Booking & TripAdvisorBooking & TripAdvisorIf you analyse hotel reviews from more than one source, you will eventually put a Booking.com score and a TripAdvisor score in the same spreadsheet. If you do it without thinking about the scales, your analysis is wrong before it starts. This is one of those mistakes that produces confident, professional-looking output that happens to be nonsense, so it is worth getting right once.
Booking.com rates out of 10. TripAdvisor rates out of 5. An 8 and a 4 both mean “good”, but they are different numbers, and a spreadsheet that sorts them together will rank an average Booking hotel above an excellent TripAdvisor one for no reason other than the scale.
The trap is that both numbers look reasonable on their own. A column of hotel scores reading 8, 4, 9, 4.5, 7 does not announce that it is mixing scales. It looks like a normal spread of ratings. Only when you notice that the 8s and 9s are all Booking and the 4s are all TripAdvisor do you realise the “top-rated” hotels in your list are simply the ones on the 10-point scale.
This is exactly the kind of error that survives review, because nothing about it looks broken. The fix has to happen at the data level, not by spotting it later.
The arithmetic could not be simpler. To bring a 10-point score onto a 5-point scale, divide by two:
Or push everything to a percentage, which is often the cleanest because it belongs to no platform: 8 out of 10 and 4 out of 5 are both 80 percent, and that reads the same to everyone.
The discipline, not the arithmetic, is the point. Two rules:
Convert before you compare, never after. The normalisation has to be in the data before any sorting, averaging or charting touches it. Convert at analysis time and you will forget on the one occasion it matters.
Keep the original. Never overwrite the platform’s own score with the converted one. Store both: the original in one column, the normalised value in another. The day someone asks “what did Booking actually rate this hotel”, you need to be able to answer without reverse-engineering your own conversion.
Making the numbers arithmetically comparable is necessary, but do not mistake it for making them mean exactly the same thing.
The platforms gather reviews differently. Booking.com only lets a guest review a property they actually booked and stayed at, which filters out a lot of noise and tends to push scores up. TripAdvisor is more open about who can post. So even after you have converted both to a 5-point scale or a percentage, a property’s Booking score and its TripAdvisor score are not measuring quite the same population.
The practical consequence: treat a cross-platform gap as a soft signal, not a precise verdict. A hotel sitting noticeably higher on Booking than on TripAdvisor might have a real quality difference, or might just be showing the booked-guests-only effect. Converting lets you see the gap. Interpreting it still takes judgement.
All of this argues for one thing: the conversion belongs in the export, not in your head. A tool that hands you raw 8s and 4s in a single rating column has left you a landmine. A tool that gives you the original score in one column and a normalised score in another has done the discipline for you and kept the source value intact.
ExportReviews’s Booking and TripAdvisor exporter records each platform’s original rating and a normalised value side by side, precisely so a cross-platform comparison is safe by default rather than a trap you have to remember to avoid. The export guide covers the rest of the workflow.
If you take nothing else from this: a Booking 8 and a TripAdvisor 4 are the same quality and different numbers, and the difference is entirely the scale. Convert first, keep the original, and treat what is left of the gap as a hint rather than a fact.
As a proportion, yes: 8 out of 10 and 4 out of 5 are both 80 percent. But they are not the same number, and you cannot put an 8 and a 4 in the same column and sort them. To compare them you have to convert both to a shared scale first, which is exactly the step people skip.
Divide by two. An 8 out of 10 becomes a 4 out of 5, a 9 becomes 4.5. The arithmetic is trivial; the discipline is remembering to do it before you compare, and keeping the original number so you have not destroyed the source value in the process.
Any of them works as long as you pick one and apply it to everything. A percentage is often clearest because it is scale-independent and reads the same to anyone. What matters is that every number in your comparison column lives on the same scale, and that you kept each platform's original figure in a separate column.
Roughly, but not exactly, because the platforms attract and prompt reviewers differently. Booking only lets guests who actually booked and stayed leave a review, which tends to lift scores; TripAdvisor is more open. Converting makes the numbers comparable arithmetically, but you should still treat a cross-platform gap as a soft signal rather than a precise verdict.
Ready to export Booking & TripAdvisor reviews?
Export hotel and travel reviews from Booking.com and TripAdvisor to CSV, Excel or JSON. Both rating scales preserved, one extension for both platforms.
See the Booking & TripAdvisor exporter