Apartments.com Wants to Prove AI Search Is a Performance Story, Not a Tech Story
For most of the past two years, the multifamily industry has talked about artificial intelligence the way it once talked about virtual tours and chatbots, as an inevitability whose business case was still being worked out. Apartments.com is now trying to move that conversation from possibility to proof. In a new whitepaper shared with Propmodo, the CoStar-owned listing giant lays out the first sizable dataset from its sitewide Smart Search tool, which replaces the familiar grid of dropdown filters with a natural-language box where renters can type or speak something closer to how they actually think. “Pet-friendly two bedroom with in-unit washer and dryer in Denver” goes in, and a refined result set comes out. The pitch is that this is not a cosmetic upgrade but a change in how renter search works.
The headline number is volume. Apartments.com says renters completed 35 million Smart Search sessions in the two months after the January 2026 sitewide launch, a figure that suggests the feature is not sitting unused in a corner of the site. The more interesting claims, though, are about behavior. Compared with renters using traditional search over February and March, the company reports that Smart Search users spent 94 percent more time on listings, viewed 61 percent more listings per session, and saved 212 percent more properties. They also went deeper into the content that signals genuine evaluation, with 73 percent more photo views, 54 percent more rent calculator usage, and 26 percent more 3D tour views. Further down the funnel the lifts continue, with 56 percent higher conversion across all lead types, 60 percent on unit-specific leads, and 144 percent on tour requests, the action that most reliably marks a renter getting close to signing.
It is a clean narrative, and that cleanliness is precisely what invites scrutiny. Every one of those figures is a comparison between two self-selected groups, and the renters who choose to describe their needs in a conversational box may simply be more motivated or further along than those clicking through filters. Stuart Richens, vice president of product at Apartments.com, says the team looked hard at that question. The most persuasive evidence, he argues, came from renters who used both search modes within a single journey. “In those cases, we consistently saw engagement increase after they switched into Smart Search, suggesting the experience itself is driving additional activity,” he said, attributing the lift to reduced friction rather than reshuffled intent. “While there’s always some level of self-selection with a new product, the consistency of the engagement lift across different renter groups gives us a lot of confidence.” It is a reasonable answer, though not a controlled experiment, and the company has not published the underlying methodology that would let outsiders judge how much of the gap survives once motivation is held constant.
The deeper question is what any of this means for the people paying the bills. Owners and operators do not advertise on Apartments.com to generate saved properties or photo views. They want leases, and the whitepaper stops short of that line, measuring lead signals rather than signed contracts. Richens does not dodge the point. “At the end of the day, leases are absolutely the metric that matters most,” he said, framing the current results as early indicators consistent with the behavior of historically higher-converting renters. His larger argument leans on track record. Apartments.com has long marketed itself on lead quality over volume, and the company points to an analysis from Perq, a multifamily marketing software firm that looked at more than 1,000 apartment communities, showing its lead-to-lease conversion running about 2.5 times higher than Zillow’s over calendar 2025. That figure is sturdier than the usual vendor stat, both because it comes from a third party and because the conversions are tracked through property CRMs and management software rather than self-reported. It also carries a caveat the company did not volunteer. The Perq window runs January through December 2025, while sitewide Smart Search did not go live until January 2026, which means the 2.5x ratio measures Apartments.com’s standing advantage and predates almost everything Smart Search has done. The company’s own data shows Smart Search users are more likely to submit these high-value leads, but the link from those leads to signed leases is still an inference rather than a measured result. Time-to-lease data is the proof the industry will eventually want before it treats the engagement lift as a leasing lift.
Then there is the gap that is hardest to wave away. Alongside the live results, the whitepaper previews a Smart Search 2.0 beta with figures that dwarf the production numbers: 268 percent more time with listings, 500 percent more 3D tour views, and 602 percent more leads submitted. A 602 percent lift sitting next to a 56 percent one is the kind of juxtaposition that makes analysts nervous, because beta cohorts tend to be small, unrepresentative, and full of early adopters behaving unusually. Richens acknowledged as much. “Beta environments also tend to move very quickly as we’re testing, learning, and iterating in real time,” he said, adding that the company expects “some normalization as the product scales broadly.” That is the right expectation to set, and it argues for reading the beta numbers as a directional signal of what removing more friction can do rather than a forecast of what June’s wider rollout will deliver.
The most strategically loaded item on the 2.0 roadmap is not a search refinement at all. It is “upfront fee visibility,” which Apartments.com is positioning as a renter-experience improvement but which lands squarely in the middle of a national fight. The Federal Trade Commission opened an advance notice of proposed rulemaking on rental housing junk fees in late January, days after settling with Greystar over allegations that it advertised rents excluding mandatory fees. A bipartisan coalition of more than two dozen state attorneys general has since urged the FTC to require total-cost disclosure in listings, and states including Colorado have enacted their own all-in pricing laws. For a platform whose advertisers are the very operators charging those fees, surfacing them more prominently is not a neutral product decision. Richens frames it as aligned incentives rather than tension. “Renters want a clearer understanding of total costs upfront, and operators want better-informed prospects who know what to expect before they engage,” he said, pointing to early markets with transparent pricing where the company is “seeing strong leasing performance.” Whether operators staring at higher advertised totals agree that transparency is costless to them is a question the rollout will answer in practice, not in a whitepaper.
None of this is happening in a vacuum. Apartments.com generated more than a billion dollars in revenue in 2024 and accounts for roughly a third of parent CoStar Group’s business, which makes it a critical front in the portal wars with Zillow, whose own multifamily rentals operation has grown quickly. CoStar has leaned into a “your listing, your lead” model that markets properties rather than selling buyer leads to unaffiliated agents, and it has been aggressive elsewhere, reportedly picking up about 1,200 apartment properties after Zillow’s rentals partnership with Redfin drew an FTC suit. The AI search push is part of the same campaign. In February, CoStar launched a voice-enabled conversational experience on Homes.com built on Microsoft Foundry and said it intends to extend the capability across its portfolio, Apartments.com included. Smart Search, in that light, is both a product and a positioning exercise meant to argue that category-specific tools beat generic search innovation.
For owners and operators, the practical takeaway is less about Apartments.com’s metrics and more about what those metrics imply for listing hygiene. If renters are increasingly arriving through a conversational layer that rewards relevance, completeness, and clarity, then thin listings with stale unit data, weak media, and opaque pricing are the ones the system will quietly pass over. The whitepaper closes with a readiness checklist asking whether listings carry high-quality media, current unit-level detail, understandable pricing, and enabled tour scheduling. That is self-interested advice, since better listings make the platform look better too, and it also happens to be correct. The honest read on the data is that it is genuinely encouraging and not yet conclusive, a strong set of leading indicators still waiting on the lagging one that matters most. The June rollout, and the lease numbers that eventually follow it, will decide whether this is the performance story the company says it is.
The post Apartments.com Wants to Prove AI Search Is a Performance Story, Not a Tech Story appeared first on Propmodo.
