Estate agents risk being excluded from buyers’ and sellers’ initial consideration as AI platforms produce much shorter agency shortlists, new research suggests.

Analysis of 10,560 responses from ChatGPT, Gemini, Claude and Google AI Overviews found that the platforms typically recommended just three to five agencies.

However, the agencies selected varied considerably depending on location, nationality and customer requirements. The AI platform being used also made a difference.

The research, conducted by Schutle, analysed 116 commercially focused prompts across London, Dubai and Marbella.

As a result, AI-led property searches could create a new visibility challenge for agents. A firm excluded from an AI response may not make a prospective client’s initial shortlist.

Independent agents can outperform bigger brands

The results also showed significant differences between markets and customer profiles.

For example, Knight Frank appeared in 63.8% of relevant Central London responses. Hurford Salvi Carr led searches covering Canary Wharf, Docklands and E14, appearing in 30.3%.

Meanwhile, the results were different in Marbella. Engel & Völkers appeared in 96.6% of responses relating to German buyers.

However, independent agency Bromley Estates Marbella featured in 81.2% of responses concerning British buyers. It appeared more frequently than several larger international firms.

In Dubai, Allsopp & Allsopp featured most prominently in buyer-support and seller searches. Fäm Properties, meanwhile, led investment and market-coverage prompts.

Results vary between AI platforms

Crucially, an agency’s visibility on one AI service did not guarantee similar prominence elsewhere.

Dubai-based Harbor Real Estate appeared in 95 ChatGPT responses and 15 Gemini responses. By contrast, it did not feature in Google AI Overviews during the study.

Therefore, the findings suggest agents cannot assume strong visibility on one AI platform will carry across to others.

Jaimie Beers, founder of Schutle, said: “The results show that AI is already forming a much smaller shortlist of businesses for customers.

“It is not simply reproducing Google’s search results, and there is no single agency that wins every type of question.

“The customer’s location, nationality and particular requirements can completely change the recommendations. The AI platform being used can also materially change which companies are included.”

What influences an AI recommendation?

The study also examined why AI systems recommended particular agencies.

Notably, agent capabilities featured in 98.8% of responses where researchers analysed the reasons behind recommendations.

Common factors included local expertise, operating history and independent reviews. In addition, professional standards, transparency and language capabilities featured.

Investment knowledge, property management and specialist market experience were among the other factors cited.

Beers said agencies therefore needed clear online evidence of the services and expertise they offer.

“An agency may offer an excellent service, but if its public digital footprint does not clearly demonstrate that capability, the AI may have little evidence on which to recommend it,” he said.

“This is not just about appearing in more places. It is about helping intelligent systems accurately understand who the company serves, where it operates, what it can do and why its claims should be trusted.”

Finally, the study measured how frequently agencies appeared in AI-generated recommendations. It did not assess service quality, transaction volumes, achieved prices, investment performance or customer satisfaction.

The researchers also acknowledged that AI responses can change as models, retrieval systems and their underlying online sources are updated.