Your food cost percentage is probably out of date

Ask most owners what their food cost percentage is and you get a confident number. Ask when they last calculated it and the answer is usually some version of "when we built the menu."

That gap is where margin goes to die.

The number you are quoting is a photograph

Food cost percentage is simple arithmetic. For the whole operation it is cost of goods sold divided by revenue. For a single dish it is ingredient cost divided by menu price. Neither formula is hard, and neither is the problem.

The problem is that both are snapshots. You calculated them on a Tuesday using the prices you were paying that Tuesday. Every day since, the inputs have moved and the number has not.

Nothing in your business tells you when that happens. The invoice arrives, it gets paid, it goes in a folder. The price on it never travels to the spreadsheet where your costing lives. So the spreadsheet keeps reporting a 29 percent food cost with total confidence, months after it became 34.

How drift actually behaves

Price drift is not evenly distributed, which is what makes it hard to eyeball.

It concentrates. Produce, dairy, oils and proteins move constantly. Dry goods and packaging barely move. So the damage lands on whichever dishes lean on the volatile stuff, and those are rarely the dishes you would have guessed.

It compounds through preps. If your marinara went up 8 percent, every dish using marinara went up, and you never priced marinara as a line item so nothing flagged it. This is the single most common place a menu quietly goes underwater, and it is worth its own explanation, which is here.

It hides behind averages. Your overall food cost can look fine while three dishes are catastrophic, because the good ones are carrying them. Blended percentages are excellent at concealing exactly the thing you need to find.

A worked example

Say you sell a pasta dish at $18. When you costed it, ingredients came to $5.04, so 28 percent food cost. Comfortable.

Since then: olive oil up 40 percent, parmesan up 12 percent, the pasta itself up 6 percent, tomatoes flat.

The new ingredient cost is about $6.10. Your food cost on that dish is now 33.9 percent, not 28. On 40 covers a week that is roughly $42 a week of margin gone, about $2,200 a year, from one dish. You have twenty dishes.

Nothing broke. No supplier did anything unusual. You simply never re-ran the arithmetic.

What "healthy" actually means

The commonly cited healthy band is 28 to 35 percent, and it is worth being careful with that number. It is an industry average across wildly different formats. A pizzeria and a steakhouse have no business targeting the same figure. Your target depends on your format, your labour model, and your rent.

The benchmark that matters is not the industry's, it is your own from last quarter. Direction beats absolute level. A steakhouse at 38 percent that was 36 last quarter has a problem. A cafe at 26 that was 26 does not, whatever the average says.

Catching it without hiring anyone

You do not need software to fix this. You need a trigger, because the failure is that nothing ever prompts the recalculation.

The cheapest version that works: once a month, take your five highest-volume dishes, pull the most recent invoice price for each ingredient, and re-cost them. Half an hour. Write the result down somewhere you will see it next month, so you have a trend rather than a reading.

If a dish has moved more than two points, either reprice it, resize the portion, or renegotiate that ingredient. If it has moved more than five, do it this week.

That is genuinely enough for most single-location restaurants. Do it consistently and you will capture most of the available margin without buying anything. You can use our free calculator for the arithmetic.

The reason people buy software for this is not that the maths is hard. It is that "once a month, without fail, forever" turns out to be extremely hard when you are also running a kitchen. Tools like ours exist to remove the discipline requirement, not the arithmetic. If you want to see what that looks like, the demo runs on sample data, and there is a comparison of the alternatives here including the ones that cost more than we do.

The one habit worth building

Whatever you do about software, build this reflex: when a price surprises you on an invoice, ask immediately which dishes it touches.

Most owners register the surprise, feel briefly annoyed, and pay it. The useful move is the next question, and it takes thirty seconds. That single habit catches most of the drift before it turns into a quarter of quietly eroded margin.