Marbella
Panorama leads investment and reputation, while Engel & Völkers leads the combined luxury analysis and German-buyer searches. Nationality changes the answer more than anything else.
AI visibility research
Real Estate AI Visibility Report: Dubai, London and Marbella
What over 10,000 AI responses reveal about how estate agencies and real estate businesses get recommended.
10,560
AI responses reviewed
116
Prompts analysed
4
Platforms monitored
ChatGPT, Gemini, Google AI Overview and Claude
Executive summary
Buyers, sellers, landlords and investors no longer begin their search with a list of links. They ask an AI platform a commercially specific question and receive an answer.
The AI system does not simply return websites. It selects a small number of companies, recommends them and explains why they may be suitable.
Schutle analysed 10,560 AI responses across 116 prompts in Marbella, Dubai and London, monitoring ChatGPT, Gemini, Google AI Overview and Claude. Only companies presented as genuine recommendations or shortlist options were counted.
Three markets, three different answers
Dubai
Allsopp & Allsopp leads buyer-support and seller searches, fäm Properties leads investment and market coverage, and Christie’s International Real Estate leads luxury.
London
Knight Frank leads Central London, Foxtons leads general East London, Dexters leads Hackney and reputation, Hurford Salvi Carr leads Docklands, and Savills leads seller outcomes.
What the research found
AI is creating a new commercial layer
Search engines present options. AI systems increasingly present conclusions, with reasons attached. An agency missing from the answer may never reach the customer’s consideration set at all.
- 3 to 5agents per responseDuring our research AI models recommended between
There is no universal winner
Leadership changes with the customer’s location, nationality, service need and the wording of the question. AI visibility is not one ranking but a portfolio of positions.
- 3Markets analysedMarbella, Dubai and London
Prompt wording materially changes the answer
Measuring one prompt is like measuring search visibility through one keyword. In the Marbella luxury analysis the top three results sat within five percentage points of each other.
- 85.5%Engel & Völkers ranked first for luxury analysis, appearing in 85.5% of responses. Drumelia was extremely close at 84.3% and ranked first for three individual prompt variants.
Local specialists can outperform global brands
Bromley Estates Marbella led the British-buyer prompt ahead of Panorama, Winkworth Spain and Engel & Völkers. A specialist local identity can beat a global network on the questions that matter locally.
- 81.2%Independent Agents Bromley Estates Marbella appeared in 81.2% of responses to the British-buyer prompt, and Hurford Salvi Carr led Canary Wharf, Docklands and E14 with 30.3%.
Category ownership matters more than broad authority
Knight Frank leads Central London and Engel & Völkers produced the strongest nationality result in the research. Brand awareness alone is not the same as owning a category.
- 63.8%Knight Frank leads Central London with 63.8% and Engel & Völkers lead Marbella prompts for German buyers appearing in 96.6% of responses
Model dependency is a hidden risk
Harbor Real Estate appeared in 95 ChatGPT responses, 15 Gemini responses and none at all in Google AI Overview. Visibility concentrated in one platform is fragile visibility.
- 37SkandiaMäklarna appeared in 37 Gemini responses and 27 Google responses but only one ChatGPT response.
AI explains why it recommends companies
Models refer consistently to reputation and trust — operating history, independent reviews, professional standards, awards, transparency and track record — and to direct relevance to the customer.
- 95.0%In London lettings, operational management capability appeared in 95.0% of responses.
Capability gaps are not evidence gaps
A capability gap means the service genuinely is not offered. An evidence gap means it is offered but the public digital footprint does not make that credible. The two need different fixes.
- 98.8%Agent capabilities were cited in 98.8% of AI responses where recommendation reasons were analysed.
Get the free report
See which agencies and brokers are leading AI recommendations. This report will help you understand why certain companies are being selected, and get practical steps to improve your company’s AI visibility.
Get the free reportMethodology
Controlled prompt monitoring.
We created a structured set of commercially relevant questions reflecting how buyers, sellers, landlords and investors may ask AI platforms to recommend a real estate company. The prompts were run repeatedly across leading AI models, including ChatGPT, Claude, Gemini and Google AI Overview where available. Geographic variation was used where supported. Prompts were written neutrally and did not direct the models toward a particular company.
Recommendation classification.
Every company appearing in an AI response was reviewed and classified before being counted. Companies were included only when they were clearly recommended or presented as part of a suitable shortlist. Incidental mentions, comparisons, citations, negative references and names appearing only in the original prompt were excluded. A company could receive a maximum of one counted appearance per response, regardless of how many times its name was repeated. Variations of the same company name were consolidated.
Balanced rankings.
Where prompts were monitored for different periods, we used prompt-balanced recommendation share. Each prompt contributed equally to its category, preventing questions with more observations from disproportionately influencing the result. Model-level results were also reviewed to identify whether a company’s visibility was consistent across platforms or heavily dependent on one AI model.
Recommendation reasons.
We analysed the reasons AI platforms expressed when recommending companies. These included reputation, reviews, local expertise, customer service, marketing reach, language capability, investment knowledge, property management and specialist market experience. The reason analysis shows the characteristics AI systems associate with a company. It does not reveal the models’ private internal decision-making processes.
What the results measure
The research measures AI recommendation visibility — how frequently and consistently companies are recommended in response to controlled customer questions. It does not independently verify service quality, transaction volume, achieved prices, investment returns or customer satisfaction. AI answers can change over time as models, source material and retrieval systems are updated.