Why your till reports show nothing — even when you're losing money

You know the feeling that brings people to this page. Takings are a bit soft. The shop seemed busy. Food is going out of the kitchen. But every night, the drawer counts out right — so you tell yourself it must be the weather, or the road works, or a quiet month. Maybe. But before you settle on that, it's worth understanding exactly what the nightly count does and doesn't prove.

What the count actually checks

The end-of-day count answers one question: does the cash in the drawer match what the till says should be there?

Every serious till-skimming method exploits exactly this: change the record and the cash together, and the count stays clean. We've broken down the six common ways it's done in The 6 ways cash leaves a balanced till — unrung sales, under-rings, after-the-fact voids, fake refunds, discount abuse and the no-sale skim. This page is about the other half of the question: if the count can't show it, what can?

Six signals hiding in reports you already have

None of these proves anything on its own — remember that, because it matters both ways. But together they can tell you whether "it's probably nothing" is actually true. All of them come from your existing end-of-day (Z) report or your till's built-in logs.

1. Cash share of takings, by shift

Card payments reconcile themselves — they go straight to your bank and can't be skimmed at the counter. So when cash is being taken, cash share drops while card takings stay steady. Split your takings into cash and card by shift for a few weeks and look for the dip that follows a person, not a day of the week.

2. Voids and refunds, by person

Every till keeps a void and refund log; almost nobody reads it. Pull it monthly and break it down by staff member and by time of day. Honest mis-rings are random — they scatter across everyone. Voids that cluster on one person's shifts, or refunds processed when the shop was empty, are worth your attention.

3. No-sale drawer openings

The "no sale" button opens the drawer with no transaction. A handful a day is normal life — change for a note, a stuck receipt. Your Z report counts them. If one shift consistently runs three times the no-sales of another, ask why.

4. Average sale value, by person

Under-ringing drags one person's average ticket below their colleagues' on the same shift pattern. It's a blunt signal — menu mix and shift timing move it too — but a persistent gap that survives those explanations is telling you something.

5. Food out vs sales recorded

Your kitchen is an independent witness. Portions used, boxes gone, kitchen tickets printed — all of it should roughly reconcile with recorded sales. If stock says you sold 620 portions and the till says 570, the gap has to be explained by something: waste, staff food, miscounts — or sales that were never rung.

6. The gut-feel test, written down

Owners' instincts are data too — they're just usually not recorded. For two weeks, jot down how busy each shift looked alongside what it took. "Rammed Friday, till says quiet Tuesday" once is nothing. Every week, on the same shift, it's a signal.

From signal to certainty

Here's the frustrating part: suppose the signals line up. Cash share dips on one person's shifts, their voids run high, and the kitchen's numbers don't match the till's. You still don't know — and you can't act fairly on what you don't know. The only thing that turns a pattern into an answer is seeing what actually happened at the counter in the specific moments where the record and reality disagree.

Watching hours of camera footage back yourself is the traditional answer, and nobody does it — it's ten hours of your evening to check one shift. TillGuard does that part for you: it cross-checks the till record against what happened at the counter, and sends only the handful of moments worth a look to your phone, each with a short clip. Most flags are honest mistakes — a mis-ring, wrong change. Those cost you money too, and they're fixable with a conversation instead of a confrontation.