Month-end stock take without the spreadsheet chaos

A month-end restaurant stock take goes smoothly when the process is fixed before anyone starts counting. For a restaurant or central kitchen, that means one agreed cut-off, every movement logged around it, blind counts by storage location in the unit you actually store, recounts only where the variance is unusual, and a reason recorded for every adjustment. Most of the chaos people blame on the spreadsheet is really a process problem that the spreadsheet can’t hide.

The result is worth the discipline. A clean count gives finance a closing stock value it can book, gives operations an inventory variance it can act on, and gives the next month a starting point everyone trusts.

What is a month-end stock take, and why does it go wrong?

A month-end stock take is a full physical count of every item held at each location — central kitchen, outlets, stores, and goods in transit — taken at a fixed cut-off so that closing stock can be valued and compared against what the system expected to be there.

It goes wrong in predictable places. The hidden cost of manual stock counts is mostly time and trust; the month-end version adds a third cost, because its numbers flow straight into food cost and the month-end close. An error here doesn’t stay in the storeroom. It becomes a margin figure someone presents to management.

Where the spreadsheet version breaks

No single cut-off. One outlet counts at 10pm, another the next morning after a delivery has arrived. The numbers are each accurate and collectively meaningless.

Units that don’t match. The supplier invoices in cartons, the storeroom holds packs, the recipe uses grams. A count entered as “4” means something different in each column, and the spreadsheet won’t tell you which one was meant.

Sheets in the wrong order. Count sheets sorted alphabetically send staff walking back and forth between the chiller, the dry store and the freezer. Items get skipped, and some get counted twice.

Stock in transit falls through the gap. Goods dispatched from the central kitchen on the last evening aren’t on the kitchen’s shelf or the outlet’s. Unless someone remembers them, they disappear from the balance sheet for a month.

Variances explained by guesswork. When the count is finally keyed in and a big negative appears on chicken thigh, the explanation is whatever sounds plausible on the day. Nobody checks, so the same variance turns up the following month.

A month-end restaurant stock take in seven steps

1. Fix the cut-off. Choose one time for the whole network — typically after last service on the final day of the month — and publish it a week ahead.
2. Log every movement around it. Late deliveries, end-of-day wastage, production jobs and transfers must be recorded before the count starts, or clearly dated after it. Anything dispatched but not yet received is stock in transit.
3. Count by location, in shelf order. Build count sheets that follow the physical layout of each chiller, freezer and dry store, so the walk is a single pass.
4. Count blind. Counters see item names and units, not expected quantities. They record what’s there, not what should be.
5. Count in the unit you store. Record packs, trays or kilograms as they sit on the shelf and let the system convert to recipe and purchase units. Don’t ask staff to do conversions at 11pm.
6. Recount only the outliers. Compare counted against expected and send a second person to recount anything beyond your threshold. Leave everything else alone.
7. Investigate, adjust with a reason, then sign off. Every adjustment carries a reason code — wastage, transfer error, receiving error, unexplained — and the outlet manager and a second approver sign off before closing stock is valued.

Steps 1 and 2 prevent most of the errors. Steps 6 and 7 are where the count starts to tell you something.

How should you read inventory variance?

Inventory variance is the difference between the stock the system expected — opening stock plus receipts and production, minus sales usage, transfers and recorded wastage — and the stock actually counted. Read the pattern, not just the size.

Variance pattern	Likely cause	First thing to check
Steady negative on a high-value protein	Over-portioning, unrecorded wastage or pilferage	Actual usage against recipe quantity; wastage log
Positive variance on a dry or packaged item	Receipt not recorded, or wrong unit of measure	Goods received notes and unit conversions
Large swing that reverses next month	Cut-off error	Deliveries and transfers dated around the cut-off
Semi-finished goods short in the central kitchen	Yield lower than the recipe assumes, or unclosed production jobs	Batch tickets and actual yields
Outlet short, central kitchen long, same item	Transfer dispatched but not received	Stock in transit records

If one item keeps appearing in the first row, the gap between theoretical and actual usage is the real story, and the count is simply where you first see it.

Central kitchen inventory needs its own rules

Central kitchen inventory is harder than an outlet’s because much of it isn’t raw material. Sauces, marinated proteins, par-baked goods and portioned packs sit between purchase and sale. A few rules keep them honest:

• Count semi-finished goods as items in their own right, by batch, with their production date, rather than trying to convert them back to raw ingredients.
• Close or record open production jobs at the cut-off, so ingredients aren’t counted on the shelf and again inside a half-finished batch.
• Treat expiry as part of the count. An item past its date isn’t stock; it’s wastage that hasn’t been recorded yet.
• Reconcile transfers both ways. What the kitchen dispatched and what each outlet received should match line by line before anyone signs off.

Where RMS fits

Codemax RMS (Resource Management System) changes what the month-end count is for. Receiving, production, transfers, outlet sales and wastage all write to the same records, so stock-on-hand is a live number throughout the month rather than something rebuilt at the end of it. Recipes and multi-level BOMs give every sale a theoretical usage, so variance is calculated against what should have been consumed, item by item.

The details that cause spreadsheet errors are handled in the system: transfers are recorded at central-kitchen dispatch and again at outlet receiving, so stock in transit is visible; and FIFO and expiry checks are enforced when stock leaves the shelf. The physical count becomes a verification of a number RMS already holds.

Between counts, the AI-Kitchen Command Center watches inventory movement at every outlet against its own baseline and flags the ones drifting out of pattern — so a variance problem can surface in week two rather than at month-end.

A ten-minute check before your next count

Ask your outlet and kitchen managers four questions:

• What is the cut-off time, and does every location count at it?
• Can a counter see the expected quantity while counting?
• What happens to goods that leave the central kitchen on the last evening?
• What were last month’s three largest variances, and what was the recorded reason for each?

If the answers differ by location, or the reasons are blank, the problem isn’t the spreadsheet. It’s the process around it — and that’s the part worth fixing first.

Frequently asked questions

How often should a restaurant do a stock take?

Most restaurants do a full stock take at month-end so closing stock can be valued for the accounts, plus weekly or daily spot counts of high-value or high-risk items such as proteins, alcohol and packaging. A multi-outlet group should count every location at the same cut-off so the network can be compared like for like.

What is an acceptable inventory variance for a restaurant?

There is no single acceptable inventory variance that applies to every restaurant. Set thresholds by category and value: tight for proteins, dairy and alcohol, looser for low-cost dry goods. The trend matters more than any one month, so a variance that repeats or grows in the same item deserves investigation even if it sits inside the threshold.

What is a blind stock count?

A blind stock count is a physical count where the person counting does not see the quantity the system expects. It stops counters from confirming the expected number instead of counting, so the result is an honest measure of what is on the shelf. Expected and counted quantities are compared only after the count is submitted.

How do you count stock in a central kitchen with semi-finished goods?

In a central kitchen, count semi-finished goods such as sauces, marinated proteins and doughs as their own items, by batch, in the unit they are stored in. Close or record open production jobs at the cut-off, and treat dispatched but unreceived transfers as stock in transit so they are neither double-counted nor lost.

Should a stock take be done before or after the outlet closes?

A stock take should happen at a fixed cut-off, usually after the last service on the final day of the month and before the next day's deliveries and production begin. Any delivery, sale, transfer or wastage that happens during the count must be logged against the correct side of that cut-off.