Kitchen Stock Count Procedure: A Weekly Routine That Holds Up

Most stock counts are wrong in boring ways. Nobody is stealing the saffron; someone counted a case of tomatoes as "1" when the sheet wanted pounds, or the Sunday count happened after the Monday delivery had been half put away. The food cost number that comes out the other end looks precise and means very little.

A good count is less about effort and more about doing the same thing, the same way, at the same point in the week. This guide is the routine we would set up in a kitchen that has never done a proper count, or one that has been doing it for years and no longer believes the numbers.

Why the weekly count matters more than the monthly one

Accountants need a month-end count. Kitchen managers need a weekly one. A month is too long to find a problem: by the time you see that protein cost jumped three points, you have paid for four weeks of it and nobody remembers what happened in week two.

Weekly counts let you calculate actual usage for a short period:

Opening stock + purchases received − closing stock = what you used

Compare that with what your sales say you should have used, and the gap is your variance. If you only do one thing from this guide, make the count weekly and on the same day. Our guide on food cost percentage walks through the variance math in more detail.

Build count sheets in shelf order

The single biggest time saver is a count sheet that follows the physical layout of the kitchen, not the alphabet and not the supplier invoice. If the walk-in's left-hand shelving holds dairy on top, then proteins, then produce at the bottom, the sheet lists them in that order. The counter walks the room once, left to right, top to bottom, and never flips pages looking for "Cream, heavy."

Split the sheet by storage area:

• Walk-in cooler
• Reach-ins and line fridges
• Freezer
• Dry store
• Bar or beverage store, if the kitchen counts it
• Chemicals and disposables, if you track them as cost of goods

When the layout changes, the sheet changes the same week. A sheet that lists items where they used to live is how things get skipped. If your walk-in has no fixed layout, fix that first; walk-in cooler storage order covers how to set one up for food safety, and the same layout makes counting faster.

Units of measure: decide once, print it on the sheet

Every line on the sheet needs three things: the item, the count unit, and the cost per that unit. The count unit is the one the counter can see without opening anything or doing arithmetic.

The rule we use: weigh anything expensive or anything where a visual estimate could be off by more than 20%. Estimate everything else in sensible fractions. Do not let counters invent their own unit on the day. If the sheet says "lb" and someone writes "1 case", the line is wrong and has to be recounted.

Prepped items are where most kitchens go quiet. If you have 6 quarts of braised short rib sitting in the walk-in, that is real money. Either count prepped items at a standard recipe cost per quart or portion, or decide deliberately that you will not count them and accept that your numbers will swing with prep timing. Counting them is better. Keep the prepped list short (your top 10 to 15 batch items cover most of the value).

When to count

Count at the point in the week when stock is lowest and nothing is moving. For most restaurants that means one of two windows:

• After close on the last service day of your week, before anything is prepped for tomorrow
• Before the first delivery of the new week, early in the morning

Either works. What does not work is counting in the middle of a delivery, or on a different day each week. If a delivery lands while you are counting, leave it on the dock or in a marked area until the count is finished, then receive it. Otherwise you will count some of it and not the rest, and the invoice will go into purchases in full.

Pick the same cutoff for invoices. Anything received before the count goes into this week's purchases; anything after goes into next week's. Write the count date and time at the top of every sheet so whoever enters invoices knows where the line is.

Two-person counts

Counting alone is slow and invites mistakes nobody catches. Two people is the standard for good reason: one counts out loud, the other writes. The writer can question a number that looks odd ("we had four cases of fryer oil last week and you're saying eleven?") while the counter stays in the walk-in.

A few practices that help:

• Rotate who counts which area every few weeks so one person's habits do not become the kitchen's numbers
• The person responsible for ordering should not be the only person counting their own area
• A manager spot-checks five random lines after each area is done, and initials the sheet
• No erasing. Cross out and rewrite so a changed number is visible

In a small kitchen with a two-hour count, this is a two-person job every week. A busy 120-seat kitchen with a large dry store might need two pairs. Either way, budget the labor. A count rushed at the end of a double is worse than no count, because it produces a number people trust.

Counting each area

Walk-in first, while you are fresh, because it holds the most value and the most partials. Then freezer, because nobody wants to stay in it long, so do it with gloves and a clear sheet. Then reach-ins and the line, then dry store.

Some area-specific traps:

• Walk-in: items stored in two places (a case on the shelf and an open one on the speed rack). The sheet should have one line per item, and the counter adds both locations.
• Freezer: frost-covered unlabeled bags. If you cannot identify it, note it, do not guess. Then deal with the labeling problem separately.
• Line: half-used containers at stations. Either count the line or establish a standing rule that the line is broken down to the walk-in before the count. Pick one.
• Dry store: cases with one item left inside, counted as a full case. Open every case you are not sure about.

Valuing the stock

Multiply each count by its unit cost and total by area. Use the latest invoice price for each item unless your accountant asks for something else; most restaurants use last price paid because it is simple and close enough for weekly management. Whatever method you choose, keep it consistent, because switching methods creates variance that is not real.

Watch for price changes. If chicken went from $2.40 to $2.90 a pound this week and the sheet still has $2.40, your stock value is understated and your food cost looks worse than it is. Update prices when invoices are entered, not once a quarter.

A worked example for one week:

If your theoretical food cost from recipes and sales mix was 29.5%, you have a 2.7-point gap, which on $36,500 is roughly $985 for the week. That is the number worth chasing.

Spotting shrinkage

Shrinkage is everything that left the building without being sold: waste, over-portioning, staff meals not recorded, spoilage, theft, and receiving errors where you paid for more than arrived. The weekly count will not tell you which one. It tells you where to look.

Break variance down by category, not just the total. If dairy is fine and proteins are four points off, the problem is not "food cost", it is proteins. Then check, roughly in this order:

1. Was anything received but not invoiced, or invoiced but not received? Check the dock log against invoices.
2. Is waste being logged? If the food waste log shows nothing for proteins all week, that is a logging gap, not a perfect week.
3. Are portions right? Weigh five plates of the item during service.
4. Did a recipe change without the costing changing?
5. Is the count itself right? Recount the high-value items before concluding anything.

Theft is usually the last explanation, not the first. When it is the answer, it tends to show up as one or two specific items (spirits, premium proteins, cheese) going missing consistently, not as a general drift.

Spot counts between full counts

High-value items deserve more attention than the weekly cycle. Pick five to ten items that make up a large share of your cost or are easy to walk off with: steaks, shrimp, specialty cheese, saffron, liquor if the kitchen holds it. Count them daily or every other day, at the same time, and compare with what sales say you used.

Daily spot counts take about ten minutes. They catch over-portioning within days instead of a week, and they make the point to the team that someone is paying attention, which on its own tends to tighten things up. Rotate one or two "mystery" items into the spot list each week so it is not always the same predictable lines.

If spot counts and the full count disagree on the same item, trust neither until you find out why.

Keeping it going

The count will drift if nobody looks at the result. Review the variance by category with the head chef the same day or the next morning. Agree on one thing to check, and check it. Put the count on the weekly schedule as a recurring task with a named owner and a due time, so it does not quietly slide to "when we have time." In TenorOps, the count can sit as a weekly checklist with a photo of each signed sheet, which makes it easy to see when it was skipped.

Once counts are reliable, they feed everything else: par levels get set from real usage instead of habit, ordering gets tighter, and the food cost number in the weekly meeting is one people actually argue about instead of shrugging at.