Real AI Performance Metrics tracked across hundreds of independent hotels
As generative search rapidly transforms how travellers plan trips, understanding direct traffic from conversational AI has become essential for hospitality brands. Our AI Assistant Traffic Barometer tracks real performance metrics across hundreds of independent hotels.
AI Assistant Traffic Barometer For Hotels
As generative search rapidly transforms how travellers plan trips, understanding direct traffic from conversational AI has become essential for hospitality brands. Our new monthly AI Assistant Traffic Barometer tracks real performance metrics across hundreds of hotel sites
-
July 2026 Barometer
-
0.3
of overall traffic
-
0.2
of overall revenue
-
97
of AI traffic from ChatGPT
July 2026
The Learnings
ChatGTP Dominates AI Assistant Traffic
AI Assistant Channel Mix
ChatGPT dominates AI-referred traffic by a substantial margin, delivering 97% of all AI-referred traffic. Gemini comes in at only 2%, followed by Co-Pilot at 1%. All other LLMs deliver <1%.
The picture is similar from a revenue perspective. However, there has been significant change over the last few months. ChatGPT-revenue sat at 80% of overall AI-referred revenue in March, increasing to 98% in July. Gemini saw the reverse trend, dropping from 11% in March to just 2% in July.
The data does deserve a caveat: Due to the traffic and revenue volume being very small in the grand scheme of things, a few bookings can make a big difference to the percentages here.
User Location Matters
International Traffic Pattern
While 55% of the analysed traffic was from Ireland, Ireland only accounted for 40% of AI traffic.
US and rest of world (all countries excluding Ireland, UK and US) are seeing the opposite trend. The US accounts for 7% of overall traffic but 12% of AI assistant traffic, while the rest of the world accounts for 14% of overall traffic and 26% of AI assistant traffic.
Our current data sets are too small to draw definitive conclusions. However, based on the currently available data, it stands to reason that users with more complex bookings to locations they’re not familiar with (e.g. US travelling to Ireland needing flights, hotels, transport), are more likely to use AI for travel research than an Irish guest going to Dublin for a night. This emerging trend warrants further research and we will continue to monitor this closely.
About This Data
All data in this report is based on real traffic data across hundreds of independent hotel websites.
It includes referral data from all major LLMs and does not include data from Google’s AI Overviews and Google AI Mode.
Prepare your hotel for AI-driven discovery
Download our comprehensive AI Discoverability guide for independent hoteliers, designed to help you focus on the areas that matter most, so your hotel is not only easier to find but better positioned to convert interest into bookings.
Want to know more?
Get access to our past barometer data and receive future updates.