FRNº 29Hotels & Groups

OpenAI Signals Data Shows Who Uses ChatGPT to Seek Travel Information

OpenAI's Signals data shows info-seeking at 20% of ChatGPT messages, led by Spain (26.3%) and France (26.1%) — but no travel section and no plan breakdown for advertisers.

· 3 min at the table

Finding your guests in the ChatGPT data
Finding your guests in the ChatGPT dataAI-generated

Menu du jour

  1. First course

    Info-seeking is about 20% of all ChatGPT messages, up from 16.7% in July 2024.

  2. Second course

    Spain (26.3%) and France (26.1%) led large markets for info-seeking in Q3.

  3. Third course

    Info-seeking rises from 18.1% among 18–24 year-olds to 25.2% among 55–64 year-olds.

  4. Fourth course

    Only Free and $8 Go plans show ads; Plus and Pro show none.

  5. Fifth course

    Signals contains no travel category, destination or hotel data.

Information-seeking now accounts for roughly 20% of all ChatGPT messages, up from 16.7% in July 2024, according to OpenAI's published Signals data — and that figure carries direct implications for hotels weighing ChatGPT advertising.

The dataset, which OpenAI released under the name Signals, contains no travel section. There is no travel category, no destination data and nothing on hotels. Any "X% of ChatGPT usage is travel" claim circulating in the coming weeks is a guess, perhaps an educated one, but a guess.

What the data does show is intent: what users try to do when they open the app. Roughly three quarters of information-seeking messages have nothing to do with work.

Which markets seek information most?

The share of info-seeking messages varies sharply by country. In the third quarter, Spain (26.3%) and France (26.1%) led the large markets, ahead of the UK (23.0%), Germany (22.9%) and the US (21.0%). India and Brazil sit near 16%.

For European hoteliers, that spread matters: Spanish and French users are the heaviest information seekers among big markets, while German and US users — core source markets for many European city hotels — fall several points lower.

How old are the users?

Info-seeking rises with age. It represents 18.1% of messages from 18–24 year-olds against 25.2% from 55–64 year-olds. The older users are the ones asking.

Among US users, 46% of ChatGPT messages come from people aged 35 and over — broadly the demographic that can afford a high-end hotel in a major European city. About 21% of US messages are information-seeking.

Who can actually see ads?

Only two subscription plans display advertising: Free and Go, the $8 tier. Plus and Pro subscribers see none.

This is where the data goes quiet. OpenAI publishes no breakdown of how many users sit on each plan, by country or otherwise. The only signals: Free carries the most personal use, about 75% non-work, while Pro is the least personal, with 68% of usage work-related.

The gap is the problem for operators. A Paris hotel drawing mostly American guests can confirm that 46% of US messages come from users over 35 and that 21% of US messages seek information. But it cannot confirm how many of those users are on a free account. A traveller paying $20 a month for Plus never sees an ad at all — though plenty of people with money still don't pay for software they use casually.

Should hotels buy ChatGPT ads?

The data cannot answer that question by reasoning alone. The recommended course is a small test: one leisure segment in one or two source markets, measured with the usual caveats. It is early, so operators should expect to learn rather than expect performance.

Before spending, operators can work the public dataset themselves:

  • Take the top three or four source markets for the property.
  • Check the age mix and the share of usage that is info-seeking.
  • Compare that profile with the property's actual guest base.

A hotel in Seville with mostly German guests in their fifties will see a very different picture from a Lisbon hostel full of 25-year-olds.

The dataset does not give all the answers, but it gives something — and that is more than hoteliers had before. Operators planning tests now will be first to know whether OpenAI ever publishes plan-level breakdowns by market.

via tell.martinsoler.com (Original)

Share this article:

More from Tom Whitfield

Tom Whitfield

Show full bio

Senior reporter covering companies, deals and strategy in European hospitality industry at Eurotables Report.

128 articles

Other tables tonight

« Previous tableNext table »