Bar theft prevention: stop loss before it drains your margins

Last updated: 26 June 2026

Most bar theft isn’t what you think it is. You’re not picturing a member of staff ringing in void sales and pocketing cash — though that happens. What’s actually costing you money is something far more invisible: a free-poured 25ml spirit that’s really 32–35ml, a cask that sat in a warm cellar for three days and dropped a litre of waste, a partial keg that got forgotten in the line, and stock counts that don’t match because nobody knows what a “proper” count actually looks like. A 1% stock loss on wet sales quietly costs a typical pub £3,000–£5,000 a year. Most licensees see this and assume they’ve been robbed. I used to assume the same thing. What I found, running my own pub on a tangle of spreadsheets and guesswork, was that the real damage comes from measurement error and forgotten wastage — not malice. This guide walks you through the actual mechanics of bar theft prevention, why most detection fails, and what a working count routine actually looks like so you can plug the leak within a fortnight.

What Counts as Bar Theft — and What Doesn’t

Walk into most pubs and ask the licensee what their biggest stock loss is, and you’ll hear one word: theft. What they usually mean is “I don’t know where the stock went.” That’s not the same thing. The most effective way to identify genuine bar theft is to first eliminate measurement error, over-pouring, and forgotten wastage — because those three things account for the majority of stock variance in any pub.

Real theft — a member of staff taking a bottle home, ringing in void sales and pocketing the till, pouring heavy pints and skimming the difference — that’s rare. It happens. But it’s not the reason your numbers don’t add up. I’ve run my pub for 15 years and I’ve had to deal with staff dishonesty twice in that span. Both times, the actual loss was smaller than the monthly variance caused by my own poor counting. The other 13 years? That was measurement error and wastage.

Here’s the distinction that matters:

• Deliberate theft: A staff member intentionally removes stock or cash without recording it.
• Measurement error: You think you had 10 bottles but actually had 9, or a spirit measure reads 25ml but pours 32ml because the optic is worn or the angle is wrong.
• Forgotten wastage: A line was blown and nobody recorded it. A cask sat in a warm cellar and dropped. A partial keg got left on the bar and nobody thought to note it down.
• Over-pouring: Intentional generosity (a customer says “make it a big one”) or accidental variance in free-pour technique.

If you’re not measuring the first three accurately, you’ll never know which is which — and you’ll spend your energy accusing staff of theft when the real culprit is your counting method.

The Three Hidden Ways Stock Vanishes

If you want to stop bar theft prevention in its tracks, you need to understand how each category of stock loss actually manifests on your shelves and in your till.

1. Spirits and Over-Pouring

A 25ml spirit measure is supposed to be 25ml. In reality, most bars pour between 28ml and 35ml when using a free-pour technique, especially under pressure on a busy service. A worn optic, a tilted bottle, a fast hand — all of these shift the pour weight. Over the course of a week, that’s a full bottle of lost margin. Multiply that across five or six spirits and you’re looking at 5–6 litres unaccounted for, which shows up in your stock count as a variance you can’t explain.

The number that actually matters is wet GP by line, not a single headline stock figure. If your till says you sold 40 measures of vodka this week, your stock should back that up. If it doesn’t, you have a pour variance problem — not necessarily a theft problem.

2. Draught and Cellar Waste

Temperature is the silent killer of draught margin. A beer line sitting in a warm cellar for three days will drop faster than one in a cool cellar. Poor line cleaning creates sediment and waste. A badly purged cask or a partially-empty keg that nobody tracked creates slack in your numbers. Most pubs I know don’t log cellar temperature, don’t have a line cleaning schedule written down, and don’t track which casks are partial and which are full — so when numbers don’t match, they assume someone stole from them.

3. Partial Stock and Forgotten Notes

This is the one that caught me out. I was running stock on spreadsheets and half the time I forgot to note down a partial keg because I was busy on the bar. A week later, I’d reconcile stock and think a bottle had vanished — when really it was still in the cellar, half-empty, and I’d just forgotten I’d moved it. That compounds: you think you’re missing stock, you assume theft, and meanwhile the actual stock is sitting where you left it.

Why Your Weekly Count Matters More Than Annual Stocktake

Most pubs rely on the brewery stocktaker to verify what they have once a year. That’s like checking your bank balance once a year and hoping you haven’t been robbed in the meantime. A proper weekly line check catches measurement error early and lets you spot trends before they become problems that cost you thousands.

Here’s the reality: if you’re losing 1% of stock and you only check once a year, you’ll lose money for 12 months before you know there’s a problem. By then, it could be a staff member, it could be a cellar temperature issue, it could be your counting method — and you’ll never know which. A weekly count, on the other hand, flags variance fast. Within a fortnight, you’ll see if it’s a consistent problem (which suggests a systemic issue like over-pouring or cellar temperature) or random noise (which suggests counting error).

At my own pub, I was running stock on a tangle of spreadsheets and still losing track of partial kegs and spirit measures. I built a simple count routine around a dipstick and a set of scales, and the weekly variance went from guesswork to a number I could trust within a fortnight. That’s when I realised most of my “loss” was actually measurement error — I’d been over-pouring spirits by about 6ml per measure, and my cellar was sitting at 18°C when it should have been at 10–12°C.

The goal isn’t perfect stock. The goal is a count method you trust, so when variance appears, you can actually diagnose it.

The Equipment You Actually Need

You don’t need a fancy app or expensive kit. You need three things: accuracy, consistency, and a record.

• A dipstick: Costs £15–£20. It tells you the depth of liquid in a cask in centimetres. That depth converts to a volume using the cask tag. No guessing. Most pubs don’t own a dipstick and are trying to estimate keg depth by weight or eye, which is why their numbers are wrong.
• A set of scales: A cheap digital scale (£10–£20) is enough. Weigh your open spirit bottles every count day. A 70cl bottle of vodka should weigh 875g when full. If it weighs 700g, you know exactly how much has been used. No variance. No argument.
• A tally sheet or app: Something to record what you counted and when. If you use a spreadsheet, that’s fine — but it has to be disciplined. A dedicated app like StockTap pub stock app is built for this and takes the thinking out of the process, but a spreadsheet with a template will do the job if you’re disciplined about it.

Do you really need special equipment? No. But you do need to be consistent. If you dip a cask one week and estimate the next, your variance will be noise and you won’t see real trends. A dipstick removes the guesswork.

Setting Up a Count Routine That Works

The key to bar theft prevention is not catching the thief — it’s building a routine that makes theft pointless and wastage visible. Here’s the template I use, and it works.

Choose a Day and Time

Pick the same day every week (I do Tuesday morning before service). Same time. Make it non-negotiable. This creates a baseline. Your staff know it’s coming, you build a routine, and variance becomes visible instead of hidden under the noise of changing count methods.

Count in the Same Order

Spirits first. Draught second. Ciders, alcopops, and other stock third. Same order every week. Your brain starts to notice if something’s out of place.

Weigh Spirits, Dip Draughts

Open bottles: weigh them. Sealed bottles: note the quantity. Casks and kegs: dip them. Partials: record the depth in the keg note, not as a separate “partial” — so you can track it week to week and see if it’s being used or forgotten.

Record Everything in the Same Format

Don’t vary how you write things down. Use: Brand | Size | Quantity | Weight (if applicable) | Notes. Same format, every week. This is where most pubs go wrong — they use different columns, forget to note partials, or log things differently each time, which makes week-on-week comparison impossible.

Reconcile Against Till Data the Same Day

Don’t wait until the end of the week. Count Tuesday morning, pull your till Z tape for the previous week (Monday closing to Monday closing, or whatever your cycle is), and compare. Did the till say you sold 45 measures of gin? Does your stock show you used 44–46 measures? That’s normal variance. Does it show 35? That’s a problem worth investigating. Does it show 55? You over-poured or someone’s pouring heavy — fix the optic or the technique.

Reconcile same day because your memory of what happened that week is fresh. Wait a month and you’ll have forgotten what you changed, what wasted, what didn’t work.

Spotting Real Theft (and Separating It From Waste)

Once you’ve got a disciplined count routine and you’re reconciling against till data, spotting real theft becomes much easier — because it shows up as a specific pattern, not just “variance I don’t understand.”

Real Theft Patterns

Genuine staff dishonesty usually shows up as:

• Consistent variance on one product (always 2–3 bottles of the same premium spirit missing).
• Till variance that matches stock variance (till says you sold 30 measures of vodka, stock shows 40 measures used — someone’s ringing in voids).
• Variance that improves when a specific staff member isn’t on shift, or worsens when they are.
• Missing cash that correlates with missing stock — suggesting they’re taking stock and selling it privately or taking cash without ringing it.

If you see one of these patterns consistently week after week, you have a theft problem worth investigating. But if your variance is spread across multiple products, varies week to week, and roughly correlates with till data, you’ve got a measurement or wastage problem, not a theft problem.

What to Do When You Spot It

If you suspect real theft, don’t confront the staff member directly. Document the variance, talk to your manager or supervisory staff, and look at CCTV or till logs if you have them. If the variance is small and sporadic, it might just be over-pouring or forgotten wastage. If it’s consistent and large, escalate it. Most of the time, a quiet conversation about pour technique or the importance of the weekly count is enough to resolve it — because the staff member realises you’re actually paying attention.

The biggest theft deterrent isn’t CCTV or constant supervision. It’s a count routine that works and is seen to work. Staff know that you’re counting accurately every week, that you’ll spot variance fast, and that you’re not going to accuse them of theft if the problem is actually measurement error. That’s honest and it’s a working environment where dishonesty doesn’t pay.