Three years ago “AI buyer matching” was a slide in a vendor advice pack. In May 2026 it is something agents are quietly running against their CRM every morning before they call out. Worth a look at where the tools actually are.

The category breaks into three rough buckets.

The first is enrichment of the existing CRM. The tool sits on top of the agent’s CRM and turns buyer enquiry data, open inspection check-in data, and the comparable interaction history into a richer buyer profile. The agent gets a list of cooler-than-average matches between their current listings and active buyers in their book. The value is real but modest. The tool is genuinely surfacing buyers the agent had forgotten about and re-prioritising the call list. The accuracy of the match is far from perfect — there is plenty of guesswork on price tolerance and on geography — but the agent gets a list that is good enough to drive a phone call.

The second is portal-data buyer-matching. The tool ingests the agent’s listings and matches them to the saved-search and inspection-tracking activity from the major portals. The match quality is better than the CRM-only category because the portals see the buyer’s actual search behaviour. The limitations are that the portal data is only partial — buyers run their search in many places — and that the buyer’s identifier is often a non-personalised match by saved search, not by named person. So the tool tells the agent “your three-bedroom in Glebe at this price range has 47 active saved searches in the last two weeks” rather than “Jane Smith should be your first call”. Useful for vendor reporting, less useful for direct buyer call lists.

The third is generative-AI buyer-correspondence. The tool drafts the personalised email or message to a matched buyer about a specific listing. The first generation of these tools sounded like marketing emails. The current generation is better — short, human-sounding, with a clear ask. The agents using these tools well are reviewing every draft before it goes, and they are using them mostly to compress the time it takes to write 20–30 personalised match emails when a new listing goes live. The lift is in the time, not in the conversion rate so far.

A few observations from the data we have:

The agencies running these tools well have spent more on training the agents than on the licence fee. The tool that the agent doesn’t trust gets ignored. The tool the agent has been shown how to use confidently — and shown where it is wrong — gets used.

The matching accuracy is heavily skewed by data quality. The agent who keeps the CRM tidy gets dramatically better matches than the agent whose CRM is a graveyard of old enquiries. PropTech tooling does not fix bad data discipline.

The agents using these tools to compete on speed are winning more than the ones using them to compete on volume. The first call to a strongly-matched active buyer when a new listing goes live is materially more valuable than the eighteenth match-email to a buyer who has been on the database for two years. The tool’s real value is reducing the time-to-first-call after a listing launches.

A note on what the tools cannot do. None of the current generation reliably reads a buyer’s “soft” pivot — the moment the buyer-with-a-three-bedroom-brief actually starts looking at four-bedrooms because the second child is on the way. The agent who knows the buyer’s life from a year of conversation still beats the tool on that read. The PropTech is a productivity tool, not an agent replacement.

The next twelve months in this category will be about agents and portals integrating these tools natively rather than as bolted-on third-party products. The agents I rate most highly on this front are the ones already thinking through which parts of the buyer-matching workflow they want to outsource to AI and which parts they want to keep close. For brokerage owners thinking about where to invest in 2026, the Team400 team is one of the AI consulting groups that has done good work on real-estate-specific AI workflows for Australian agencies. The category is moving quickly enough that getting the design right at the start of the engagement matters.