Kerfuffle PropTech Intelligence

AI won’t make you the best estate agent. So what will?

How artificial intelligence is changing the way agencies win instructions, build relationships and run their businesses, and why the real competitive advantage might have nothing to do with technology.

By Elaine Keep, Head of Marketing at Kerfuffle

Every estate agent has heard the promises by now.

AI will write your property descriptions. Answer your phones. Chase your leads. Find your next instruction. Manage your diary. And, if you believe some of the more enthusiastic sales pitches, possibly make the tea while it’s at it.

It’s enough to make you wonder whether you’ll need any staff left by Christmas.

When the independent agency with three negotiators can use the same intelligent tools as the corporate with 300 branches? When everyone can produce polished marketing materials in seconds? When responding to an enquiry at 11pm becomes perfectly ordinary?

Where does the competitive advantage come from then?

It’s a question that came into focus at the Kerfuffle Conference, where Dr Neil Cobbold of Reapit discussed the importance of customer insight, connected data and AI-powered decision-making.

The argument was compelling. The next generation of market leaders won’t necessarily be the businesses with the most technology. They’ll be the ones who own the customer relationship.

And that’s where this conversation starts to get interesting.

Working with estate agents and property technology suppliers at Kerfuffle, we see both sides of the AI conversation. The genuine operational improvements, the impressive new tools and, occasionally, the rather ambitious promises that don’t quite survive contact with a busy branch.

That’s why we wanted to get beyond the sales pitches and hear directly from agents about what’s genuinely working.

01 / Adoption

The AI gold rush has a problem

Spend five minutes looking at property technology today and you’ll find an AI solution for almost everything.

Some are genuinely impressive. Others appear to have taken an existing piece of software, added a chatbot and discovered a remarkable new marketing opportunity.

The difficulty for estate agents is working out which is which.

What agents told us

Alex Pelosi-Buchanan, Beacons

“Start with the basics. Ask AI to teach you how to use it. And start before a competitor masters it.”

02 / Discovery

Your next instruction might start somewhere you can’t see

Imagine a homeowner sitting on the sofa one evening, thinking about moving.

A few years ago, they might have typed ‘estate agents near me’ into Google, visited a handful of websites and requested three valuations.

Now they might ask ChatGPT:

It’s a small change in behaviour with potentially enormous consequences.

Rather than being presented with a traditional list of search results, the homeowner may receive a conversational answer drawing on information gathered from several sources.

Suddenly, the question isn’t simply whether your agency ranks on Google.

It’s whether AI knows enough about your business to recommend you at all.

This is where AI search optimisation, sometimes called GEO or generative engine optimisation, becomes important.

For agents, this isn’t about abandoning traditional SEO. It’s about making expertise, local knowledge, reputation and services easier for AI systems to discover and understand.

Art Division is already helping agencies understand this emerging challenge. Its AI Search Optimisation Guide for Estate Agents explores how agencies can improve their visibility in AI-generated responses.

Explore Art Division’s AI search resources

But getting discovered is only half the battle.

Because the homeowner who eventually invites you into their living room might arrive with an AI-generated valuation estimate, a comparison of your competitors and a list of questions prepared by ChatGPT.

In other words, agents are no longer the only people arriving at valuations armed with market data.

The customer has an AI expert in their pocket, too.

That raises the stakes for local expertise. It’s no longer enough to recite what comparable properties have sold for. Agents need to explain the context, challenge questionable assumptions and demonstrate what an algorithm might miss.

03 / Lead management

The 9pm enquiry shouldn’t have to wait until 9am

It’s 9.15pm. A buyer spots a property they love and sends an enquiry.

They’re excited. They’ve already discussed where the furniture might go. They want to arrange a viewing.

Unfortunately, your office closed three hours ago.

By the following morning, they’ve enquired about another four properties with three other agents.

This is one of the more convincing practical arguments for AI.

And it’s becoming increasingly interesting when you consider what’s happening to the estate agency website itself.

For years, websites have been designed to attract visitors, show them properties and encourage them to complete a form. But what if a website could do more than simply wait for an enquiry?

Homeflow has been exploring this through its AI-powered Live Chat and Lead Manager technology, helping agencies respond to enquiries, collect useful information and identify potential opportunities.

But Homeflow co-founder Jeremy Harford Tapp makes an interesting point about where AI actually adds value.

In a recent webinar on AI and lead management, he introduced the idea of a lead value pyramid.

Both deserve a response. But do they really need the same level of human attention?

Harford Tapp argues that the problem with traditional lead management isn’t that agents don’t know what good service looks like. It’s that maintaining it consistently, across hundreds of enquiries, is incredibly difficult.

A busy negotiator might forget to ask whether an applicant has a property to sell. An enquiry might sit unanswered for half an hour. A potentially valuable conversation might never make it into the CRM.

AI doesn’t get distracted by the phone ringing or an urgent sales progression problem.

Handled properly, AI can take care of repetitive qualification, gather the right information and identify opportunities before passing the conversation to someone who can build the relationship.

As Harford Tapp explains, the quality of the system depends on its configuration. It needs to know which questions it can answer confidently, when to refer commercially sensitive conversations to a person and when to stop.

That’s arguably a much more useful way to think about AI in estate agency.

Not as a replacement for experienced people, but as a way to ensure they spend their time where their experience matters most.

And Homeflow isn’t alone in tackling the challenge.

Bridge AI manages conversations across WhatsApp, email and websites, qualifying leads, arranging appointments and recommending other suitable properties.

FlowProp AI focuses on telephone enquiries, while Enqira AI brings communications across multiple channels together, combining automated responses with human handovers.

Different approaches, but a common objective: fewer missed opportunities and more consistent service without losing the human judgement that turns an enquiry into a relationship.

04 / Your database

There’s probably money sitting in your CRM

Every estate agency has a database.

Not every agency has a particularly good relationship with the people in it.

There are buyers who stopped looking six months ago. Former tenants who are now considering buying. Landlords who haven’t heard from their agent in a year. Valuation prospects who said they weren’t quite ready.

And there are thousands of little pieces of information scattered across notes, email threads and half-completed records.

The problem isn’t necessarily that agents don’t care about these people.

It’s that finding the right opportunity, at the right moment, among thousands of contacts takes time most teams don’t have.

This is where AI gets commercially interesting.

MoxiWorks, for example, is bringing its AI-powered RISE relationship engine to the UK market.

Rather than simply storing contact information and waiting for someone to search through it, RISE is designed to identify promising opportunities within an agency’s database, suggest follow-up and support personalised marketing activity.

The UK platform also incorporates local property market intelligence powered by Dataloft.

Explore RISE and register your interest

Meanwhile, Rex Software is developing AI capabilities covering administration, prospecting, nurturing and management, helping agents identify who might be worth contacting and what the next conversation should be about.

The latter could fundamentally change how negotiators spend their time.

But there’s a catch.

AI cannot magically repair years of inaccurate records, missing information or badly managed customer relationships.

If your database is a mess, a cleverer interface won’t automatically make it valuable.

The agencies that benefit most may be the ones willing to do the decidedly unglamorous work first: cleaning records, improving processes and understanding what information they actually need.

05 / Productivity

The best AI might be the stuff your customers never notice

Not every use of AI needs to be about winning instructions.

Some of its most valuable applications may happen quietly, behind the scenes.

Consider the letting agent trying to answer a complicated compliance question while dealing with a maintenance emergency and three landlords requesting updates.

Letwise AI has been developed specifically for this challenge, providing an AI-powered compliance assistant drawing on property legislation and curated legal resources, with access to solicitor verification.

That’s an attractive proposition in an industry where regulatory mistakes can be expensive and keeping up with changing requirements is practically a job in itself.

Similarly, Stitch AI helps agencies simplify repetitive communications and administrative workflows.

And while nobody is likely to put ‘reduced WhatsApp administration’ on a conference stage in enormous flashing letters, the commercial benefit is obvious.

Every hour saved on repetitive work is an hour that could be spent on something more valuable.

06 / Cost

The hidden price of ‘cheap’ AI

There’s one rather awkward subject that doesn’t always feature prominently in the sales pitch.

What does all this actually cost?

Mal McCallion and Matt Goddard, hosts of the Property AI Report podcast, examine the realities of property technology alongside the latest AI developments.

McCallion brings more than 25 years of industry experience, including involvement in the launches of Zoopla and PrimeLocation. Goddard’s background spans estate agency and proptech, including Foxtons and Reapit.

Their industry discussions are a useful reminder that agents need to look beyond what AI can do and consider what it costs to operate.

And one of the less glamorous issues is the price of processing AI requests.

A free trial or generous introductory allowance might make an AI tool look inexpensive. But once your agency begins relying on it for thousands of conversations, property reports, document checks and other automated tasks, usage-based costs can add up.

Some services charge according to the amount of AI processing consumed, often measured in tokens or credits. Others charge for actions, calls, conversations or the number of users.

The problem isn’t necessarily that those charges are unreasonable.

It’s that they can be difficult to predict.

An agency might discover that a workflow which looked wonderfully economical during testing becomes considerably more expensive when rolled out across multiple branches.

And changing suppliers isn’t always straightforward once workflows, customer data and staff habits have become embedded.

An AI tool that saves your team two hours a day might be excellent value. An AI tool that costs twice what you expected and creates more work probably isn’t.

The cleverest technology still needs to make commercial sense.

07 / Risk

The other side of AI: when the customer uses it against you

There is another consequence of making sophisticated technology available to practically everyone.

Not everyone will use it responsibly.

Take tenant referencing.

For years, spotting fraud often meant identifying a badly altered payslip, an inconsistent employment reference or a suspicious bank statement.

AI has made producing convincing documents significantly easier.

Research published by Goodlord, examining more than one million tenancy references, estimated that suspected tenancy fraud could expose the UK’s private rented sector to up to £4.1 billion in potential annual losses.

The figure is an estimate of potential exposure across the private rented sector, not a confirmed amount lost to AI-generated fraud.

Nevertheless, the findings are concerning.

Technology that can generate convincing documents, recreate identities or alter financial information makes it harder for agents to rely on traditional visual checks.

And it’s not only lettings applications that are becoming more complicated.

Property managers are also navigating a world in which customers can ask AI to draft detailed formal complaints, challenge service charges or construct elaborate compensation demands.

Some complaints will be entirely legitimate. Others may contain incorrect legal references, exaggerated claims or arguments that don’t reflect the actual circumstances.

Either way, an AI-generated complaint can take seconds to produce and considerably longer for an agent to investigate properly.

The answer isn’t to dismiss a complaint because it sounds like ChatGPT wrote it.

It’s to verify the facts, understand the underlying issue and respond appropriately.

This is where AI creates an interesting contradiction for the industry.

It can help agents work faster, but it can also make the information arriving on their desks harder to trust.

And as customers become more comfortable using AI, agents will need stronger processes for checking documents, interpreting claims and recognising when something doesn’t add up.

08 / Accuracy

When AI starts making things up

Of course, agents themselves aren’t immune from the problem.

An AI-generated property description can sound wonderfully convincing while quietly inventing details that don’t exist.

An edited photograph can turn a cramped room into something that looks surprisingly spacious.

Virtual staging can help buyers imagine how an empty property might look furnished, but the same technology can create expectations that reality can’t possibly meet.

And when customers discover the difference, it isn’t the AI platform they’ll blame.

It’s the agent.

UK consumer protection rules require property marketing to be accurate and not misleading. Using AI doesn’t remove an agent’s responsibility to provide reliable information.

Clearly disclose virtual staging, check factual descriptions and ensure digitally altered images don’t conceal anything a buyer or tenant reasonably needs to know.

There’s a considerable difference between making marketing more attractive and making the truth less visible.

09 / The advantage

The agent still matters. Perhaps more than ever.

Let’s return to that question from the beginning.

What happens when everyone has access to the same technology?

Perhaps the genuinely interesting consequence of AI isn’t that it makes human relationships irrelevant.

Perhaps it’s that human relationships become the thing agents have left to compete on.

When every business can respond instantly, speed alone becomes less impressive.

When everyone can produce beautifully written marketing, polished language becomes less distinctive.

When every CRM can suggest who to contact next, the quality of the resulting conversation matters more.

And when every homeowner can arrive at a valuation armed with an AI-generated market report, the ability to explain, interpret and challenge information becomes a genuine advantage.

An algorithm can inform that conversation.

It can’t automatically earn the relationship.

That’s why the discussion around AI needs to move beyond features and functionality.

The question isn’t which business can automate the most.

It’s which business can use technology to become more knowledgeable, responsive and genuinely helpful to its customers.

So, what’s hot and what’s slop?

Use the technology. Keep a person in charge.

That’s precisely what we wanted to explore in our new Kerfuffle guide.

AI in the Property Industry: What’s Hot & What’s Slop is the first in our Kerfuffle PropTech Intelligence series.

We asked estate agents how they’re actually using AI, what they’re getting out of it and what advice they’d give to others.

Their answers were rather more grounded than some of the industry’s biggest promises.

Sarka Wilde of Distinct Property Consultants encourages agents to master one tool before adding another, while keeping the human element firmly in place.

That’s probably a better starting point than another subscription.

Because AI isn’t going away.

But neither is the need for good estate agents.

Free Kerfuffle guide

Find out what agents really think about AI

Our free guide, AI in the Property Industry: What’s Hot & What’s Slop, looks at real-world applications, agent experiences and the opportunities worth exploring.

Your turn

Got something to say about AI in agency?

We’re gathering more real-world experiences for the next edition. Whether you’ve found a tool that has transformed your working day or something that has failed to live up to expectations, we’d love to hear from you.

About the author

Elaine Keep is Head of Marketing at Kerfuffle, with more than 14 years experience in content marketing, digital strategy and B2B communications.

Elaine specialises in property technology, estate agency marketing and emerging technologies, including artificial intelligence. Her experience includes working with estate agency technology providers as well as enterprise AI models themselves.

Elaine works closely with proptech suppliers and the wider estate agency sector, producing industry research, educational resources, expert interviews and content designed to help agents make informed technology decisions.