Rightmove’s H1 2026 numbers look unremarkable, predictable even - right up until you notice the joins. Revenue up 7% to £225.8m, underlying operating profit up 2% to £148.2m, margin slid to 69%. Earnings per Share (EPS) up 5%, dividend up 3%. If you’re an agent, you’ll read that and think: “So we’ve paid more, again, and it’s still printing money.” If you’re a shareholder: “Fine - but where’s the growth?”
The split tells you what’s really going on.
Agency revenue is up 9% to £163.9m, ARPA up 8% to £1,636, and - here’s the figure that should make every estate agent owner wince - retention at 96%, their best first half in over a decade. Membership up 206 branches. Loud complaints, quiet renewals.
New Homes is the canary with a hard hat on. Revenue up a measly 2% to £38.2m, membership down 6% year-on-year, conditions among the toughest since the financial crisis, with “really not one” tailwind. Extraordinary thing for a CEO to say out loud, and it matters beyond developer advertising: don’t expect the builders to rescue your year.
Then there’s “Other” - commercial, rentals, mortgages - up 14% to £17.5m, rental services up 67% off the Enquiry Manager rollout. Bundle, distribute, repeat.
And this is the danger at the heart of Rightmove’s stealthy changes. It’s spreading, oozing across more surface areas of agent operations. Online Agent Valuation took 45,000 consumer submissions in six months - that’s the top of your funnel. Voice Intelligence transcribes and analyses your calls, so that’s your coaching, compliance and conversion analysis. Four million Enhanced Leads is your CRM’s job. Lead to Keys, with utilities and referencing bolted on, is your tenancy stack. Forty-six strategic AI deployments in the half, up from thirty-one. Each one is, somewhere, a product a supplier used to sell.
Why it works isn’t about product quality. Rightmove’s marginal cost of putting a feature in front of your eyeballs is roughly nothing - the login, the habit and the direct debit are already there. So it doesn’t have to win its category. It has to be adequate and already in the box. Then it does its actual job: nudging ARPA up and shortening the renewal conversation.
Guidance has been trimmed to 6-8% revenue growth from 8-10%, with profit growth held at 3-5% - helped by the real story of the day: a £350m buyback inside a plan to return over £400m by July 2027, part-funded by a new £200m revolving credit facility. So they’re bankrolling a land grab across agency operations and borrowing to hand back a fifth of a billion, off a 69% margin. Your subscription is paying for the thing that narrows your pool of suppliers, inexorably … to Rightmove.
The market gave the game away with a Friday session that swung 12% intraday, as traders digested “guidance cut”, then re-read “£350m buyback”. The buyback won, because buybacks always win when the growth narrative is doing a bit of a Sponsored Walk.
For agents, 96% retention alongside 8% ARPA growth is precisely what claimants in the Newman/CAT action will call dominance behaving badly rather than product-market fit - now costing Rightmove £4-7m in exceptional legal fees for this year, which is no longer background noise. But the claim could win on pricing and still hand nobody back ownership of their funnel. Litigation is a rebate. It isn’t a structure. Every function you hand over is one you can’t take with you - each a sensible efficiency until you add them up and notice you’re now renting relationships you used to own.
For suppliers, that margin compression says the AI programme is a P&L event now, not a slide deck. And anything which is a feature will eventually be bundled, because it’s cheap to add and it helps the renewal. A hundred point solutions, each competing hardest with the other ninety-nine, is a market eaten from underneath in sequence. Everyone can see it happening to somebody else’s category.
Which brings us back to what can be done. This is a moment in time, when AI is changing how consumers behave. It happened with the internet’s arrival, which slayed the print beast but replaced it with a much more dangerous Rightmove. Now, consumers are getting their intel from platforms that aren’t commercially incentivised to show specific featured properties. They don’t want to see the properties that Rightmove wants them to see - they want to see the ones that are perfect for them. They’re able to discover properties that are on agents’ websites via AI - so being on Rightmove could, with care, become less necessary. It takes thought, sure - but now’s the time to get thinking, searching for alternatives, because they’re out there and they work already.
Because one number makes the timing urgent. Rightmove reports over 85% of traffic direct and organic, and under 0.5% from Large Language Models (LLMs) like ChatGPT and Claude. That’s there to reassure investors and today it’s true. But the AI layer will be the front door eventually, and when a model answers “three-bed near a decent school” it builds that from whatever structured property data exists. If the best-organised corpus in the country sits behind Rightmove’s login, the moat deepens; if agents and suppliers hold portable, jointly accessible data of their own, the aggregation advantage is negotiable for the first time since 2000. That window is open while they’re mid-build. It isn’t open in 2030.
Rightmove is still operationally strong and the tools genuinely work - which is exactly what makes them so hard to resist. The real question isn’t whether the courts, the regulator or the competition eventually decide the thing agents mutter privately, that “retention” can sometimes mean “stuck”. It’s what you’d have to own for the word to mean something else. The technology exists, and it’s cheap set against what the industry already spends fighting the symptom; the only missing ingredient has ever been a reason urgent enough to move.
A 12% intraday swing and a £350m buyback funded by your subscription ought to be reason enough.