A diagram of the method, not any hotel’s actual prices.
Why a crossed-out price proves nothing
Anyone can write a higher number next to a lower one. The only way to know whether the higher number means anything is to watch it over time — and that is the part almost nobody does, because it costs a request every day for every hotel and every date.
We have recorded 26,690,589 price measurements across 25,414 hotels in 140 destinations, over 69 arrival windows. A discount appears on PriceChrono only when it is below a number we measured ourselves.
Three ways a price can be genuinely low
- Below its own median
- We recorded this hotel’s price for these dates repeatedly and it is now under its own usual level. Shown only after at least ten measurements over fourteen days — before that we would be inventing the reference.
- Cheapest of its class
- Measured against other hotels of the same star rating, in the same place, for the same dates, from the same supplier response. Not against a different hotel, a different week or a different source.
- Renovated and still underpriced
- Refurbishment year from HotelChrono, shown only where the hotel matched unambiguously. A hotel that was rebuilt last year and still sits under its class is a different kind of find.
What we throw away
A gap deeper than 60% is rejected, not displayed. Past that point the hotel’s own star rating is more likely wrong than its price low — we were once offered a “4-star for 11 EUR a night”, which is a miscategorised hotel, not a bargain. In the last run alone this rule discarded thousands of candidates.
Where we have no history of our own, we say so instead of showing a number. An empty field is honest; an invented reference is not.
What this check cannot tell you
Six dates are a sample, not a proof. A price that moves across them is behaving like a market price — it is not a guarantee that today’s offer is good. And a fixed markup is not fraud: some hotels genuinely price that way. It only means the crossed-out number carries no information.
We run this check on our own supplier’s data too, including when the answer is unflattering.