The chain: from fair to scan

Five concepts, in the order they happen. Almost every column in the data hangs off one of them.

Fair event

A fair event (Fairevent*) is one edition of an event: a name and a year, with a start and end date. "Trade Fair 2026" and "Trade Fair 2027" are two fair events, not one.

Several editions over the years are grouped by a fair brand (FairBrand*) — that is the thing whose growth you compare year over year.

Filter events by name and year separately. Never fold the year into the name, and never match names with contains — similarly named events are common and will silently merge.

Article

An article (Article*) is what is actually sold or issued: a day ticket, a multi-day pass, an exhibitor badge, a press pass, a parking ticket. It carries the price, the validity and the rules.

ArticleType distinguishes what kind of thing it is — a ticket, a voucher, a service fee and so on. See Datasets → Column values for the full list.

Articles are the natural grain for "what sells", and they are what links a ticket to its statistics groups — which is how attendance reporting is structured. See Counting visitors.

Sale

A sale (Sale*) is one purchase transaction: who bought, when, through which channel, paid or unpaid, test or real, and with which campaign parameters (SaleUtm*).

A sale carries four separate status columns — overall status, payment, registration and receipt — because those four things genuinely progress independently. Their value sets are in Column values.

One sale can produce many tickets — somebody buying five badges for their team is one sale, five tickets.

Sales carry a SaleIsTest flag, and test sales genuinely exist in production data. Excluding them is not optional.

Ticket

A ticket (Ticket*) is one admission entitlement, belonging to one person. It is the centre of this data model, and the Tickets dataset flattens everything above onto each ticket row: its event, its article, its sale, its buyer and its holder.

A ticket has a status, may be cancelled, may be free, and may be a day ticket or a permanent one. It also carries links to its digital forms — PDF, mobile ticket, Apple and Google Wallet.

Usage

A usage (the TicketUsages dataset) is a scan: someone presented a ticket at an entrance and a turnstile or handheld device read it. Each usage records when, at which entrance, on which terminal, whether it was an entry or an exit, and whether it was the first scan of that day or of the whole event.

A ticket can have zero usages — sold but never used, a normal and important case — or many, because visitors come back on several days and leave and re-enter.

This is where the sold-versus-attended gap lives, and it is why TicketUsages is a separate dataset rather than a column on the ticket.

Two more things attached to tickets

Legitimation (Legitimation*) is the proof someone showed to qualify for a restricted article — a press card, a student ID, a trade-visitor credential. Where an article is restricted, the legitimation says on what grounds it was granted, from when to when it is valid, and how far its verification got (LegitimationStatus).

Survey answers (the Surveys dataset) are what the visitor told you during registration or afterwards. They join back to the ticket, so any answer can be segmented by anything the ticket knows.