The Global Property Industry Is Entering Its AI Era

Online property search has been stable for twenty years: dominant portals, rigid filters and an experience that has barely changed since the early 2000s. That stability is eroding.  

Buyers now think and communicate in ways the old portals can’t support. AI finally gives the industry the tools to match those expectations. 

The rapid sequence of AI announcements from Homes.com, Rightmove and Redfin suggests the industry recognises this shift and knows it can’t postpone modernisation much longer. 

Late 2025 marked an unusual period of activity across several major portals. For example, Homes.com released an AI-based “Smart Search” tool that allowed users to describe what they wanted in everyday language.  

Shortly after, Rightmove announced that it would significantly increase its investment in AI over the next three years. The share price fell more than a quarter as investors questioned the impact on margins. 

More recently, Redfin, a publicly listed US brokerage and portal, introduced a conversational search product of its own. They also followed REA Group’s August investment in Jitty, signalling that even established global players are backing AI-driven entrants. 

Portals of this scale rarely make comparable product moves in such close succession. Their decisions pointed to a shared recognition happening globally that the habits of their users were changing faster than their products. 

Buyer behaviour has outpaced the portals serving them. Younger cohorts now expect to express intent in natural language rather than adapt themselves to a form. 

Their comfort with conversational interfaces is well established. Surveys from Deloitte and YouGov show that adults aged 18 to 34 are the heaviest users of chat-based AI tools and expect software to understand ordinary language and structure early steps automatically. 

Communication norms point in the same direction. Millennials avoid unnecessary calls, and Gen Z often ignores them entirely. Many buyers now default to async, text-based interaction. 

Buyers are speaking one language, and legacy portals are listening in another. 

People have always described homes in practical terms. How much light does a place get, is there a good school nearby, is the garden large or whether a renovation is realistic.  

Traditional portals can’t interpret these intentions. They reduce a buyer’s needs to a small set of filters, forcing people to do the hard work themselves: clicking into bad-fit listings, cross-checking details and manually deciding which homes truly fit what they are looking for. 

LLMs can extract far more detail about properties than traditional portals ever captured. Much of what makes a home appealing is already in the property photos, the floorplan or the listing description. These tools unlock that data. They can identify rooms, identify box rooms, estimate natural light, classify materials and key features, and judge the condition of a property. 

They also let buyers describe what they want in their own words. Instead of forcing people into rigid filters, the system can interpret a buyer’s description and match it to relevant attributes in the listings. 

These capabilities extend beyond the listing itself. New tools can pull in external data on transport, amenities and environmental factors, then combine it with listing-level information to give a fuller picture of what a home offers. 

Together, these advances give buyers much more power to express what they want and focus on the homes that genuinely match those needs. Portals that surface this depth and context are standing apart from those that do not. 

Inventory alone is no longer the primary differentiator. The advantage now lies in how well a portal can interpret each listing, understand each buyer’s needs and match the two.  

The pattern mirrors what has happened in other sectors when a product design aligns with contemporary behaviour. Users shift quickly once a more intuitive model appears. Buyers are not attached to legacy search patterns. They were accommodating them because no viable alternatives existed. Now that is changing, and the industry is being forced to change with it. 

Clearly, the large portals are not completely resistant to AI. But they are constrained by the systems and, more critically, the incentives that made them successful. Their platforms were built long before machine learning became foundational, which makes integrating new models slow and expensive.  

Years of high-margin stability have also encouraged optimisation rather than reinvention. Big changes are hard, because they could negatively impact revenue in the short-term, which they want to avoid at all costs. This gives newer, faster moving entrants like Jitty an opening.  

The assumptions that shaped the first generation of property portals no longer hold. Buyer expectations have changed, and AI can meet them in ways traditional systems cannot. The next phase of competition will be defined by intelligence, not just inventory.