Beverage stock deficiencies : What the number is actually telling you
- David Holden
- Jun 16
- 7 min read
Updated: 1 day ago
Following beverage stocktakes across the UK, every month, in pubs, restaurants and hotels, the same meeting happens. The stock deficiency comes up. Someone offers an explanation. The room nods. The number gets accepted, reluctantly.
Next month, stocktake deficiency, same areas. Same explanation.
This is how bar shortages survive. Not because they're complicated. Because the conversation that would surface them never happens.
A stock loss isn't a mystery. It's a blue light flashing. And that blue light is almost always telling you one of three things: something is being lost, something is being missed, or something is being taken.
In my experience, persistent variances aren't random. They repeat because the cause hasn't been identified and, consequently, hasn't been addressed.
This post breaks down what each category actually looks like — and why it matters which one you're dealing with.
What a stock deficiency actually is
A stock deficiency is the gap between what your stock should have produced in revenue, based on what you have purchased, what was recorded as waste, and what was counted… and what it actually produced in revenue.
It's expressed as a retail value (selling price) and as a percentage of sales. In a well-controlled bar operation, that figure should sit between 1–3%. Above that, something needs explaining. Persistently above that, something needs investigating.
The percentage matters more than the pound figure. A £500 variance in a bar turning over £5,000 a week is a different problem to a £500 variance in a bar turning over £25,000 a week. One is a serious control failure. The other is noise.
The number on its own tells you nothing. The trend tells you everything.
Something is being lost
Operational loss is the most common cause of stock deficits, and the most accepted. It covers everything that genuinely disappears between delivery and the customer's glass, through no deliberate act.
Spirit over-pouring is one of the most widespread causes. The free-pouring of spirits, by flamboyant or rushed bar staff, contributes here. Ignoring cocktail recipes and building drinks by eye will also damage your stock in a serious way. A spirit measure 10% over on every serve might not seem like a problem, but that's 3 measures gone out of each bottle used. Across a month's trade, that's significant.
Line-cleaning wastage is unavoidable but variable. Beer lines obviously need cleaning. There are ways to minimise loss, such as cleaning as and when kegs are emptied. Know the volume of your lines, in pints, then measure what you're pulling off to minimise the amount of beer loss when pulling beer back through. Record what is being lost accurately and check whether the volume is reasonable.
Poorly maintained beer lines will produce more waste than a well-maintained one.
If line waste is being recorded as higher than expected, that's a cellar management question, not just a waste question.
Cask ale deserves its own category. Cask is the most variance-sensitive product behind any bar, for several reasons:
Poor cellar temperature causes cask to condition too quickly or turn prematurely. Stock that should have a three-day drinking window closes in one. The remainder goes down the drain.
Poor ordering is equally damaging. A cask ordered to cover a week's forecast that doesn't sell fast enough will go out of condition before it's finished. Over-ordering on cask, particularly on slower lines, is one of the most consistent causes of avoidable loss.
Inconsistent dispense means yield varies by who's behind the bar. A well-pulled pint and a poorly-pulled pint will obviously produce different yields. At volume, that difference can be significant.
Cask should be ordered tightly, rotated properly, and monitored closely. If your cask lines are producing consistent variance, the answer usually sits in the cellar before it reaches the bar.
Spillage and service waste are real but frequently overstated. A busy service produces genuine spillage. It does not produce the level of waste that some sites record. If spillage is being used as a blanket explanation for variance, the figure needs scrutiny.
Something is being missed
Control failures are different to operational losses. The product isn't necessarily going anywhere it shouldn't, but it's not being recorded correctly, which means the variance figure carries losses that have no explanation attached to them.
Delivery failures are the most common and most overlooked. Short deliveries signed for as complete. Substitutions accepted without sign-off. Damaged stock received and logged as good. Each of these enters your stock count at full value but at reduced volume or quality.
The variance appears at the next stocktake with no obvious cause.
If nobody is counting deliveries before the driver leaves, you are relying entirely on the supplier, and that is open to abuse. That is not a control.
Recording failures compound the problem. Wastage that happened but wasn't logged. A cask that turned before it was finished and went unrecorded. A line change that wasn't captured. Every unrecorded loss becomes unexplained variance.
The sheet exists to capture these moments. When it's blank, the information is gone, and the loss is just a number that provides no indication of what happened.
Something is being taken
This is the category nobody names in the meeting.
Theft and fraud in a bar environment don't look like an empty till or a missing bottle. They look like variance. They hide in the overall figure, absorbed by the operational losses around them, explained away by the same reasons that explain everything else.
The patterns are consistent once you know what to look for:
Variance that is disproportionately high on specific shifts, specific tills, or specific team members
Voids and refunds that cluster around one user or one time of day
Cash variances that repeat on the same till
Stock losses that don't correlate with volume — high variance on a quiet week, normal variance on a busy one
Wastage records that are always completed but always round numbers
None of these are proof. All of them are reasons to look more carefully.
The operational loss category and the control failure category both have straightforward remedies and, once addressed, they focus the attention on the remaining shortages. These then require a different kind of conversation — and someone willing to have it.
Why categorisation matters
A genuine operational loss, a control failure, and a behaviour problem all show up in the same variance figure.
They need completely different responses.
Retrain the team on pour discipline if the problem is over-pouring. Tighten the delivery process if stock is arriving short. Review ordering and cellar management if cask is going out of condition. Investigate the individual if the pattern points somewhere specific.
Treating a behaviour problem like an operational loss is how it continues. The team gets retrained, the process gets reviewed, and the variance carries on — because the actual cause was never addressed.
The categorisation has to come first. The response follows from it.
What a proper investigation looks like
Most stock deficiency investigations start and end with asking the bar manager for an explanation.
A proper investigation works differently:
Start with the data. Pull variance by week, not just by period. Look for patterns — which weeks are high, which are low, whether the variance correlates with volume or not.
Break down the components. Separate draught from spirits from wine from soft drinks. A variance that sits entirely in draught beer is a different problem to one spread evenly across all categories.
Review the records. Wastage sheets, delivery notes, void reports, refund logs. Not to find the answer — to find where the recording breaks down. The gap in the records is usually where the loss lives.
Count what matters. A line check on your top-selling products, comparing consumption to till sales, tells you quickly whether the variance is in the pour or elsewhere.
Follow the pattern. If the data points to a shift, a till, or a person — follow it. Don't look away because the answer is uncomfortable.
The goal isn't to produce a report. It's to identify which category the loss sits in and take the appropriate action.
The question that needs asking
If your bar variance has been explained by "we were busy" or "I forgot to log it" for more than one period in a row, it hasn't been explained. It's been accepted.
The question that closes a variance isn't "what happened?" It's "why does it keep happening?"
One is a conversation. The other is an investigation.
Your bar shortfalls won't be brought into line until someone asks the question that makes the room uncomfortable.
Common questions about bar losses
What's an acceptable bar variance percentage?
In a well-controlled operation, 1–3% is a reasonable benchmark. Above 3% consistently warrants investigation.
Above 5% persistently is a significant control problem. The percentage matters more than the pound figure — always compare variance as a proportion of bar sales, not as a standalone number.
How often should we be doing a bar stocktake?
Monthly as a minimum for the overall figure.
Weekly line checks on high-value or high-volume products give you the early warning system that monthly stocktakes can't provide.
By the time a monthly stocktake surfaces a problem, you've already had three or four weeks of the same loss repeating.
Should the bar manager be investigating their own variance?
They should be part of the process — they know the operation. But they shouldn't be the only voice in it. A bar manager investigating their own variance has an obvious conflict of interest, particularly if the cause involves their team or their own practice. An independent review, even an informal one, changes the dynamic.
What's the most common cause of bar variance in a pub estate?
In my experience, over-pouring and cask management failures account for the majority of genuine operational losses.
Control failures around deliveries and recording are the most common compounding factor. The two together can easily produce a 4–6% variance that looks inexplicable but has entirely preventable causes.
If Your Bar Variance Keeps Reappearing
That's the conversation worth having.
I work with hospitality operators, from independent pubs and restaurants to multi-site groups, to identify where bar losses are actually coming from and put in place the controls that stop them recurring.
If your variance figure has been carrying forward for longer than it should, get in touch.
Related reading:
Comments