Refundable vs Non-Refundable: The OTA Markup You Are Actually Paying
Every time you pick the "free cancellation" rate instead of the non-refundable rate on Booking.com or Expedia, you are buying an insurance product with no risk modeling on your end. Let us actually do the math.
What the premium looks like
Across major chains and destinations, the "free cancellation" rate is typically 12–22% higher than the non-refundable advance-purchase rate. On a 4-night stay at €200/night, that is €96–€176 of pure optionality cost per booking.
What you are actually paying for
Statistically, OTAs report cancellation rates between 18% and 28% on refundable bookings (varies by market). That means roughly 75–80% of travelers who pay the premium never exercise the cancellation option. The premium is effectively funding a pool that pays out to the 20–25% who do.
The hotel's perspective
Here is what most travelers do not realize: the hotel itself charges a very small premium for refundability — typically 3–6%. The other 9–16% is OTA margin on uncertainty. You are insuring at retail OTA rates, not at hotel cost.
When to pay the premium
- Volatile travel (business, visa-dependent, medical)
- Bookings more than 90 days out where anything can change
- Destinations with known weather/political risk
When to skip it
- Dates you are 80%+ certain of
- Stays of 1–2 nights (low sunk cost if plans change)
- Bookings inside 21 days of travel (your plans are already concrete)
The arbitrage
Book the non-refundable rate. Put the 15% saved into a dedicated travel-change fund. Over 8 bookings, you have roughly paid for one complete re-booking. The OTA's refundability premium is a 15% tax on indecision, not an insurance product priced on your actual risk.
Articoli correlati
Come guadagnano le OTA con gli hotel? (Breakdown completo delle commissioni)
Booking.com, Expedia, Agoda — sembrano motori di sconti. Sono motori di commissioni. Ecco esattamente quanto prendono, a chi e perché.
23 aprile 2026
Perché i prezzi degli hotel sono diversi su ogni sito? (Non è casuale)
Stesso hotel, stesse date, cinque prezzi diversi. Non è un bug. È un'architettura di distribuzione di 40 anni. Ecco cosa sta succedendo.
23 aprile 2026
Qual è il momento migliore per prenotare un hotel? (Dati da 1,2 M di tariffe)
I consigli "prenota presto" e "prenota tardi" sono entrambi sbagliati. La vera risposta dipende da classe di hotel, destinazione e giorno della settimana. Ecco la curva.
23 aprile 2026