The AI tools investigated claim people in cities including Glasgow, Manchester, and Sheffield can’t afford to go abroad, recommending staycations instead. But real booking data from experts tells a very different story.
The study compared AI’s travel recommendations to AttractionTickets.com’s real-time internal booking data, allowing the experts to see if there was a correlation between where AI thinks Brits should be travelling and where they’re actually buying tickets to travel.
The study found that AI assumed people living in locations like Glasgow and Manchester couldn’t afford to travel abroad to places like Walt Disney World and would be more likely to spend their holiday in the Lake District.
The experts can reveal that the AI made a wrong assumption on 57% of the UK cities included in the study, with the tools suggesting that people living there could only afford staycations as opposed to overseas holidays.
How badly did AI get it wrong?
Out of the 35 major UK cities analysed, the 20 cities listed below were capped by AI on the places they could “afford” to travel to.
(AI’s suggested “affordable” staycation vs. real-world 2024 overseas booking popularity)
- Glasgow – AI suggested: Lake District | Actual overseas booking rank: 2nd
- Manchester – AI suggested: Lake District | Actual overseas booking rank: 4th
- Sheffield – AI suggested: Peak District | Actual overseas booking rank: 11th
- Cardiff – AI suggested: Pembrokeshire | Actual overseas booking rank: 14th
- Norwich – AI suggested: Norfolk Broads | Actual overseas booking rank: 15th
- Stoke-on-Trent – AI suggested: North Wales | Actual overseas booking rank: 18th
- Swansea – AI suggested: Gower Peninsula | Actual overseas booking rank: 19th
- Plymouth – AI suggested: Cornwall | Actual overseas booking rank: 20th
- Aberdeen – AI suggested: Scottish Highlands | Actual overseas booking rank: 21st
- Dundee – AI suggested: Scottish Highlands | Actual overseas booking rank: 22nd
- York – AI suggested: Yorkshire Dales | Actual overseas booking rank: 24th
- Durham – AI suggested: Lake District | Actual overseas booking rank: 25th
- Lincoln – AI suggested: Norfolk Broads | Actual overseas booking rank: 26th
- Portsmouth – AI suggested: Isle of Wight | Actual overseas booking rank: 27th
- Chester – AI suggested: North Wales | Actual overseas booking rank: 28th
- Cambridge – AI suggested: Suffolk Coast | Actual overseas booking rank: 29th
- Perth – AI suggested: Scottish Highlands | Actual overseas booking rank: 30th
- Stirling – AI suggested: Loch Lomond | Actual overseas booking rank: 31st
- Inverness – AI suggested: Isle of Skye | Actual overseas booking rank: 32nd
- Salisbury – AI suggested: Dorset Coast | Actual overseas booking rank: 34th
The Study’s Key Findings:
Both Glasgow and Manchester were assumed by AI to have a low to moderate chance of booking overseas holidays, whereas AttractionTickets.com’s internal data showed that they were in the top 5 locations for making overseas holiday bookings.
AI also seemed to have the biggest bias towards the north of England, with five locations featured on the list including Manchester, Chester, Durham, York, and Sheffield.
In addition to the north, AI also appeared to have a bias towards navy cities, with both Plymouth and Portsmouth featured on the list.
30% of cities AI decided would only be able to afford a staycation as opposed to traveling abroad were actually classed in the top 50% of AttractionTickets.com’s real-time data affordability ranking.
Chris Bradshaw, Marketing Director at AttractionTickets.com, commented on the findings: “I was surprised to see just how strikingly inaccurate the results were on the AI portion of this study. This technology made assumptions that people residing in Manchester, for example, were one of the least likely to travel abroad when in actuality our booking data shows that Manchester ranks among the top five cities for our Walt Disney World Florida bookings.
Places like Glasgow, Sheffield, Cardiff, Norwich, and Stoke-on-Trent were also underestimated, despite the strong booking numbers we can see internally.
My concern is that if people are using tools like Gemini or ChatGPT to plan their holidays, there may be a bias that the AI is holding against them, and potentially restricting dream holiday destinations because it has assumed their level of affordability. People should be aware of this bias before using these tools to plan their holidays.
AI doesn’t account for human determination because no matter where we’re located, many families will save, plan, and find ways to make their dream trips a reality.”