### **Reapit Reboot**
“Fresh investment” always gets written up as a victory lap. Reapit’s £20m-plus from Accel-KKR and its claim of £40m a year going into product and innovation certainly *looks* like that sort of story.
But it reads more like a stopwatch moment to me.
Reapit is not some lightweight bolt-on. It’s deep in the engine room of British agency: big, operationally messy businesses with legacy processes, compliance needs, and lots of people doing lots of small jobs. That installed base — from Savills and Leaders Romans Group down through strong regionals — is a moat. It’s also a hostage to fortune, because AI doesn’t politely enhance yesterday’s workflow. It rewrites it.
The old CRM world was built around human effort. Diandra in Esher sits down between viewings and spends half her life doing system admin: updating records, rekeying enquiries, chasing a chain, copying and pasting emails, scanning a list of applicants that *might* be relevant. The CRM was the place where work went to be processed by people.
Now imagine the CRM as the place where AI does that processing first.
A swarm of software agents can already do the unglamorous, high-volume tasks: monitor inbound leads, rank applicant matches, draft vendor updates, summarise calls, build valuation packs, spot fall-through risk, and tee up the tiny number of decisions that need a human being. The point isn’t to create a fancy chatbot. The point is to get Diandra out of the system and back in front of clients — and to stop agencies paying skilled staff to behave like an interface.
That’s what Reapit is really buying with £40m a year: time. Time to make its core platform AI-native, rather than an API-connected museum of workflows designed for 2014.
And it’s not just Reapit. Rightmove is publicly flagging around £18m into AI and adjacent initiatives in 2026, inside a broader investment phase to 2028. When the incumbents start putting proper numbers on the table, it’s a sign the threat has become measurable.
But the bigger issue isn’t the tech. It’s the economics.
CRMs in this sector have typically made money in ways that mirror the shape of the agency: branches, modules, and users. More negotiators, more admins, more managers — more seats, more revenue. It’s not a sinister model; it’s just how value used to be delivered. Humans operated the machine.
AI flips that. If one well-trained AI agent can do the CRM-heavy grind that used to be spread across 20, 30 or 50 people, that’s a phenomenal outcome for an agency: faster response times, fewer missed opportunities, less leakage, and more humans doing the persuasive, local, relationship-based work that still wins instructions.
It’s also awkward for a software provider if its best product shrinks the number of chargeable humans.
So yes, Reapit will spend on features. But it will also spend on something much less demo-friendly: working out what it charges for when the “user” becomes less central.
Does pricing move to transactions? To properties? To workflow outcomes? To AI actions? To volume of managed conversations? To instruction-generation performance? Some hybrid platform fee plus consumption? Or the most controversial option of all — taking a slice of the value created?
None of this is simple. Agents will pay for measurable wins: more instructions, better fee defence, smoother pipelines. They will not enjoy feeling metered for every automated task their own data made possible.
Reapit’s advantage is that it’s already embedded in the operational heart of the agent. That gives it workflow context, distribution, and a privileged position in the data. But it also makes the pricing question impossible to dodge. When you’re that central, whatever you do will be felt.
So the investment is sensible — necessary, even. Standing still would be reckless.
But if you’re an agent, don’t just watch the new AI buttons. Watch the commercial model behind them. Because the real question isn’t whether AI creates efficiency. It’s who gets to keep it.