Using AI in Your Estate Agency

A UK government survey published in December 2024 asked 3,225 adults to describe AI in one word. The most common answer was “scary”. The second most common was “worried”. Out of the top twelve responses, only one was positive. ¹ That tells you a lot about the gap between where AI actually is and where most people think it is. For estate agents, that gap represents either a significant risk or a significant opportunity – depending on which side of it you’re standing on.

A 2024-25 survey by Viewber found that 48% of UK estate agents are actively using AI – up from 38% the year before – with a further 19% experimenting with it.² But as Toby Martin, a consultant and trainer who works with agency teams across the UK, put it on the Kotini Coffee Table: “We are barely scratching the surface as an industry of what is capable.”

Kotini is a UK property onboarding and compliance platform used by hundreds of estate agencies to manage seller and buyer onboarding, anti-money laundering (AML) checks, material information collection, and digital compliance workflows.

What “using AI” actually looks like for most agents

For the majority of estate agents, AI use is ad hoc – typing a prompt into ChatGPT to draft a listing description, generating a social media post, or running a quick search. That’s a useful starting point, but it’s not a process. Toby Martin described what he sees in training rooms across the country: “An awful lot of people are still in the starting blocks. Either sticking their head in the sand or just not feeling confident enough to really embrace AI.”

The problem isn’t enthusiasm – it’s consistency. Without a defined process for how and when to use AI tools, individual agents default to their own habits. Some use it every day; others never do. The result is inconsistent output, inconsistent quality, and no compounding benefit for the business.

The clearest marker of an agency that’s actually capturing value from AI is not the tool they’re using. It’s whether they’ve built a process around it – and almost none have. According to Martin, barely any of the agencies he works with have an AI usage policy: a document that tells team members what tools to use for which tasks, how to handle client data, and what to do with the output before it goes anywhere near a client.

Where AI is already working for agents who’ve committed to it

Prospecting and call analysis

The most measurable use case in agency right now is AI-assisted prospecting and communication analysis. John Paul, founder of Prospector Pro – an AI communication and lead generation platform – described how the technology works in practice: every call is recorded, summarised, and returned to the agent’s CRM. But the more powerful function is what happens next.

AI analyses the conversation for what should have been asked but wasn’t. On a first call with an applicant, a trained agent will ask whether the applicant has a property to sell, a mortgage in place, and a solicitor. If those questions weren’t asked, the system generates a coaching report. It also flags language that shouldn’t be there – conditional selling, pressure tactics, and other patterns that a manager would catch on a call listening session, but that most managers simply don’t have time to do at scale.

Sentiment analysis adds another layer: how much empathy was the agent showing? How were they closing? The data produces specific, actionable coaching points rather than a general sense that someone needs improvement.

On the commercial side, the same technology can analyse which types of calls are actually generating revenue. John Paul described a common revelation for agencies that first see their call data: “I didn’t realise we weren’t speaking to many revenue-generating leads. How many we weren’t speaking to? Withdrawn, previously sold, home anniversaries. It’s always tenants checking in, applicants checking in, rather than those conversations that are going to generate the revenue.”

That visibility alone changes agent behaviour without any additional training.

Marketing and content creation

Marketing is where most agents have experimented first, and where the efficiency gains are most immediately visible. Gemma Warne, founder of Agent Marketer, described the transformation for agencies that build AI into their content workflow properly: instead of spending hours trying to come up with something to say, agents can feed their expertise into AI tools and generate content ideas, hooks, scripts, and visuals that speak directly to their target clients.

The key word is “properly”. Warne was direct about the difference between useful adoption and wasted time: “If someone’s just popping in there going, I don’t know what to post today, let me just type in a couple of things and see what it spits out – that’s going to look like everyone else’s content.”

The agents seeing real results are setting up their AI tools with a full brief: brand values, tone of voice, target client types, current feed examples, recordings of how they actually speak. With that context loaded, AI can produce content that sounds like the agent – not like a generic estate agency.

One useful exercise Warne described: ask ChatGPT to crawl your website and describe back to you what your tone of voice is. The result is sometimes surprising, and often a useful prompt to tighten up how you’re presenting yourself online before you start using AI to amplify it.

AI agents and database prospecting

The most significant operational shift happening in estate agency right now – and the one where the gap between early adopters and everyone else is widest – is AI agents working through databases at scale.

John Paul described a test with one of Prospector Pro’s clients: an AI agent was sent to contact 50 to 60 people in a realistic text conversation. One of the recipients was the client’s friend, who didn’t recognise they were talking to an AI and started flirting with it. “I’ve employed worse,” he said. “It’s a 7 out of 10 as a negotiator. Not a 10. But I’ve employed worse.”

The practical application isn’t to replace negotiators – it’s to make the impossible possible. Mal McCallion, founder of ModelProp (a voice and AI property search platform), described the version of this that his platform runs: an AI agent going through a database of 300,000 contacts in the background, identifying the two or three who flag up as genuinely ready to sell.

That’s work that used to require either a large prospecting team or not getting done at all. Gemma Warne gave a parallel example from Agent Marketer’s chatbot, which her clients name and give a personality. Leads were calling the agency’s office and asking for “Sarah” – the name they’d given the chatbot. They had no idea they’d been talking to AI.

The trust question

Across all of this, Mal McCallion identified what he believes will be the defining competitive factor for agents as AI use becomes more widespread: trust.

“The currency that is going to be most important is trust. How do you show up? How do you make sure your brand is associated – as an agent – with that kind of upfrontness? Because we are going to see so much stuff that we just don’t know is real online.”

• Mal McCallion, founder, ModelProp

Being transparent about how you use AI – labelling virtually staged images, noting when AI has assisted with content, being clear about what’s automated and what isn’t – doesn’t make you look worse. According to McCallion, it makes you look trustworthy. And trustworthy is what clients are about to start specifically looking for.

What’s holding most agencies back

Toby Martin identified the gap clearly: “I believe that the vast majority of agents are still lacking are clear AI processes – set procedures that for this task we do this in this platform.”

The barrier isn’t the technology. The tools are accessible, affordable, and in many cases free. The barrier is treating AI as something you experiment with rather than something you build into how you work.

Agents who are most effective with AI have made two decisions. First, they’ve defined what AI is for in their business: which tasks it handles, which tools handle them, and what the output standard should be. Second, they’ve accepted that their business will have a competitive advantage for a limited window – and that window is now.

The agencies running AI agents through databases, using sentiment analysis to coach their teams, and using AI-assisted content to stay consistently visible are already pulling ahead. The ones waiting until it’s “more settled” are compounding the gap.

Where to start

If your agency’s AI use is currently ad hoc, three changes will move you from experimenting to operating.

First, pick one process and make it official. Prospecting callbacks, listing description drafts, viewing confirmation messages – choose the task where AI is most obviously useful and write a one-page procedure for how your team uses it. That single document starts to build the consistent habit.

Second, build a GPT or Claude project that knows your business. Feed it your website, your tone of voice, your typical client types, and your brand standards. Every piece of content that comes out of it will be anchored to who you actually are, rather than a generic version of an estate agent.

Third, look at your database. If you haven’t contacted every person in it within the last four months, you’re leaving instructions on the table. AI agents can make that outreach systematic and scalable in a way that a human prospecting team rarely manages consistently.

Frequently asked questions

Can estate agents use AI to analyse their sales calls?

Yes. AI communication platforms used by estate agents can record, transcribe, and analyse prospecting and negotiation calls. The technology identifies questions that weren’t asked, flags language patterns that indicate poor practice, and scores calls for empathy and closing technique. This produces specific coaching data rather than a general sense of how a call went.

What are the risks of estate agents using AI for client communication?

The main risks are transparency and consistency. Using AI to generate or manage communications with clients without disclosing it can damage trust if it comes to light. Using AI inconsistently – some agents following a process, others not – means the business carries compliance and quality risk without the efficiency gain. An AI usage policy that defines what tools are used for which tasks, and what must be reviewed before it reaches a client, addresses both.

How do estate agents use AI for database prospecting?

AI agents can send personalised messages at scale to an agency’s entire contact database, identify which contacts respond and with what intent, update the CRM with that information, and flag the contacts most likely to be ready to instruct. This allows a small team to maintain consistent contact with thousands of past clients, applicants, and landlords in a way that would be impossible manually.

Is AI going to replace estate agents?

No. The evidence from agencies already using AI at scale points in the opposite direction: AI handles repetitive, high-volume tasks – confirmation messages, database outreach, content first drafts, call transcription – and frees agents to spend more time on the work only they can do: negotiations, valuations, client relationships. As Toby Martin put it, “Your AI should be like a good pair of pants. It should be very supportive but largely invisible.”

What should an estate agent’s AI usage policy include?

An AI usage policy for an estate agency should cover which tools are approved for which tasks, how client data may and may not be used in AI tools, what review is required before AI-generated content or communication reaches a client, and how the team reports issues or suggests improvements. Without a policy, AI use is inconsistent and carries unnecessary risk for the business and its clients.

If you’d like to see how Kotini’s onboarding and compliance tools fit into an efficient, AI-assisted agency operation, get in touch with the team for a walkthrough.

References

1. Department for Science, Innovation and Technology (DSIT), Public Attitudes to Data and AI: Tracker Survey, Wave 4, December 2024
2. Viewber, Voices of Property survey, 2024-25