USA

The U.S. housing market in 2025 stayed in a slow-but-stable regime: demand was capped by affordability while supply remained structurally constrained. Existing-home sales totaled about 4.06 million (a ~30-year low) as elevated home prices and near-7% mortgage rates limited mobility; the national median existing-home price still rose 1.7% to $414,400. Rates eased late in the year (AP notes ~6.15% by year-end), supporting a pickup in December activity, and Freddie Mac's weekly 30-year fixed average was 6.06% as of Jan 15, 2026-down from 7.04% a year earlier. Supply remains the market's core driver: Freddie Mac estimates the U.S. is undersupplied by roughly 3.7 million housing units (as of 3Q 2024), helping keep prices resilient even when transaction volume is weak. New construction cooled but remained meaningful (October 2025 housing starts ~1.246 million SAAR), with builders leaning on incentives as financing costs stayed elevated. In rentals, a wave of multifamily deliveries kept advertised rent growth flat in 2025 (0% YoY), improving renter choice but tempering near-term rent upside for some assets. Consumer behavior is increasingly bifurcated: first-time buyers fell to a record-low 21% share and their median age hit 40, while repeat buyers leaned on accumulated equity (30% of repeat buyers paid all cash). Portals and brokers are competing on end-to-end conversion and AI-assisted discovery. Realtor.com launched AI-powered natural-language search, Redfin rolled out conversational search, and Zillow introduced an in-ChatGPT home-search experience. Risk and opportunity signals for 2025 centered on commission and listing-rule shifts (NAR settlement), rising fraud and content-rights disputes (FTC scam data; CoStar v. Zillow photos), and consolidation as platforms integrate financing, search, and closing.

The US residential portal market in 2026 is navigating two simultaneous upheavals: the structural repricing of agent compensation following the NAR settlement and the aggressive AI product push from all three major portals. Together, these forces are compressing the traditional buyer-agent workflow and creating new competitive space for portal platforms that can own more of the transaction journey.

The NAR settlement, effective August 2024, ended mandatory buyer-agent compensation offers in MLS listings and introduced mandatory written buyer-broker agreements before property tours. The practical effect has been a re-negotiation of buyer-agent fee norms, with commission rates trending downward in competitive markets. For portals, this creates both opportunity and risk. Opportunity: portals that can help buyers understand their compensation rights, compare agent performance, and manage their own search process more effectively will capture higher engagement from the growing self-directed buyer segment. Risk: if traditional agent relationships weaken and transaction volumes stay compressed, portal ad revenue from agent subscriptions — the primary revenue model for Realtor.com and Homes.com — will come under pressure.

Zillow's response to this environment is the most strategically coherent. Zillow Group's 'Super App' strategy — integrating search, agent matching, Zillow Home Loans, ShowingTime scheduling, and title services into a unified transaction platform — positions it to capture value from buyer intent regardless of how agent compensation evolves. The ShowingTime acquisition created a proprietary scheduling data layer that quantifies buyer intent at a granularity that no other portal matches. Zillow's AI tools, including the Listing Media Suite automated photography descriptions and the natural-language search improvements rolled out in 2024-2025, are reinforcing listing engagement metrics.

CoStar's Homes.com is the most significant new competitive entant in the residential portal segment. CoStar's strategy is differentiated: rather than aggregating MLS data, Homes.com focuses on agent-branded listing pages, positioning each listing as the agent's personal advertisement rather than a platform commodity. The theory is that agents will pay more for a platform that builds their individual brand. In the first two years of operation, Homes.com traffic has grown significantly but remains well below Zillow. The critical test for CoStar is whether the agent-brand model can generate subscription volumes at a scale that justifies the $1B+ investment in the US residential portal market.

For operators in the US market in 2026: Zillow remains the dominant consumer portal and the non-negotiable starting point for buyer exposure. Homes.com is worth monitoring as a potential long-term challenger with significant capital behind it. The AI search and listing quality features across both platforms reward agents who invest in high-quality photography, accurate descriptive data, and consistent price management. The rate environment remains the most important macro variable — affordability recovery will unlock meaningful volume, and the portals positioned to capture that volume surge will be those with the strongest listing quality moats built during the slow period.

AI's role in the US market is advancing faster than any other national market tracked by GPPI. Natural language search, automated listing description generation, AI-powered property valuations, and predictive buyer-agent matching are all live products across Zillow, Realtor.com, and Homes.com. The practical implication for agents is that listing quality expectations are rising: AI-enriched listings with consistent attributes, professional photography, and accurate pricing set a bar that manually-managed, sparse listings increasingly fail to meet. Agents and brokerages that invest in listing quality systems — whether through portal-native tools like Zillow Showcase or third-party photography and copy services — will see better engagement metrics and faster time-to-offer as AI-powered search surfaces the best-presented inventory first. Tracking the spread between Enhanced Markets penetration and standard market performance in Zillow's quarterly results will be the most reliable indicator of whether the super-app thesis is delivering on unit economics for operators evaluating US portal allocation.