Most restaurants do not count inventory, and the ones that do often count it in a way that produces a number they cannot use. That is a shame, because without a count you do not have a food cost figure — you have a purchases figure, which moves with delivery timing rather than with what you actually used.
Why purchases are not cost of goods
Cost of goods sold is opening inventory + purchases − closing inventory.
Skip the count and you are tracking purchases. A month with two big deliveries in the last week looks dreadful and is not. A month where you ran stock down looks excellent and is not. Both are timing, and both will send you looking for a problem that does not exist — or reassure you when there is one.
This is the single most common reason a restaurant's food cost percentage bounces around for no apparent reason.
Counting in a way that is actually useful
Same day, same time, every week. Before a delivery, after close. Consistency matters more than precision — a count that is slightly wrong the same way every week still gives you a reliable trend.
Same person, or the same method. Two people counting differently produces variance that looks like a real change.
Count what matters. You do not need every item. Your top twenty by value will be most of your food cost. A full count of everything, done monthly and dreaded, is worse than a focused count done weekly.
Count in units you buy in. Not "half a bag" — weigh it or convert consistently. Half-bag estimates are where most counting error enters.
What it tells you that nothing else does
Theoretical against actual usage. Your POS knows how many portions of a dish sold. Multiply by the recipe spec and you get what you should have used. Compare to what you actually used, and the gap is portioning drift, waste, or loss.
That single comparison is the most powerful thing inventory counting gives you, and it is invisible without a count. A dish consistently plated over spec does not show up anywhere else.
Where the waste is. Not that there is waste — which item.
Whether a price rise happened. Suppliers move prices and it is easy to miss on an invoice.
Whether your recipe costings are still true. Most restaurants cost dishes once and never revisit. Prices have moved since.
Making it survivable
The reason inventory does not happen is that it is tedious, so design it to be short:
- Order the count sheet to match the physical layout of the storage, so you walk once
- Fix the sheet order permanently, so people learn it
- Count high-value items weekly and the long tail monthly
- Two people, one counting and one writing, is faster than one doing both
- Do it at the same point in the week, always
Twenty minutes weekly is achievable. Two hours monthly is not, in the sense that it will quietly stop happening by March.
The numbers to watch
Food cost percentage, properly calculated with the count.
Variance between theoretical and actual usage, by item. Where to look.
Inventory value. Cash sitting on shelves. Restaurants are frequently surprised by how much is tied up in stock they are not turning.
Days of inventory on hand. Stock value divided by daily food cost. Rising days means money going onto shelves rather than through the till.
Where it connects
Food cost is half of prime cost, and prime cost is the number that actually tells you how the restaurant is running — because the two halves trade against each other and neither is meaningful alone. Prime cost covers that.
The other connection is menu work. Accurate item costs are what makes menu engineering real rather than theoretical, and item costs come from counting and from current invoices rather than from a spreadsheet written last year. Menu engineering has that side.
And the food cost percentage itself is more slippery than it looks — food cost percentage: the formula and the trap goes through where the formula misleads.
Starting this week
Pick your top twenty items by spend. Write a count sheet in storage order. Count on Sunday night after close. Do it again next Sunday.
Two counts give you a real cost of goods for that week. Four give you a trend. That is the whole thing, and it is a genuinely different way of running the kitchen from looking at invoices.