To Win The Consideration Phase, Portals May Have To Switch Sides

People looking for a new place to live usually don't send a lead as they're searching; they do so after weighing up the pros and cons of listings they've been looking at during the consideration phase. It follows that real estate portals need to keep users on their site or app during the consideration phase to keep ownership of the lead.

Last week I argued that this, not search, is where the real battle between portals and LLMs will be fought. The question I left hanging was whether portals can actually win it.

Here portals face several challenges. Some are technical and overcomeable; some relate to the nebulous and elusive nature of internet user behaviour, and one is a fundamental clash of interests.

For 25 years, brands like Rightmove, Realestate.com.au and Idealista have been justifying the invoices they send their agent customers by bringing the most attention to their listings and maximising the outcome for sellers. Portals have never really had to work for the buyer.

But like people seeking legal advice, medical reassurance or relationship counselling, home buyers are increasingly being guided and empowered by LLMs that excel at consideration.

Portals can win the consideration phase, but only by doing something most of them would consider heresy... Switching sides.

Buyer vs Seller

Simon Baker regularly reminds listeners of the PPW Podcast that there are only two people who really matter in a housing transaction. He's stating the obvious, but he's right to do so.

Portals, agents, mortgage brokers and everyone else are middlemen who stand to lose very little on any given deal. The seller, and to a greater extent, the buyer, are involved in the most stressful transaction of their lives.

Yes, both want a smooth and successful transaction, but housing transactions are fundamentally adversarial; there are life savings, sleepless nights and tears at stake here. Sellers and buyers are opposing forces, sometimes at an awkward peace and sometimes in a proxy battle.

Broadly speaking, one side benefits from restricting or controlling information (days on market, price drops and, famously, climate data) while the other benefits from any disclosure about the listing and what's around it.

Nowhere are the differences between buyers and sellers so stark as in the USA, where the country's largest brokerage, Compass, is locked in a fierce legal battle with Zillow. At stake is who gets access to listing information when. Strip back all the nuance and bluster, and the case becomes a very simple question of whether buyer or seller interests should matter more.

Can portals change their loyalty?

If asked publicly about their loyalties, portal executives would probably channel their inner politician and say something slightly disingenuous like: "At the end of the day, the seller and the buyer have a common goal. We can serve both." But their loyalty has always been and remains with the seller.

Most big portals operate in sellers' markets where the demand has been greater than supply for a decade. As a result, they have acquiesced to seller interest, removing or never building data layers that would have made them better consideration engines. Meanwhile, some of the data held back from portals (crime stats, travel times, planning and zoning) has been made available elsewhere or can be interpreted to an acceptable level by Claude or ChatGPT.

They're now playing catch-up as they try to integrate the data their product department wanted to use all along while simultaneously trying to educate or at least not upset their agent customers.

Above: We classified more than 18,000 statements made by portal executives since 2019, drawn from our database of investor calls, podcast and PPW sessions transcripts, by whether they mention the buy side (buyers, renters, house hunters) or the sell side (sellers, vendors, agents, landlords). The sell side gets mentioned six times more often, although the gap has been closing since 2023.

Three ways out

That closing gap in the chart suggests portals know something has to change. The interesting question is which way they jump, and so far there are three answers on the table.

The first is the hedge. OLX's CompassGPT and Rightmove's ChatGPT app are decision tools built inside a third-party interface: the portal parks its consideration engine in the LLM's living room, hoping to be wherever the user is while keeping the lead for itself. But does the plugin have all the data about the user that the portal has? The saved searches, the viewing history and the mortgage decision in principle all live back on the portal, which is precisely the advantage Johan Svanstrom was describing when he told investors that consumers "go back and do much more to the experience" on Rightmove. And the loyalty hasn't changed either; a portal-built tool answering questions inside ChatGPT is still a portal-built tool.

The second answer is to stop hedging and side squarely with the buyer. In June, Huispedia, which claims to be the largest independent platform in the Netherlands with 3.8 million monthly visitors, launched Huispedia Intelligence, a "Collective AI" it describes as a digital shield for buyers. The company says that Dutch consumers bidding without professional guidance overpay by an average of €41,000 through emotional overbidding, and that seven in ten Dutch buyers have no buyer's agent at all.

"The most expensive problem in the housing market today is no longer in the information, it's in human behavior under pressure," said CEO Maxim Bours at launch.

These are launch-week claims from a challenger portal, complete with a "world's first" label and a declaration that traditional property sites are now obsolete, and none of it is independently verified. But the strategy matters more than the product here. A portal with real traffic has decided its market position is "the buyer's shield". In its own words, Huispedia is backing the side where the financial and emotional damage actually occurs.

The third and most radical answer keeps the portal loyal to the seller but changes the job. Ziuljenas Galeckas, who built a seller lead business and sold it to Baltic Classifieds Group, recently asked on LinkedIn what would happen if the portal owned the price rather than just the listing. "We price the most expensive purchase of our lives like we'd price a second-hand sofa," he wrote, arguing that portals sit on live demand signals that could work a price against the market the way airlines and hotels do. He's honest about the consequence: "The moment a portal owns the price, it owns the result." A portal paid for outcomes rather than listings is, as he puts it, "a different business, and a different balance of power."

Galeckas and I agree on the premise: the notice-board model, paid per listing, looks to be running out of road and portals can and perhaps should have a greater role in setting property prices. We differ on whose outcome the portal should own. His version puts it to work optimising the seller's result. I'd argue the bigger prize sits on the other side of the table, because the buyer's trust is the thing LLMs win by default if portals don't contest it.

The portal best placed to defect

If any portal could make the switch, it's Zillow. The MLS system means it doesn't depend on agent goodwill for inventory the way European portals do; the listings land on Zillow's results pages whether or not agents want them to. It has the market-leading consumer brand in the world's biggest real estate market. It has a history of moonshots (not all successful) and of showing buyers data that agents would rather they didn't see:

• The listings themselves back in 2006
• Its flagship AVM tool, the Zestimate
• In 2025, Zillow launched Offer Insights, a tool to suggest to buyers how much to bid for each listing

Yet Zillow is also the clearest illustration of how hard the switch is. It's fighting Compass for access to listings while simultaneously inching towards brokerage economics, where its revenue depends ever more directly on transaction sides and commission pools. The portal best equipped to side with buyers is being pulled towards the sell side by its own business model.

Whichever way Zillow goes, the underlying question already has a moral and political right answer. The biggest issue in most countries is that people can't afford to buy houses, not that people are getting lower prices than they wanted when they sell. I believe that being pro-buyer can also, counterintuitively, be the right answer for portal shareholder interests in the long term.

By unashamedly siding with buyers, portals not only increase their traffic and their trust, but they can also contribute greater affordability and liquidity to the housing markets they operate in. Portals that can combine internal and external data with a commitment to the truth about what a house should cost, how long it has actually been on the market, whether it's likely to be flooded and what a good offer looks like will enhance their brand, win the trust battle and keep ownership of the lead.