POS vendors design their home screen for the demo, not for you: today's revenue in a big font, maybe a line chart. It's the number you can do the least about — the night already happened, and revenue alone doesn't say whether it was a good night. The reports that change decisions exist in every system I've used, buried three menus deep and exported by almost nobody. Five of them earn their place in your week.
1 — The item-level sales report
What sold, item by item, with quantities. This is the raw material for everything: the sales mix, menu decisions, and — once each item is connected to a recipe cost — the margin the night actually earned. If you export only one report, export this one, weekly at minimum. A venue that reads totals but never the item lines is running a shop without knowing what it sells.
2 — Voids, comps and discounts
The money that rang and then un-rang. Each line has an innocent explanation — a mis-ring, a gesture for a regular, a manager fixing a mistake — and in aggregate they have a shape: one station voiding more than the others, comps clustering on one shift, a discount that was supposed to be occasional running nightly. On a venue doing 40,000 a month, voids and comps drifting from 1% to 3% is 800 a month leaving quietly. You don't need to suspect anyone; you need to read the report, because the pattern is invisible ticket by ticket.
3 — Sales by daypart
Revenue sliced by service window — lunch, afternoon, dinner, late. This is the report that tells you which opening hours earn their keep and which exist out of habit. Crossed with the wage bill of the minimum crew, it answers the most expensive question an independent can ask: should this daypart exist at all? A slot that can't cover two people's loaded wages three weeks out of four isn't a slow shift, it's a subscription you forgot to cancel.
4 — Average ticket and covers
Revenue is covers times average ticket, and the split matters because the fixes are different. Covers falling with a stable ticket is a traffic problem — marketing, hours, the street. Ticket falling with stable covers is an in-house problem — the mix shifting cheap, upsells dying, a new menu quietly lowering the average. Watching revenue without the split means treating both diseases with the same medicine.
5 — Same-weekday comparison
This Tuesday against the last four Tuesdays. Weekdays have personalities — comparing Tuesday to Saturday is noise, and comparing this week to last week hides a slide that's been running for a month. Four same-weekdays in a row tell you more than any monthly total: a drift of three soft Tuesdays is a signal while it's still three nights, not thirty.
The habit that makes them pay
None of these reports pays as a one-off read — the value is the streak, the same numbers every morning until an odd day stands out on its own. That's the whole method of the restaurant KPIs guide: five numbers, one coffee. And the item-level report only reaches its full value when items carry real recipe costs — that connection is what turns sales lines into theoretical vs actual margin, which is where the money hides.
