Voids and refunds: what's normal and what's theft
Every kitchen has voids and refunds — they're part of real service. But the same buttons that fix honest mistakes can hide real losses. Here's how to tell the difference without becoming the paranoid boss.
Voids and refunds are normal. A server rings the wrong table. A dish goes out wrong and comes back. A card declines after the order’s in. Real service generates these every day, and a kitchen with zero voids probably isn’t recording them honestly.
So this isn’t a post about treating your team like suspects. It’s about knowing where the line sits — because the exact same buttons that correct honest mistakes are also the most common way money quietly leaves a venue.
Why voids are the classic blind spot
A void removes an item or a whole order before payment. That’s its job. But think about what it also does: it makes a sale disappear from the record.
If cash came in and the order was then voided, the receipt is gone and the drawer is over — and a quick “adjustment” balances it. No customer complains (they paid and left happy). No report flags it (the sale was “never” made). It’s the cleanest way to skim, and it hides in the noise of legitimate voids.
The same logic applies to refunds that don’t match a return, and to comps stacked on top of voids. None of this means your team is stealing. It means these are the levers if something’s wrong — so they’re the ones worth watching.
What “normal” actually looks like
There’s no universal number, but there are honest questions:
- What’s your void rate — voids as a share of transactions — and is it stable? A steady 2–3% is one thing. A number that jumps on certain shifts is another.
- Are voids concentrated? One register or one person with a void rate far above the rest could be the busy bar station… or could be worth a look. The data doesn’t accuse; it just points.
- Do refunds line up with reality? A refund should usually map to a return, a redelivery, or a card issue. Refunds with no matching story are the ones to glance at.
- When do they happen? Voids clustered right before close, or during the one unsupervised shift, are a pattern worth noticing.
Notice the framing: patterns, not single events. One big void is a Tuesday. The same big void every Tuesday on the same register is a question.
How to watch without poisoning the room
The worst thing you can do is turn this into surveillance theatre. The best operators do the opposite — they make it boring and transparent:
- Tell the team you look. Not as a threat — as a norm. “We review voids and refunds weekly” is a sentence that prevents far more than it punishes.
- Set a threshold, review the exceptions. You’re not auditing every void. You’re glancing at the handful outside the normal range. Five minutes a week.
- Lead with curiosity, not accusation. “Hey, Tuesday had a run of voids on register 2 — what happened?” Usually there’s an honest answer. Occasionally there isn’t. Either way you learn something.
Where VentaLens fits
Loyverse does the foundational work here — it records every void and refund on every receipt, reliably and for free. Having that honest, complete log is what makes any of this possible. VentaLens reads it and totals your void and refund exposure, shows the rate by register and shift, and flags the patterns that sit outside normal — the Watchdog view surfaces the dollar exposure and the where/when, so the unusual stuff is visible instead of buried. We’re careful about language: Loyverse is your source of truth, and we just surface patterns on top. We don’t call anything theft, and we don’t accuse anyone — the judgment is always yours. (No POS yet? Loyverse is free — start there.)
If you’re on Loyverse, start a free trial and look at your void rate by shift. Most owners find it’s fine — and sleep a little better for having checked.