Almost every cash-handling business — restaurants, salons, shops — eventually notices the same thing: the cash drawer at close doesn't quite match what the day's sales say it should. Most owners' first instinct is to suspect theft. Usually that's wrong, and usually there's no way to actually know, because nothing was tracked precisely enough to tell the difference between a one-off mistake and a real pattern.
What actually causes variance, in order of likelihood
- A wrong change given during a rush — far more common than deliberate theft, and invisible unless caught in the moment.
- A sale not rung in properly, so the cash in the drawer doesn't match a recorded transaction at all.
- A discount or void applied without being logged, so the 'expected' total used for comparison was never accurate in the first place.
- Genuine theft — real, but statistically the least common cause relative to the other three, and the one owners jump to first anyway.
Why tracking it matters even when the amount is small
A single day's IQD 5,000 variance looks trivial. The same gap recurring every shift, with the same staff member, over a month, is a completely different signal — but you can only see that pattern if every shift's variance is actually recorded somewhere searchable, not just noticed and shrugged off individually.
What a system should actually flag
Automatic variance calculation at shift close — expected total from recorded sales versus actual drawer count — with a record kept per shift, per staff member, is the baseline. A variance that's small and random is noise. A variance that's consistent in size or direction, or tied to one specific person's shifts, is the pattern worth actually investigating.
TU2's restaurant Business tier includes shift open/close with automatic variance calculation, with a manager PIN required for any over-limit void — built specifically so this pattern becomes visible instead of staying a vague feeling.