Method

This project follows the sourcing discipline of institutional tourism publishing, where every claim carries its origin and its date. The rules below are the ones we hold ourselves to, and they are published so you can check our work.

Sources are weighed, not ranked in public. Official statistics anchor the market context, industry research that discloses its sample carries the behavioural measures, and curated platforms are used only to find primary documents. We apply that hierarchy when deciding what belongs here; we do not stamp a grade on a publisher's name, because the grade describes disclosure rather than honesty, and a well-documented vendor survey can answer a question an official series never asks.

One figure, one document. Every indicator is a single number that appears verbatim in the document at the link given. We do not compute averages across sources, we do not round for effect, and we do not carry a number forward from a press summary into a context its publisher did not intend.

Every figure is dated twice. Once for the period it refers to, and once for the day we last checked it against its source. A statistic without a reference period is not a statistic, and in a field moving this fast an unchecked figure decays within months.

Primary where possible, labelled where not. When we locate a figure inside a second document that cites the original, we say so on the page, link both, and mark confidence as medium. Several widely quoted numbers in this field reach everyone through intermediaries, and pretending otherwise would be the easiest way to look more rigorous than we are.

Forecasts are never shown as measurements. Every indicator carries an evidence type: survey, official statistic, platform telemetry, estimate, forecast or analysis. The USD 1.4 trillion figure that dominates industry commentary is a forecast, and it is labelled as one everywhere it appears.

Samples are shown, not hidden. Where a publisher discloses sample size, field period and population, we reproduce it. Where a publisher does not, we say that too — that absence is itself information about how much weight the figure can carry.

Disagreement is reported, not resolved. When credible sources diverge, the question pages explain the mechanism — usually question wording, population or date — rather than picking a winner. Averaging two incompatible instruments produces a number that is wrong in a new way.

Corrections are visible. If a figure here is wrong, misattributed or has been superseded, write to us and the page will be corrected with the change noted. Accuracy is the entire product.