A quarter of decided bookings on this desk were cancelled — more than 8,640 out of more than 34,500.
That single figure is almost useless, and publishing it alone would be the least interesting thing I could do with the file. The number that matters is the spread underneath it.
Depending only on where the booking was going, the cancellation rate ran from under 14% to close to 40%. Same desk. Same agencies. Same booking process. Nearly a threefold difference driven entirely by the destination. The full table is Finding 03 in the Gulf Outbound Hotel Index.
The two ends of the range
| Cancelled most | Rate |
|---|---|
| Bodrum | close to 40% |
| Geneva | close to 40% |
| Vienna | close to 37% |
| Munich | close to 36% |
| Paris | close to 36% |
| Cancelled least | Rate |
|---|---|
| Makkah | under 14% |
| Kuwait City | close to 15% |
| Salalah | close to 16% |
| Madinah | close to 17% |
| Doha | close to 17% |
Every city above clears the minimum observation count required to enter a ranking in this index — that threshold, and every other definition, is on the methodology page. Makkah alone carries more than 1,950 decided bookings, so the low end is not a small-sample effect.
At country level the same pattern holds and is easier to read. France, Switzerland, Germany and the United Kingdom all sit around or above 30%. The United Arab Emirates, Thailand and Turkey sit meaningfully below them.
Why the spread exists
Read it next to the booking window and it stops being mysterious.
The destinations that cancel least are the ones that get booked a day or two before travel: short-haul, visa-free, repeat trips, and pilgrimage. There is very little time between the decision and the trip, so there is very little room for the decision to change.
The destinations that cancel most are booked three to four weeks out and usually involve a visa, a long flight and a plan with more moving parts. A month is enough time for a visa to be refused, a schedule to move, or a family to change its mind — and each of those is a cancellation that has nothing to do with the hotel or the agency.
Cancellation risk is not a measure of customer reliability. It is a measure of how much time and how many dependencies sit between the booking and the trip.
What this means for a desk
Most agencies price cancellation risk as a single number, if they price it at all. A desk that applies one assumption across its whole book is pricing Bodrum risk and Makkah risk identically, and is therefore wrong by close to three times on one of them.
Being wrong by that margin in one direction leaks margin on every booking. Being wrong in the other direction prices you out of the business you were most likely to keep.
The practical version — which desk pays for what, and how to build a destination-tiered assumption without a data team — is a separate post, because it deserves the room.
Before you compare this to a published benchmark
You will find cancellation figures elsewhere that look nothing like these, and the reason is almost always definitional rather than real.
The rate above is cancelled divided by decided — that is, cancelled over confirmed plus cancelled, with pending bookings excluded from the denominator. Measured against every row in the file instead, the same data reads close to 24%. Both numbers are true. They answer different questions, and the definitions are stated on the methodology page rather than left for a reader to guess.
If you are comparing two cancellation benchmarks and they disagree, check the denominator before you check the market. Nine times out of ten that is the whole story.
Source: the Gulf Outbound Hotel Index, version 1.0 — more than 36,100 B2B wholesale hotel bookings from one Gulf desk. Findings and full tables · Methodology · CC BY 4.0 · 10.5281/zenodo.21796038