visibility score tracks the one page AI almost never cites. The full report breaks down where citations land, page by page.
One renter question is actually 11 different searches
Testing against a single prompt is incomplete at best. LLMs take one question and break it up, and you need to measure coverage against each one.
So what: One prompt shows you a fraction of what AI looks at. We traced a single renter message through every search, and found the word AI searched most in one test. None of the property pages we checked use it.
Showing price still wins. The edge is half what it was
Properties with clear rent and commute info get recommended more when renters ask about them. That advantage has shrunk since winter, from 45% to 20%. But 64% of properties still fall short on both, so it's still an easy win.
So what: Put real rents and commute times where AI can read them. The four-question check in the report shows where your property falls short, and what the weaker sites are missing once AI gets in.
72% of AI citations come from ILSs, and the share is growing
Up every month since May, to 73% in August, the highest in our three editions. Nearly 9 in 10 trace back to Apartments.com and Zillow.
So what: Most of what AI cites for apartments comes from listing sites, not property websites. And the Zillow vs. CoStar race flipped again: in June we said Zillow's lead had nearly collapsed. This summer it came back. The report shows which listings to prioritize now.
Three myths the data doesn't support
- "Reddit drives AI answers." Under 0.1% of AI citations come from Reddit, every month since December.
- "AI checks your reviews first." Under 1% of AI searches in August looked up reviews. The models searched price and commute.
- ...and the one about your website most teams get wrong. See the third myth →
Searches: ChatGPT, Gemini and Claude, Aug 2026.
So what: An AI plan built on any of these is aimed at the wrong target. The report lays out what the models search for instead.
Read the full third edition
Read the third edition of the AI Discovery Black Book.
Want to see where your own communities stand? Get a free AI visibility score for any property.