- The questions
- They are written for each property in a traveller’s words. Most give a neighbourhood or a landmark and something the traveller needs, and do not name the hotel.
- The three assistants
- Each question goes to ChatGPT, Gemini and Perplexity with web search switched on. Every answer is saved word for word with its date.
- Whether the property is named
- We record whether the property appears in each answer and where it sits in the list. An assistant sometimes asks a question back, and those replies are counted separately.
- Which hotels are named instead
- We list the competing hotels the assistants name, with the number of answers each one appears in.
- Rooms, facilities and policies
- We read what each assistant says about the property and mark each detail that is wrong or out of date.
- The websites each answer cites
- We record the websites each answer links to. We note whether the hotel’s own website is one of them or whether the links go only to booking websites.
- The property’s own sources
- We then check the hotel’s website, its Google Business Profile, its reviews and its listings against what the assistants said.
- Restaurants, bars and event spaces
- Where a group runs them, they are checked with their own questions.
- What was not checked
- The report states anything we could not check, with the reason.