What you learn about your audience, and where the line is

Most ticketing dashboards show you tickets sold and revenue. That is a receipt, not analytics. It tells you what happened and nothing about what to do. Every figure here is computed from your own tickets, and refunded tickets are excluded from every sales number — leaving them in inflates the picture exactly when an event is going badly, which is when it is being read most carefully.

The numbers that change a decision

• Sales over time, with the dead days drawn as dead days — a chart that skips them turns a stalled fortnight into a straight line between two good weeks
• A sell-out projection that states its own window, and refuses to give a date when the rate cannot support one
• Tier mix by revenue, which is rarely the same order as by count — a top tier selling out first is a top tier priced too low
• When people actually arrived, in fifteen-minute buckets around the advertised start, measured from real door scans
• How far ahead they book, in bands coarse enough to act on
• Repeat buyers across everything you run, counted by events attended rather than tickets bought

The forecast shows its working

It is a straight line from the last fortnight, and it says so on the card. Not because a better model is impossible, but because an organiser can check this one against their own sales in their head — and a forecast nobody can check is a forecast nobody should staff a door on. On a young event the window shrinks to the days the event has actually existed, so twenty sales yesterday reads as twenty a day, not as one and a half averaged over a fortnight it was not on sale for.

When it cannot support a date it says which kind of nothing it has: sold out, no recent sales, or a rate that will not clear the remaining stock before the doors. That last one is the useful one, and rounding it to a date after the event would be arithmetic rather than information.

On the night

The check-in page shows live occupancy for every zone as people scan through, against the limit you set. Duplicate scans are refused and shown to the door staff with the time the ticket was first used. What is not built is a single live operations screen: scans per minute, and a comparison against a forecast arrival curve, are not there — the arrival curve is read after the event, not against a prediction during it.

The reporting boundary

Organisers see aggregate audience data. They see the attendee list for their own event, because they need it to run the door and handle problems. They do not get an exportable behavioural profile of an individual across the platform — what else that person attends, what they browsed, what they nearly bought.

This is a deliberate limit and it costs us a product we could otherwise sell. Cross-organiser behavioural data is the most commercially valuable thing a ticketing platform holds, and it is also the thing customers least expect to be handing over when they buy a ticket. Aggregate insight, yes. A dossier, no.

What is deliberately absent

There is no comparison against other organisers’ events, no checkout drop-off funnel, and no traffic-source attribution, because none of those are built. Sponsor reporting is not built either. It is worth naming them: an analytics page is the easiest place in a product to imply a number exists because a chart could hold one.