The Real Estate Market in Australia Is Shifting Under the Portals

A lot of the recent debate around REA and Domain has been treated as a pricing story. One portal is framed as slightly more aggressive, the other as slightly more restrained, and the market is invited to read that as competition. But the more I look at what is happening here and overseas, the less I think pricing is the main story. It sits on top of something bigger. The more important shift is happening in discovery, in distribution, and in who gets to sit at the front of the property journey. REA is still vastly larger on audience than Domain, with Semrush estimating about 42.1 million January 2026 visits for realestate.com.au versus about 10.7 million for domain.com.au, while Domain’s own reporting says its residential platforms averaged 8.1 million Australians a month in Q4 2025.

That is part of why the Google real estate test in the US matters more than it first appears to. In select markets, Google has been testing property listings directly inside search, complete with images, prices and a “Request a tour” button. It still looks small, experimental and geographically limited. Even so, it changes the shape of the question. Once Google starts keeping more of the property journey inside its own ecosystem, this stops being only a fight between portals. It becomes a question of who controls discovery in the first place.

That matters in Australia because the portals are not as insulated from search as people sometimes assume. Public traffic analysis discussed by Stepps, drawing on Similarweb data, puts direct traffic to realestate.com.au closer to 53%, with organic search as its second-largest source, rather than the much higher direct-traffic assumptions that are sometimes quoted. Domain is also meaningfully exposed to search. If Google eventually rolls out a fuller property-search product in Australia, even a partial interception of that search traffic would put a very large slice of portal audience at risk before a buyer ever reaches REA or Domain. And if that happens, the obvious follow-on question is not just what it does to visits, but what it does to their ability to keep charging as if that discovery layer still belongs to them.

REA itself seems to understand where this is heading. On 11 February 2026 it launched a realestate.com.au app inside ChatGPT, allowing users to search live listings conversationally inside the AI interface. That does not look like the move of a business that thinks the portal homepage will remain the unquestioned front door forever. It looks more like a business trying to stay visible as the front door starts moving.

That shift is still early, and it is worth being measured about that. Similarweb’s February 2026 AI traffic pages suggest AI referrals are still small in percentage terms, about 75,200 visits to realestate.com.au and 45,500 to domain.com.au, with ChatGPT the dominant source for both. So this is not a case of AI having already replaced the portals. It has not. But it is no longer hypothetical either. There is now visible, measurable traffic flowing from AI tools into Australian property sites, and that matters because these things rarely stay still for long.

At the same time, discovery is starting to fragment in other ways. Aussie is not just pitching itself as another portal. Its Aussie for Agents offer is built around VPA-free listings across off-market, pre-market and on-market stock, while leaning on a claimed base of more than 5 million customers, 1,300 brokers and 220 stores. Cotality, meanwhile, says one in three agencies now publish new listings on their own website first, and Ray White AKG says 16 per cent of its sales now happen in the first week, often before a wider launch. That is not just noise around the edges. It suggests more of the market is trying to capture attention, enquiry and even sales before the most expensive portal phase begins.

The US is also starting to offer live examples of where this could lead. In February, Compass and Rocket announced a three-year alliance under which Compass “Coming Soon” listings would appear on Redfin immediately, with “Private Exclusive” properties to follow. Compass said the arrangement had the potential to bring more than 500,000 additional listings to market through Redfin. That matters because it shows a major search destination trying to win not simply by copying the standard open-listings model, but by getting privileged access to staged, semi-exclusive inventory. Seen from Australia, it is not hard to connect that back to the local rise in pre-market, off-market and website-first publishing.

Once you look at the market through that lens, the pricing discussion starts to look narrower than it first seemed. The current portal model asks the vendor to carry almost all of the risk upfront. A large amount is spent at the beginning of the campaign, regardless of whether the property sells quickly or sits there for weeks, and regardless of whether the eventual buyer comes through the first enquiry or much later. That is part of why more agencies are testing website-first and pre-market distribution. If you can reach the market earlier on your own terms, at much lower cost, the logic of handing over thousands upfront to a portal starts to look less inevitable than it once did.

What Google, AI discovery and Aussie really point to is not just more competition, but different pricing logic. Aussie is explicitly offering a VPA-free model. Google’s core advertising system is built around cost per click, where advertisers pay when someone actually engages. Carsales already shows that a classified marketplace can also charge on a pay-per-enquiry basis, with official dealer packages built around monthly fees plus enquiry charges. In other words, the alternatives now building around the portals are increasingly based on interaction, response or distribution leverage, not simply on asking the vendor to write a large cheque upfront and hope the campaign works.

REA does already have a Pay on Sale option, but even that remains a modified version of the old logic rather than a true rethink. It is tied to Campaign Flex, billed at the contracted Premiere rate plus 20 per cent, and still sits inside a structure designed mainly for agencies already doing meaningful volume. That is not the same thing as saying, “we will back ourselves and share risk properly.” If Domain really wanted to start a genuine price war, that is where the opportunity would sit. Not in raising prices a little more slowly than REA, but in saying to agents: give us the listing first, let us prove we can help sell it, and if it does not work you pay little or nothing. A two-week first-window model, a success-based fee, or a genuine pay-per-enquiry structure would be a far more meaningful competitive move than a lighter version of the same old portal invoice.

The pattern itself is not unfamiliar in Australia. Back in 2008, Sensis effectively conceded web search to Google, partnering to have Google power searches through the Sensis portal and place Yellow listings on Google Maps. Years later, Sensis was openly acknowledging that in metropolitan Australia the internet had become the main means by which people searched. Property portals are not a perfect replay of Yellow Pages, but the family resemblance is there. A directory-style business can look immovable right up until a better discovery layer starts doing the job more naturally.

That is why I think the real question now is not whether Domain can look slightly kinder than REA on next year’s invoice. It is whether the portals recognise that the market under them is starting to shift. Discovery is fragmenting. Inventory is becoming more strategic. Agencies are testing earlier distribution. AI is becoming a real, if still small, referral source. Google has shown what a different front door could look like. And once that combination starts to matter at scale, the old habit of charging more each year simply because the market has tolerated it may become much harder to sustain. The headline, to me, is not that one portal may now be a little less aggressive than the other. It is that the real estate market in Australia is shifting under the portals, and the most important competition now building may not come from one portal getting a bit better at the old game. It may come from the game itself changing.