Source notes: where every figure on this site comes from

Every figure carries its source on the page where it appears, which is the right place for a reader checking one number. This page exists for the different question: what kind of evidence does the whole index rest on, and how much of it is first-hand.

The short answer is that presence data is first-hand and pricing data mostly is not, and that asymmetry is worth understanding before relying on either.

Which markets a firm operates in comes from that firm's own published pages. This is the strongest evidence on the site, because a firm claiming a city on its own website is making a statement it can be held to.

It is also the narrowest. A firm that operates somewhere without publishing a page for it is invisible to this method, and there is no way to correct for that without inventing coverage.

Almost no firm in this category publishes its own pricing, so fee figures come from comparison sites, trade press and general reporting. That is second-hand evidence and is labelled as such wherever it appears.

One structural caution worth naming: some comparison sites are themselves operated by firms in the market. Where a roundup is published by a competitor, that is noted alongside the source rather than left for the reader to discover.

Dates and venues come from the organiser wherever one publishes them, then from official tourism and government bodies, then from reference works. The order matters, because an organiser publishing its own date is the only party who can confirm it.

Where a fixture moves with a lunar calendar, with Easter, or with a bloom, no confirmed date is given at all and the season carries the entry instead.

Each source carries the date it was read, not the date it was published. A page consulted in one month may have changed by the next, and the access date is what makes a claim checkable rather than merely attributed.

Where a figure has moved since it was recorded, the recorded figure and its access date remain the honest statement of what was true when the index was compiled.

A source is a named publication with a URL and a date on which it was read. That is a deliberately low bar in one respect and a high one in another: it admits trade press and third-party comparisons alongside a firm's own pages, and it excludes anything that cannot be pointed at, including figures that circulate in the category without an origin.

The distinction that matters most is between a firm speaking about itself and someone else speaking about it. A firm's own location page is good evidence that it claims a market and poor evidence of what that claim means operationally. A published fee in the trade press is evidence that a figure was reported, not that it is the current price. Both are recorded with the difference visible rather than flattened into a single authority.

Where a claim could not be traced to either, it is absent. Several things a reader might reasonably expect to find here are missing for that reason rather than because they were judged unimportant.

Two kinds of fact age at very different speeds in this market. A recurring season is stable; an organiser's confirmed date is not. A firm's engagement model persists for years; a published fee can be superseded without any announcement. Attaching the read date to each citation is what lets a reader tell which kind of fact they are looking at, and how much weight the age of it deserves.

It also makes the site checkable in a way an undated reference is not. Any figure here can be taken back to the page it came from and compared with what that page says now, and where the two have diverged the citation is the record of when they agreed.

They do, particularly on fees, and the disagreements are rarely small. Different publications describe different tiers of the same firm, quote figures from different years, or report a joining fee as though it were an annual one. Averaging those into a single number would produce a figure nobody published and would look more authoritative than any of the inputs.

The rule here is to keep every reported range and show the disagreement. Where a firm carries two incompatible figures, both appear with their sources, and the reader can decide which publication and which year to trust. The methodology sets out how that rule is applied.