If there is one thing Zillow Group’s chief executive wants everyone to understand about the future of real estate, it’s this: people still want to find their own homes. Jeremy Wacksman made the point plainly during the company’s quarterly earnings call, and his words carried an almost defiant warmth for a tech executive talking about artificial intelligence. Yes, he acknowledged, we’re entering an era where AI agents can cancel subscriptions, book dinner reservations, and handle a thousand small administrative annoyances that clutter modern life. But buying a home? That’s different. That’s personal. Wacksman argued that house hunting is not a tedious errand waiting to be automated—it’s an experience people genuinely enjoy, even when it’s exhausting. He put it beautifully: no one wants an AI agent to watch a movie or play a game for them, because the enjoyment is the point. Dreaming about a new kitchen, scrolling through photos of sun-drenched living rooms, imagining a backyard where kids might play—that’s not a chore to be outsourced. It’s the fun part of moving, the part people want to feel with their own eyes and hearts. In a world increasingly eager to hand responsibility to machines, Wacksman is staking his company’s future on a very human truth: a home is not just an asset, and choosing one is not just a transaction. It’s a milestone, and milestones are meant to be lived, not delegated.
That might sound like sentimental talk from a CEO, but it’s actually a crucial business argument. Zillow makes most of its money from the home-buying process—connecting hopeful buyers with real estate agents, generating leads, and running a mortgage lending operation that profits when people move from browsing to borrowing. Every one of those revenue streams depends on a simple behavior: humans visiting Zillow, clicking through listings, and reaching out to another human for help. If AI agents ever get so good that they can quietly search all available properties, compare them against a buyer’s preferences, and schedule viewings without ever opening a single tab or scrolling past a single photo, Zillow could find itself cut out of the equation entirely. That is the nightmare scenario for the company—a world where the dream of homeownership is reduced to a background process, and where the digital storefront that millions of people visit becomes obsolete. Wacksman is not naive about that risk; he’s confronting it head-on. But he’s also betting that the threat is smaller than some investors fear, because the search itself is part of the reward. There’s something deeply human about wandering through a house and mentally rearranging furniture, about lingering over a listing that feels like it could be the start of a new chapter. Automation might make the process more efficient, but efficiency is not what people are looking for when they decide to make one of the biggest purchases of their lives. They’re looking for a feeling. And feelings cannot be programmed, scheduled, or optimized.
Of course, the AI revolution is not waiting for anyone, and Wacksman knows Zillow cannot afford to bury its head in the sand. The last few months have shown just how quickly personal AI agents are becoming part of daily life—and just how messy the transition is proving to be. Meta’s AI agent, Muse, launched in September with the ability to book travel and make purchases on a person’s behalf, and within two weeks Amazon had slapped a block on it, saying the agent never identified itself and had no permission to shop on the site. Amazon is also locked in a legal battle with Perplexity over the shopping agent in its Comet browser; a federal appeals court recently overturned an injunction Amazon had won, leaving the dispute unresolved. These skirmishes are the early tremors of a much larger shake-up in how people use the internet. If AI assistants become the gatekeepers to commerce, then companies like Zillow lose their direct connection with customers. Wacksman, however, is not seeing the apocalypse yet. He reported that Zillow isn’t getting meaningful traffic from AI platforms so far—not even a trickle. Instead, about 80 percent of the company’s traffic still comes straight to the site, with people typing in the URL or tapping an app icon because they want to be there. That is a remarkably strong moat, at least for now. Zillow is also making sure it stays in the game rather than getting locked out: the company has integrated with OpenAI and Google’s Gemini, and is now working with Anthropic as well. The strategy seems to be simple—embrace the technology, build bridges to the big AI players, but keep insisting that the human spark remains the heart of the experience.
It would be a mistake, though, to think Zillow is simply crossing its fingers and hoping people stay attached to their screens. The company is investing heavily in its own AI tools, and those tools are designed to make the search feel more natural and more helpful without replacing the human element. Zillow’s AI mode—a conversational search feature that lets people ask questions the way they might ask a friend for advice—is now available to everyone who signs in. The results have been striking: more than 2.5 million people used it between March and September, asking 8.5 million questions. That is not a niche experiment; that is a crowd of curious buyers who want to do their own hunting but are happy to have a smarter guide along the way. The company is also using AI behind the scenes to empower the real estate agents who depend on its platform. Follow Up Boss, Zillow’s software for agents, is getting new AI features that can spot leads needing attention and recommend next steps—reminding an agent to call a couple who toured a house last weekend, or suggesting an email to a buyer who has been searching for weeks without finding the right place. Agents can approve or reject those suggestions, keeping the human firmly in control. And in a telling detail about the company’s internal culture, Wacksman noted that engineers at Zillow are shipping 60 percent more changes to production than they did in 2024, thanks to AI-assisted coding tools. In other words, the company is not just talking about AI; it is using AI to work faster, make better decisions, and give its users a more responsive experience. But every one of those innovations is in service of a larger vision: a world where technology helps people find homes, not where homes are found by machines.
The financial picture, however, is a reminder that even the most carefully crafted strategy cannot escape the realities of the housing market. Zillow’s third-quarter revenue rose 7 percent to $724 million, which sounds solid until you remember that the company had expected to bring in between $745 million and $760 million. The shortfall came down to something that no AI tool or clever feature can fix: a sharp rise in mortgage rates, which slowed home sales to a crawl and put a damper on the entire industry. When borrowing gets more expensive, fewer people move, fewer people buy, and fewer people need Zillow’s services—at least in the short term. Rates are not a glitch in a user interface or a mismatch in a search algorithm; they are the cold, hard math of affordability, and they affect everyone whether they search on their own or with a bot. To make matters worse, Zillow posted a $158 million net loss for the quarter, with the bulk of that stemming from a settlement in a shareholder lawsuit over Zillow Offers, the company’s ill-fated home-flipping business. That defunct venture was an attempt to move beyond listing and into buying and selling homes directly, and it ended in a costly disaster, a cautionary tale about the dangers of straying too far from what the company does best. The loss is a bitter pill, but it is also a kind of accounting closure—a final payment on a mistake that has haunted Zillow for years. Wacksman did not dwell on it, but the message was clear: the company is ready to move on, even as it navigates a market that remains stubbornly difficult for buyers and sellers alike.
In the end, Wacksman’s remarks amount to a bet on human nature, and it is a bet worth taking seriously. The rise of AI agents is real, and soon we may live in a world where machines handle more of our daily chores than we ever imagined possible. They will negotiate our bills, answer our emails, and maybe even choose our groceries. But Wacksman is betting that when it comes to the places we call home, we will always want to look with our own eyes and feel with our own hearts. That instinct may be irrational—a good AI agent could, in theory, do a more thorough search, spot better deals, and avoid the emotional mistakes that lead people to overpay for a house with a beautiful view and a leaky roof. But buying a home is not a purely rational act. It is about identity, belonging, family, and the quiet thrill of imagining a life somewhere new. Those are experiences no algorithm can share. So Zillow is walking a careful line: embracing AI as a tool to make the search easier and more intuitive, while insisting that the search itself remains a fundamentally human journey. It is a risky strategy in an industry being reshaped by automation, but it is also a deeply reassuring one. Because no matter how advanced our machines become, the dream of finding the perfect home—the one that feels like it was waiting for us—is still a dream we all deserve to live ourselves. And if Wacksman is right, we will never have to give that up.
