AI is transforming how buyers explore homes, potentially shifting influence from traditional real estate platforms to the technology guiding their searches. For years, sites like Zillow and Domain have dominated by controlling listings, buyer data, and search functionalities. However, a recent Harvard Business Review examination suggests this dominance may weaken as AI tools take over the discovery phase.
Sangeet Paul Choudary, an expert in AI and platform strategy, argues that consumers are increasingly delegating the discovery and comparison of products to AI assistants, potentially weakening the influence of the marketplaces they have relied on for decades. This trend is already evident in Australia, where REA Group embedded its realestate.com.au listings into ChatGPT in February 2026. Domain has also introduced AI-powered interactive floorplans and expanded virtual property viewing capabilities to simplify the buyer’s experience.
The critical issue for agents is no longer whether AI will participate in the process, but which entities will shape buyer decisions before they make initial contact. Choudary observes that AI can now access multiple property services through a single interface, allowing buyers to refine searches by factors such as commute times, school districts, or renovation potential. Instead of beginning with a specific portal, buyers may instruct AI to cross-reference listings across platforms, narrowing choices before visiting any website.
For portals, this development presents a significant challenge. Even the largest audience or most recognizable brand may no longer guarantee leadership if buyers turn to independent AI for decision-making. Choudary points to the U.S., where Zillow became the first major real estate app integrated into ChatGPT in late 2025. By March 2026, Realtor.com followed, embedding its listings into the same conversational platform. As a result, buyers can now evaluate homes based on highly personalized criteria that traditional filters cannot match.
Adobe’s findings show that traffic from generative AI sources to U.S. retail sites increased by 393% year-over-year in the first quarter of 2026. Visitors arriving through AI spent 32% more time on sites, viewed 10% more pages, and were 27% less likely to leave immediately. While these metrics suggest AI could drive highly engaged traffic, Choudary warns the relationship may favor AI over portals. The technology may already guide initial decisions, narrowing options before buyers interact with any platform.
For real estate professionals, this means engaging with buyers who have conducted extensive pre-research. Traditional marketing metrics—such as page views or inquiry counts—remain relevant, but their importance may diminish compared to post-enquiry interactions. The focus could shift to what occurs after the initial contact, as buyers arrive with well-defined preferences.
Agents possess data that portals lack: years of client conversations, inspection feedback, and transaction outcomes stored in agency CRMs. While portals track searches and saved listings, agents understand why buyers inspect one property but reject another, or why an offer succeeds or fails. If this data is structured and connected to AI tools, it could become a strategic advantage, particularly as assistants expand beyond research to manage alerts, inspections, and preliminary negotiations.
