What does your agency actually own?
You own the brand. You own the relationships your team has built. You might own the building, or at least the lease. But the intelligence — the years of prospecting data, landlord contact trails, market analysis, campaign performance records, the institutional knowledge of your patch that distinguishes you from the agency down the road — where does that live? In a platform someone else controls, in a format someone else designed, on a server you could not access tomorrow if you stopped paying the invoice.
That is not ownership. That is tenancy. And the property industry has been signing this lease, year after year, because the alternative always looked more complicated than the renewal. The alternative is now simpler than it has ever been. And the regulatory, commercial, and technological landscape is shifting in ways that make the calculus very different from three years ago.
How have the three PropTech eras shaped agent data ownership?
The first era of property technology was about moving offline processes online. Window cards became portal listings. Paper files became digital records. The platforms that won were the ones that built audience fastest and charged for access to it. Agents had no leverage, so they paid.
The second era was about bundling. CRM, data, marketing tools, and eventually AI, assembled into integrated platforms priced to make switching feel irrational. The integration worked. The value proposition was genuine. But the architecture was deliberate: data generated inside the platform stayed inside the platform, in proprietary formats, with limited portability. The agent was a productive tenant, and the landlord was getting rich.
The third era is different in kind, not just in degree. It is defined by ownership — of data, of workflows, of the intelligence outputs that AI tools generate when applied to property data. It is being driven by three forces converging simultaneously.
- Regulatory pressure. The main provisions of the UK's Data (Use and Access) Act 2025 came into force on 5 February 2026, including the Smart Data framework that allows government to mandate open data-sharing schemes sector by sector. The CMA's cloud market investigation named egress fees a barrier to switching, and in March 2026 AWS and Microsoft committed to removing or reducing them. For vendors serving EU customers, the EU Data Act adds a further layer of obligation. The cost of running a closed ecosystem is rising, and it will not plateau.
- Market maturity. Agents who have been on the same CRM for five years are asking, for the first time in significant numbers, the questions they should have asked at sign-up. What do I own here? What happens to my data if I leave?
- AI commoditisation. As foundation AI models become cheaper and more capable, the moat of 'we have AI' is disappearing for every CRM vendor simultaneously. The sustainable advantage is shifting to the data — specifically to who controls it. The agent who owns their data and can apply any model to it is structurally better positioned than the one who can only use the model the vendor chose for them.
What makes open property intelligence practical today?
Open property intelligence is not a theoretical position. It is built on two specific infrastructure components that are available now.
The first is a UK property data API — a daily-refreshed, well-documented endpoint that delivers sales, lettings, landlord portfolios, EPC ratings, environmental risk, and demographic data programmatically. The data is no longer locked inside a platform's user interface. Any system, any workflow, and any AI tool can consume it. The agent chooses the tools. The agent owns the outputs.
The second is a property MCP server. The Model Context Protocol is an open standard — backed by Anthropic, supported by OpenAI — that allows AI models including ChatGPT and Claude to connect directly to live external data sources rather than relying on what they learned in training. When a property data provider operates an MCP-compatible server, connecting property data to ChatGPT takes minutes. The connection is direct. No vendor mediates it. Every output the agent generates using that connection is theirs to keep, use, export, and build on.
Together, these two components make it possible to build a property intelligence stack that compounds in value over time — where the data is current, the tools are chosen by the agent rather than the vendor, and the intelligence generated through prospecting accumulates in your business.
What distinguishes agents who build a proprietary data advantage?
There is a version of the next five years of estate agency where some agents build a genuinely proprietary intelligence advantage in their local market. They know their patch more completely than any competitor. They act on more current information, faster. Their AI tools are connected to live data, not to training sets from two years ago.
And there is another version, where a different set of agents continues paying for access to shared intelligence from a platform that owns the relationship more than they do.
The difference between those two agents is not the AI. It is who owns the data the AI reads from. Open platforms are how you build the moat. The tools to do it exist now.
📌 KEY TAKEAWAY
PropTech's third era is about ownership. Open, API-first platforms compound value in your business — not your vendor's. propalt.ai is built on the principle that the intelligence you generate from your market belongs to you — not to the platform you generated it through.
Sources
—EU Data Act — empowers users with data portability rights — European Commission
—Data (Use and Access) Act 2025 — legislation.gov.uk
—Propalt API — daily-updated, MCP-compatible UK property data. From £49/month — Propalt
