How AI Decides Which Estate Agents to Recommend

When someone asks ChatGPT or Google’s AI Overview to recommend an estate agent, they get one answer. Maybe two or three. Not a page of results to browse through. If your agency isn’t one of those names, you don’t exist in that moment. Understanding what drives those recommendations is now one of the most commercially important questions in estate agency marketing.

The Search Behaviour That’s Changing Everything

The traditional path to finding an estate agent, a Google search, a scroll through results, a check on Rightmove, is still common. But a growing number of buyers and sellers are skipping straight to AI. Research from Veterans United found that 39% of prospective homebuyers used AI tools in their property search, up from 34% the previous quarter. Over the same period, AI chatbot traffic grew by over 80% year on year, while traditional search engine traffic fell.

That shift matters for estate agents because AI doesn’t behave like a search engine. A search engine surfaces options. An AI assistant makes a call. It reads what’s available about you across the web, weighs it up, and either names you or doesn’t. The agents it names tend to share certain characteristics. The ones it skips tend to share others.

Confidence, Not Rankings

The most useful way to think about AI recommendations is not in terms of rankings but in terms of confidence. An AI model recommends a business when it has high confidence that the business is genuinely active, genuinely trusted, and genuinely relevant to the query. Incomplete data, inconsistent listings, sparse reviews, or vague positioning all lower that confidence. When confidence drops below a certain threshold, the AI simply doesn’t mention you.

This is a different problem from traditional SEO, where a well-optimised page can rank even if the underlying business is thin on credibility signals. AI systems pull from a much wider range of sources: review platforms, third-party citations, structured business data, editorial coverage, and the actual language people use when they write about you online.

What AI Systems Actually Read?

Reviews: Not Just the Star Rating

When an AI tool evaluates an estate agent, it reads the content of reviews, not just the aggregate score. The stories clients tell, the neighbourhoods they mention, the types of transactions they describe, all of that builds a profile. Research into which agents ChatGPT recommends found that those it surfaces average 4.3 stars. But the volume, recency, and specificity of reviews matter too. An agent with 30 recent reviews that mention specific streets, property types, and outcomes is far more legible to an AI than one with three reviews from three years ago.

Responding to reviews also signals an active, engaged business. Silence reads as neglect.

Data Accuracy Across Platforms

SOCi’s research found that business profile information was only 68% accurate on ChatGPT and Perplexity, compared to 100% accuracy on Gemini, which draws directly from Google Maps. That gap is significant. If your address, phone number, opening hours, or service description differs between your website, your Google Business Profile, and third-party directories, AI systems register the inconsistency. Inconsistent data is a credibility problem, and AI treats it as one.

Gemini’s reliance on Google Maps data also explains why a complete, well-maintained Google Business Profile is no longer optional. It’s the single most direct input into one of the most widely used AI recommendation engines.

Structured Data on Your Website

AI systems read structured data well. Clean Schema.org markup for your business, your locations, your services, and your team makes your agency machine-readable. An AI processing a query like “which agent in [area] specialises in period homes?” can only match you to that query if your website clearly and structurally answers it. Marketing language that sounds good to a human but doesn’t answer a specific question clearly is invisible to AI.

This is one reason the technical foundations of your website matter more than they used to. Our guide to AI-friendly website content for estate agents covers how to structure your pages so AI tools can actually parse and cite them.

Portal Presence vs. Direct Recommendations

The picture varies by platform. Research by Blue Array found that ChatGPT and LLaMA tend to favour property portals in their results, with Rightmove achieving the highest average probability of recommendation across multiple AI models. Google’s Gemini, by contrast, shows a bias towards recommending individual estate agents directly. That means portal presence and direct brand authority are both worth pursuing. Depending solely on Rightmove to carry your visibility into AI-powered search is a gamble on which AI tools your prospective clients happen to use.

Third-Party Citations

AI systems draw heavily on third-party sources: editorial sites, business directories, local news, and professional bodies. Research into which URLs major AI models cite found that Reddit, LinkedIn, and YouTube featured among the top sources. This is off-site territory, and it means the question isn’t just “what does your website say?” but “what does the broader web say about you?”

Propertymark membership, coverage in local property press, mentions in trade publications like The Negotiator, Estate Agent Today and Letting Agent Today, and well-maintained LinkedIn content all contribute to the picture an AI assembles about your agency. None of these are controllable in the way a webpage is. But they’re not random either. They respond to deliberate effort.

The Signals That Carry Most Weight

The table below summarises the key factors and how they map to the AI platforms most relevant to UK estate agents.

What This Means for Independent and Smaller Agents

The adoption gap is real. Around 82% of large corporate chains with 50 or more branches are already using AI tools internally, compared to 24% of independent agents with fewer than ten branches. But the gap in being recommended by AI is not yet fixed. The dominant position is still open in most markets.

That’s an opportunity, but it won’t stay open indefinitely. Agents who build credibility signals now, through reviews, consistent data, structured content, and local editorial presence, are the ones who will occupy those positions as AI-powered search continues to grow.

For independent agents, a well-built, technically sound website with clear local content is the foundation. Our work on how estate agents can appear in Google AI Overviews goes into the specific on-site steps worth taking first.

Quick Reference: AI Recommendation Signals

Next Steps

If you’re not sure where your agency currently stands in AI-powered search, start with your Google Business Profile. Check that every field is complete and that your information matches your website exactly. Then look at your reviews: how recent are they, and do they mention specific areas and property types? Those two steps alone will tell you a great deal about your current visibility.

If your website isn’t structured to answer specific, local questions clearly, that’s the next thing to address. It’s not a cosmetic issue. It’s the difference between being citable and being invisible. If you’d like to talk through what that looks like for your agency, book a free consultation with the team at Art Division.

Frequently Asked Questions