Antony Slumbers published a piece this morning that everyone in property should read. His argument, in one line: almost every AI initiative in real estate is aimed at doing today's work more efficiently, and efficiency you share with all your competitors isn't an advantage; it's a cost of staying in the game.
He calls it H1: the same work, done faster and cheaper. Quicker lease abstraction. Faster reporting. Cheaper admin. Useful, real money sure, but almost entirely undefendable, because the moment a gain comes from applying a widely available tool to a standard workflow, competition redistributes it. Clients take it as lower fees. Model providers take it as margin. You keep it for a quarter, maybe two.
Then he asks the harder question: what happens when everyone has the same efficiency? Without a bigger pie, the answer is a smaller industry. Fewer, better people, running the 2019 business at 2019 fee levels, minus the compression.
I read that and recognised the race most of proptech is currently running. I also recognised why we chose not to run it.
Look at the AI tools flooding into residential right now, and almost all of them are H1 by Slumbers' definition. They sit on top of a fragmented tech stack and try to stitch it together. Drafting the email faster, triaging the ticket faster, summarising the document faster. They automate the chaos; they don't remove it.
And they inherit the industry's real problem, which was never a shortage of automation. It's that most property data lives in systems built for accounting, not operations. Incomplete, inconsistently formatted, trusted by no one. AI is only as good as the data you feed it, and a model that fills gaps in incomplete data with confidence isn't intelligence. It's a liability. Confidently wrong is worse than no answer at all. Ask the professional services firms currently refunding government clients for hallucinated reports.
So a wave of tools will make everyone marginally faster at the same time, the savings will be competed away in fees, and the vendors will reprice sharply once the subsidised economics of the AI race normalise. That's not a strategy. That's a treadmill that speeds up.
Antony's framework says the durable opportunity is H2: things that were never impossible, merely uneconomic, and that you earn the right to H2 by fixing the foundations first.
That's been our thesis from the start. We didn't bolt AI onto a messy stack; we replaced the messy stack. A single rental operating system covering the whole lifecycle, with documented, consistent, compliant workflows as guardrails because you cannot automate a process that was never properly defined, and automating an inconsistent one just scales the inconsistency.
The unglamorous truth is that we already do the heavy lifting the AI tools are promising. Lead management, tenancy progression, communications, chasing. Automated, rules-based, reliable. Around 90% of applications complete without human intervention. Seven hours of admin saved per tenancy. We just never needed to call it AI.
What that discipline produces is the thing AI actually requires: clean, structured, enriched data from day one, on every resident, every tenancy, every property. A golden record, not four leaking buckets.
With that foundation, the genuinely new becomes economic. Not a faster ticket queue - the questions that actually drive net operating income, answered continuously. Property teams, portfolio managers and asset managers working from the same live data, making pricing, renewal and investment decisions quickly and with confidence, across any market, with any team. Yield managed as a discipline rather than reviewed in arrears. That's the capability that was always possible and never viable.
Apple didn't build the first MP3 player, the first smartphone or the first smartwatch. They built the best ones, at the right moment, on the right foundation. Being a fast follower on AI isn't a consolation prize; it's how you avoid paying for someone else's experiment. When the models are reliable enough, and the economics are sane, we plug in, and unlike operators running AI across fragmented point solutions, we'll have something worth feeding it.
Antony says the divergence will show up in revenue mix within five years: firms earning meaningful income from services that didn't exist in 2025, versus firms running the old business minus the fee compression. We agree. Better, faster, cheaper is table stakes. The foundation is the strategy.
We've been building it all along.
