The AI Threat: Why This Time it might be Different for Property Portals

In previous discussions, we examined how property portals became resilient “digital fortresses,” successfully repelling challengers for over a decade. Giants like Google and Facebook, as well as VC-backed and broker-backed challengers failed to dislodge them, underestimating the power of the network effect of buyers, sellers and agents, combined with the reassurance of verified content and efficient momentum of embedded, professional workflows. But as the late futurist Ray Amara noted, “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.”

With Artificial Intelligence, we may be moving past the initial hype into a phase of fundamental disruption. This time, will it be different? While past threats attacked the portals’ business model, AI attacks their very essence. It doesn’t try to outspend portals; it rewires the entire search and discovery funnel.

Our central thesis:

AI doesn’t kill marketplaces per se, but it severely punishes those that cling to legacy buyer and seller experiences.

SHIFT IN USER EXPECTATIONS

As users become more accustomed to AI in other areas of their lives, their demand for personalized, predictive, and instant service in property search will  grow exponentially

Agentic Workflows: Portals Risk Being Bypassed

One of the property portals’ most pressing threats is agentic AI; autonomous systems that reason, plan, and act independently without constant human input. In real estate, AI does so much more than just enhancing listings - the most common application today. Agentic AI can orchestrate entire workflows, from discovery to viewing to negotiations - potentially bypassing portals altogether.

In the traditional model, buyers go to a portal (usually directly or via SEO), browse listings, and generate leads for agents by clicks and calls. AI introduces the possibility of “agentic workflows”, autonomous systems that act on behalf of the buyer or seller.

Imagine telling an AI assistant: “Find me a two-bedroom apartment near Hyde Park, under £1 million, with high ceilings and natural light.” Instead of sending you to a portal, the assistant can scour multiple sources, including agent websites, private listings, the smaller portals, social media etc. and cross-check data, book viewings, and map out the steps of the transaction..

OpenAI's Operator and Deep Research products, released in early 2025, stand out as agentic AI tools. They allow consumers to specify complex property requirements ("three-bedroom homes near a train station with a large garden priced 8% less than comps") and have an AI agent search not just the big property portals, but also real estate agent websites, social platforms, and fragmented listing sources to deliver highly customized and unique results. This development is so significant that leading figures such as Mal McCallion, who was one of the original team at Zoopla, has publicly stated that property portals like Zoopla could face an existential threat from AI agents.

If portals fail to counter this functionality, existing buyers may never visit them again, while new buyers might never find them.  The fall out is that agents would stop paying portals to be promoted on their platforms.

Premium Placement and Legacy Monetization at Risk

Today, premium slots and featured listings are the monetization backbone of many portals. They work because the search process is imperfect: buyers more often click on promoted results.

Property portals have thrived on a simple but powerful loop: aggregate listings, attract buyers, and sell visibility and leads back to agents. Their monetization has been built on subscriptions, premium placements, and advertising, essentially monetizing inefficiency in the search process. 

AI undermines this. If the system can instantly identify the “best fit”  3- 5 properties for a buyer, there is little value in paying to appear at the top of a results page. The better the AI gets at matching, the less room there is for monetization based on skewing the search display .

Zillow has already experimented with Flex, a commission-sharing model that moves away from monthly subscriptions and premium slots, to claim a share of the agent commission instead. Expect more portals to re-engineer their monetization as AI will compresses the value of placement.

AI does not just promise to enhance search. It also promises to rewire the user relationship end to end, guiding buyers and sellers from discovery through to transaction, and potentially bypassing the monetization streams portals rely on.

MONETIZATION REQUIRES A RETHINK

The legacy model relies on selling user attention (leads) to agents. Agentic AI risks bypassing this entirely by owning the user relationship and guiding them through the entire transaction. 

AI-Native Platforms and Empowered Agencies

The real disruption may come less from incumbents adapting slowly, but from AI-native entrants. These platforms will not start as listing boards. They will start as agents for the consumer, personalized, conversational, and embedded in daily workflows.

Instead of a portal being the first stop, an AI assistant could become the gateway: surfacing listings, verifying details, arranging tours, even handling the transactional workflow. In that (currently theoretical) world, the portal risks being relegated to a back-end database rather than the front door of discovery.

Emerging residential agentic search players like HomeHapp are experimenting with experiences built around a chat-based interface that lets users describe their needs in natural language. The backend is purpose-built for real estate, using AI to interpret queries and personalize property discovery, rather than relying on predefined filters or bolt-on modules. Its platform leverages AI not just for search, but for personalized recommendations, deep neighborhood context, and real-time updates, surfacing relevant options traditional portals often overlook. 

AI-native commercial real estate platforms Skyline AI are redefining how real estate decisions are made by embedding artificial intelligence at the core of their operations. Rather than layering AI tools over legacy systems, Skyline AI was purpose-built to autonomously analyze and interpret vast datasets, drawing on thousands of property attributes and hundreds of disparate sources to identify opportunities, forecast asset values, and guide investment strategies in real time. For example, Skyline AI can instantly evaluate any multifamily building in the US by analyzing over 10,000 attributes such as rental income, occupancy trends, sales history, demographics, and local amenities. What would normally take weeks of human effort is delivered in minutes: risk-adjusted projections, optimal bid prices, and hidden value-creation opportunities that human analysts might overlook.

KEY AI VECTORS

AI excels in areas where most portals are weakest, moving beyond simple data aggregation, to intelligent data interpretation and user guidance

Why AI Punishes Legacy Models

AI excels in the very areas where portals are weakest: conversational discovery, predictive valuation, hyper-personalization, and automated matching. Instead of navigating clunky filters and scrolling through dozens of irrelevant listings, users can now receive precise recommendations tailored to their individual preferences.

This is more than a user-experience upgrade. It strikes at the heart of how portals make money.

• Premium placement could become less relevant. If AI delivers a perfect match in seconds, the concept of paying to sit at the top of a long results list loses all value. Portals have thrived by monetizing inefficiency; AI thrives by eliminating it.

• Subscription models lose appeal. Agents have accepted rising subscription prices because portals deliver reliable leads. If AI agents connect buyers and sellers directly, agents might start questioning the need to pay monthly fees.

• Commission-sharing becomes more viable. In the old world, tracking lead conversion was too messy for performance-based pricing to scale. AI changes that by potentially following the buyer journey from first query to closed sale, making it easier for whoever provided the lead to prove the end value they have delivered.

The danger is not that AI makes portals redundant overnight, but that AI agents and AI native platforms challenge the pillars of their existing monetization model. Those that cling to subscriptions and premium visibility risk being trapped in a shrinking business while new AI-native platforms aim to redefine what value add is.

This is why AI punishes legacy models. It does not destroy marketplaces, but it threatens to destabilize any business that fails to reinvent how it creates and captures value

What Comes Next

In next week’s post, I will explore how portals can fight back. Which strategies can help them embed AI before AI-native platforms take a swipe at their lunch? How can they evolve monetization beyond premium placement? And how can they leverage their data, brand, and agent relationships to remain indispensable?

Because while AI does not kill marketplaces, it will punish those that refuse to evolve.