The Gulf Outbound Hotel Index is one business-to-business wholesale desk. More than 36,100 bookings, more than 7,340 properties, more than 720 cities, more than 110 countries.
It is not a market. It has never been claimed as one, and this post exists so that nobody has to work that out for themselves after quoting it.
Four things it cannot tell you
It cannot tell you what the Gulf outbound market did. One desk is one desk. Its destination mix reflects the agencies it served and the contracts it held. Any figure here describes that book of business. Read across to a national market and you are extrapolating from a sample that was never random.
It cannot tell you what is happening now. The file has a fixed start and end, both stated on the methodology page. It closes before the current regional conditions, and travel demand in this region has not been stable since. Every finding here is a description of a period, not a forecast.
It cannot tell you anything about a named hotel, agency or traveller. Only city- and country-level aggregates are published. No client, no passenger, no booking reference, no supplier appears anywhere in the dataset or on the pages — by design, not by omission.
It cannot tell you about consumer behaviour. These are wholesale bookings placed by travel agencies. The terms, the cancellation rules and the incentives are different from a consumer booking on a refundable rate, which is one of the reasons cancellation benchmarks disagree so violently.
There is one more limit worth stating plainly: fewer than all bookings resolved cleanly to a city. More than 90% did. The remainder are excluded from city and country figures rather than guessed into a location, which is why some totals in the destination tables do not add up to the headline count. Guessing would have produced neater tables and worse data.
Three things it tells you better than almost anything published
The exact hour a booking was created. Not the travel date — the creation timestamp, for every row. That is how more than 60% out of hours is knowable at all. Most published travel data is aggregated to the day before anybody outside sees it, and the day is where this finding disappears.
Lead time and cancellation rate on the same rows, by destination. Because both come from the same records, you can see that the destinations booked a month ahead are the ones that cancel most, and the destinations booked two days ahead cancel least. That link is invisible in any dataset where lead time and cancellations come from different sources.
A single market's week and day, unaveraged. Friday carrying less than half of Wednesday is a real property of this market. It exists in this file precisely because nothing was pooled with Europe or North America to make the sample larger.
Why I am telling you where it breaks
Two reasons, and only one of them is principle.
The principled one: a number without stated limits is not research, it is marketing with a chart.
The practical one is more interesting. A source that states its own limits is a source somebody can safely quote. A journalist, an analyst or an operator putting their name next to a number needs to know where it stops being true — and if you do not tell them, the responsible ones will not use it at all.
So: the definitions, exclusion rules, minimum observation counts and exact row counts are on the methodology page. The scope questions are answered in the FAQ. The dataset is versioned, licensed CC BY 4.0, and carries a permanent identifier so that a number quoted from it can still be traced to the exact version it came from.
Take the numbers. Use them. Attribution is the only condition.
Just do not make them say more than one desk can say.
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