AI Discovery Black Book
Benchmarks, trends and findings from the fastest growing source of organic leads
We analyzed nearly 2,000 top competitors to find out what makes AI discovery engines recommend one apartment over another.
How We Got Here
The first industry report on what actually drives AI visibility in multifamily.
We started testing AI discoverability in mid-2025 with a simple goal: understand a growing source of lead traffic to our unit-level 3D touring platform. As we spent more time bringing our reporting solution to owners and operators, we kept hearing the same questions: Which ILS platforms matter for AI discovery? Does my website vendor make a difference? Why are my competitors showing up and I'm not?
This report is what came out of it — a snapshot across 416 assessments, 1,991 top competitors, and 326K data points. We'll update it quarterly as this landscape evolves.
Key Findings
Seven takeaways from our analysis of how LLMs and AI discovery engines find and recommend multifamily properties. Throughout this report, we use AI Discovery Score — the percentage of relevant AI discovery queries where a property shows up — as our core visibility metric.
Two data sources control AI visibility.
Zillow-sourced data powers 57% of ILS citations feeding LLMs. CoStar adds 28%. Two listing databases account for 85% of what AI discovery engines see.
Property websites still matter — especially at the community level.
While ILS dominates in aggregate (71% of citations), individual community websites carry significant weight for top performers. The highest-performing single community in our data had 249 mentions sourced from its own property website. Claude in particular over-indexes on property website content relative to other LLMs.
Properties underperform significantly when missing pricing and nearby POI/commute-focused data.
Top performers outscore assessed properties by 46% on price queries and 35% on commute. Well-structured price and commute content gives listings significantly more uplift in AI discovery than amenity descriptions do.
Top performers are using Entrata and Repli websites.
Top performing website vendors are 2x more likely to have top performing properties. Entrata (60.3%) and Repli (50%) powered websites perform at the top of their markets most frequently.
Gemini amplifies winners most aggressively.
The top-performer gap is largest on Gemini (+41%) and smallest on Claude (+13%). Each LLM creates different competitive dynamics.
AI discovery is earned at the community level, not the portfolio.
Some of the highest-performing PMCs have modest portfolios, while some of the largest barely show up. It's listing accuracy, content depth, and ILS coverage at each individual property — not sheer scale — that determines who LLMs recommend.
Redfin's data is Zillow's data.
Zillow paid Redfin $100M to exit multifamily ILS advertising (Feb 2025). Redfin now syndicates Zillow listings exclusively. That means listing on Zillow, Trulia, and Redfin puts you in front of three platforms — but only one data source.
Want to see how your properties stack up?
Get your property's AI discovery score against local competitors.
The Data Landscape
ILS platforms account for 70.7% of all LLM citations. Two parent companies control 85%. And breadth across data sources matters more than breadth across platforms.
When someone asks ChatGPT or Gemini to recommend an apartment, where does the answer actually come from? Mostly ILS platforms — but the platform names are misleading. Syndication deals mean many of these sites pull from the exact same listing database.
ILS Citation Share by Platform
Why Breadth Across Data Sources Wins
Does it matter how many ILS platforms you're on, or which ones? Both — but the kind of breadth matters. The goal isn't maximum ILS count. It's maximum data source coverage.
Market-Relative Source Breakdown
Top-quartile competitors aren't just on more platforms — they're on platforms from different data sources. The bottom quartile averages just 2 ILS sources, almost all from the same one.
So Should I List on Every Site That Syndicates from Zillow?
Breadth matters, but understand what you're actually getting. Listing on Zillow, Trulia, and Redfin gives you presence on three "different" platforms that all pull from the same database. You're not diversifying — you're increasing surface area for one source. The real play is making sure you're also on CoStar-network sites (Apartments.com, ApartmentFinder, Homes.com) and Yardi's RentCafe, which operate on independent listing databases.
It Varies by Market
The national averages mask real differences at the market level. Zillow dominates the ILS citation share in 18 of 20 states we analyzed — but the degree of concentration varies, and so does the competitive landscape.
Highlights show where a market deviates most from the national average — green for below-average concentration (more balanced), red for above-average (more concentrated).
Which owners and operators are winning in AI Discovery?
The PMCs with the strongest AI discovery score share common traits — and portfolio size isn't one of them.
We looked at which management companies are punching above their weight in AI discovery — and which aren't showing up despite their size.
Top PMCs by AI Discovery
For every property we assessed, we identified the top 5 competitors that AI discovery engines recommended most often. "Top-5 Appearances" counts how many times a PMC's properties showed up in those top-5 lists across all assessments. The higher the number, the more often AI is recommending that company's properties over yours.
What the Top Performers Have in Common
The standout PMCs share a few things: broad ILS distribution, strong property website content, and consistent listing data across platforms. Veris Residential, with a 35.3% mean AI discovery score and 16.2x overrepresentation, shows that a focused, well-managed portfolio can dramatically outperform larger competitors. Edward Rose & Sons at 24.3x overrepresentation reinforces the pattern.
Morgan Properties leads on raw volume (45 top-5 appearances), while Equity Residential and Nuveen hit the highest AI discovery scores among larger portfolios — scale and quality aren't mutually exclusive, but quality comes first.
What Low Performers Have in Common
The patterns at the bottom of the leaderboard are just as telling. Low-performing PMCs tend to share a few traits: thin ILS distribution (often only 2–3 platforms), outdated or template-heavy property websites with little unique content, and inconsistent listing data across platforms — different pricing, missing unit details, or stale availability. Many rely on a single ILS data source (usually Zillow-sourced) without coverage on CoStar or Yardi. The gap isn't about budget or brand — it's about execution consistency across every community in the portfolio. The common thread is a lack of visibility into what each community's listing data actually looks like across platforms — something unit-level auditing can surface quickly.
Your competitors are already in the data
See how your properties compare — personalized for your market.
What Content Actually Moves the Needle
We compared top-5 competitors against assessed properties across five content categories. The biggest visibility gaps aren't where most marketing budgets go.
The biggest gaps aren't in amenities or lifestyle copy — they're in pricing and commute content. Properties with well-structured data in those categories get significantly more AI discovery.
AI Discovery Score by Category: Top Competitors vs. Assessed Properties
Where Most Properties Fall Behind
The gaps above aren't outliers. Across all assessments, most properties underperform their top competitors in the same categories — and the shortfall is significant.
How Each LLM Plays It Differently
Not all models treat your listings the same way. The top-performer advantage ranges from +11% on ChatGPT to +41% on Gemini.
Each AI discovery engine weights listings differently. Here's how the gap between top competitors and assessed properties varies by platform — a larger gap means that LLM is more likely to show your competitor instead of you.
Claude Weighs Property Websites More Heavily
One notable pattern: Claude places significantly more weight on property website content than Gemini, ChatGPT, or Perplexity do. If your community has a strong, well-structured property website, Claude is disproportionately likely to surface it. The balance between property websites and ILS isn't fixed — it depends on which AI tool the renter uses.
Your Website Vendor Matters More Than You Think
The platform your property website runs on correlates with how often LLMs cite you.
We identified the website vendor for 382 top-performing competitor properties and compared them to mid-tier communities (ranked 6–100). Repli and Entrata are significantly overrepresented among top performers.
Vendor Distribution: Top Performers vs. Benchmark
Community Count by Vendor & Performance Tier
Top Performers: rank 1–5 | Mid-Tier: rank 6–50 | Invisible: rank 51–100. Excludes unidentified vendors.