Where Is Your Bar Stock Going? 10 Common Causes of Inventory Loss
A bar can hit its sales target and still lose margin through stock that was purchased but never converted into recorded revenue. The warning sign is often a persistent gap between actual cost of goods sold and theoretical cost of goods sold.
That gap is inventory variance. It can come from overpouring, waste or theft, but bad stock counts, outdated recipe costs and incorrect POS mapping can create the same result. Before blaming the team, the numbers themselves must be tested.
This guide explains how bar inventory variance works, the ten most common causes of a positive variance, and the order in which each cause should be investigated.
What actual and theoretical COGS mean
Theoretical COGS is what the bar should have consumed based on the items sold and the standard cost attached to each recipe.
Theoretical COGS = Sum of units sold × standard recipe cost per item
Actual COGS is what inventory records show was consumed during the same period.
Actual COGS = Opening inventory + purchases + transfers in − transfers out ± approved adjustments − closing inventory
Variance = Actual COGS − Theoretical COGS
A positive variance means actual consumption cost is higher than the cost predicted by sales and recipes. This may indicate product loss, but it may also mean that the theoretical model or inventory records are wrong.
For a broader explanation, see the practical guide to beverage KPIs, cost, margin and
A simple bar inventory variance example
Assume a bar reports:
Net beverage sales: $20,000
Theoretical COGS: $4,000
Actual COGS: $4,700
The dollar variance is $700. Expressed as a percentage of sales, it is 3.5%.
The bar used $700 more product than its sales and recipe data predicted. The number identifies the size of the problem, not its cause.
Turn variance into a management view
The Beverage KPI Dashboard Toolkit connects POS sales, inventory-based Actual COGS and item-level Theoretical COGS, helping managers review variance by category and identify where investigation should begin.
1. Inventory counts are inaccurate
A wrong closing count immediately changes actual COGS. Missed bottles, duplicated items, guessed partial quantities and incorrect units can all produce a false variance.
Count the same locations in the same order every time. Define one counting unit for each SKU, such as bottles, litres, cans, kegs or prepared-batch containers. Partial bottles should follow one consistent estimation method.
Counts should also be completed before stock moves again. Deliveries, transfers or service activity during a stocktake make the period unreliable.
2. The stocktake periods do not match
Actual and theoretical data must cover the same opening and closing points. A purchase entered after the cutoff, or a POS report exported for a different date range, shifts consumption into the wrong period.
Use fixed cutoff times. Confirm that opening inventory, purchases, transfers, closing inventory and POS sales all refer to the same reporting window.
3. Supplier prices changed but recipe costs did not
If a bottle now costs more but the recipe still uses the previous cost, theoretical COGS will be understated. The bar may execute every drink correctly and still show a positive variance.
Recipe costs should be updated from current invoices. Check imported spirits, wine, fresh produce and products affected by pack-size, discount, delivery-charge or currency changes.
The guide to choosing, comparing and managing bar suppliers explains how cleaner purchasing records support cost control.
4. Recipes, yields or batch costs are incomplete
A recipe may specify the correct pour while carrying the wrong cost. Missing garnishes, house-made ingredients, prep loss and batch yield can make theoretical COGS artificially low.
Every sellable item should include the full standard build. House preparations need a defined production yield and cost per usable unit. A batch that starts at 5 litres but produces 4.6 litres after transfer and filtering cannot be costed as if all 5 litres were saleable.
5. Bartenders are overpouring or changing specifications
An extra 5 ml in one drink may appear insignificant. Repeated across high-volume serves, it becomes measurable stock loss.
Common causes include free-pouring without calibration, different jigger techniques, unclear specifications and informal changes during busy service. Use one approved recipe, measured tools where appropriate, practical station training and occasional controlled pour tests.
6. Waste, spills and remakes are not recorded
Product can leave inventory legitimately without producing a sale. Spilled drinks, broken bottles, spoiled juice, flat sparkling wine, draft loss, incorrect orders and remakes all affect actual consumption.
Logged losses become explained variance. Ignored losses remain unexplained, making it difficult to identify whether the cause is training, purchasing, prep or storage.
The bar wastage and inventory-loss guide provides a practical structure for reviewing these categories.
Make legitimate product loss visible
The Waste, Comps & Staff Drinks Cost Log Toolkit helps record waste, complimentary drinks and staff drinks by item, quantity, reason and approval status, separating recurring leakage from unexplained stock loss.
7. Comps, staff drinks and tastings bypass the POS
A complimentary drink may be authorised, but the ingredients still leave stock. The same applies to staff drinks, training pours, guest tastings and promotional samples.
Create clear POS buttons or approved logs for each category. The objective is not to eliminate every free pour. It is to distinguish authorised consumption from unapproved use.
8. POS items and modifiers are mapped incorrectly
Theoretical COGS depends on the relationship between sales buttons and recipes. If a premium spirit is sold through a generic button, a double is recorded as a single, or a modifier is not connected to its ingredient cost, theoretical depletion will be wrong.
Review duplicate buttons, happy-hour items, sizes, doubles, substitutions, add-ons, removals, open-price items and package deals. Test the highest-volume and highest-cost products first.
9. Receiving, transfers and credits are incomplete
Actual COGS can be distorted when delivered quantities differ from invoices, credits are not entered, stock moves between outlets without a transfer record, or emergency purchases are omitted.
Receiving should compare the purchase order, delivered quantity, invoice price and physical condition. Multi-outlet operations need transfer records showing the item, quantity, source, destination, date and authorisation.
PAR levels can reduce emergency purchasing. See how to set and maintain PAR levels in a cocktail bar.
10. Theft or unauthorised consumption is occurring
Theft is possible, but it should not be the first conclusion. Many variance reports are damaged by counting errors, stale costs, poor mapping and unrecorded legitimate use.
Investigate theft after process and data causes have been tested. Look for repeated item-level patterns, unexplained movement in high-value products, weak access controls and discrepancies that remain after counts, recipes, waste and POS mapping have been corrected.
The goal is evidence, not accusation.
How to investigate a positive variance
Use the same order every time:
Confirm reporting dates and inventory cutoffs.
Recheck opening and closing counts for the largest gaps.
Confirm purchases, credits and transfers.
Update supplier prices and recipe costs.
Test recipe quantities, yields and batch outputs.
Audit POS mapping, modifiers, doubles and promotions.
Compare logged waste, comps and staff drinks with variance.
Observe portion control during service.
Review patterns by item, category, shift or outlet.
Escalate unresolved repeated losses for a security review.
This sequence prevents a common management failure: treating an unreliable report as proof of poor staff behaviour.
Frequently asked questions
Is actual COGS always higher than theoretical COGS?
No. Actual COGS can appear lower when purchases are missing, closing inventory is overstated, recipes overstate usage, or sales mapping depletes more product than was served. A negative variance should also be investigated.
How often should a bar review inventory variance?
The cadence should match risk and volume. Weekly review is useful for high-value or fast-moving categories because the cause is easier to trace while the period is recent. A full monthly review may still support financial reporting.
Should variance be measured in money or percentage?
Use both. Dollar variance shows financial impact. Percentage of net sales supports comparison between periods or outlets. Item-level quantity variance helps locate the operational cause.
Can inventory software eliminate variance?
No. Software can organise counts, recipes, purchases and sales mapping, but it cannot correct inconsistent counting, incomplete setup or weak service discipline by itself.
What is unexplained variance?
It is the part of the gap remaining after recorded waste, comps, staff drinks, transfers and approved adjustments have been considered. It deserves investigation because no documented reason exists for the product use.
Inventory variance should lead to action
Inventory variance is not a verdict on the bar team. It is a diagnostic signal produced by connected systems: stocktaking, purchasing, recipes, prep, POS configuration, service standards and loss reporting.
Validate the data first. Then isolate the largest item or category gaps and assign one corrective action to each verified cause. A smaller number of well-explained variances is more useful than a perfect-looking report built on weak inputs.
Apply the system to a real bar operation
For venue-specific support with inventory routines, cost control, station standards and implementation, explore The Double Strainer bar consulting services.
Join The Double Strainer Newsletter and get the free Bar Essentials guide.Practical tools for better prep, smarter batching, and cleaner service.
Written by: Riccardo Grechi | Beverage Manager, Bar Consultant & Trainer
Comments