Your Food Cost Is Too High. Here's Where to Look First.

What causes high food cost in a restaurant?

High food cost is usually caused by the gap between what your recipes say a dish should cost and what it actually costs to produce. That gap comes from portioning, yields, purchase price changes, waste, spoilage, incorrect POS recipes, employee meals, theft or counting errors. Raising menu prices treats the symptom. Finding the gap treats the cause.

When an owner sees food cost climb, the reflex is to reprice the menu. Sometimes prices genuinely are too low and need to move. But repricing before diagnosis means you may raise prices, annoy regulars, and still have the same leak running underneath.

Theoretical versus actual food cost

This is the single most useful concept in food cost control, and most independent restaurants never calculate it.

Theoretical food cost is what you should have spent. You take everything you sold, multiply each item by what its recipe costs, and add it up. If you sold two hundred burgers and each one costs three dollars and ten cents to make, your theoretical cost for burgers is six hundred and twenty dollars.

Actual food cost is what you really spent. Opening inventory, plus purchases, minus closing inventory. That is what genuinely left your storeroom.

Variance is the difference.

If theory says eighteen thousand and reality says twenty-one thousand, three thousand dollars of product left the building without being sold. That is not an opinion. That is the number, and everything below is about finding where it went.

Why the single monthly percentage hides the problem

Most operators track one number: total food purchases divided by food sales. It is easy to pull, and it is genuinely useful as a trend line.

But it cannot tell you anything about cause. A thirty-four percent food cost could be a purchasing problem, a portioning problem, a waste problem, a menu mix problem or a counting problem. The percentage is identical in all five cases.

Variance is diagnostic. The percentage is only a thermometer.

Where to look, in order

1. Recipe costing that reflects today's prices

Before anything else, check that your recipe costs are current. If a dish was costed two years ago, your theoretical number is fiction and the variance you calculate is meaningless.

Cost your top ten sellers at today's invoice prices. Include the parts people leave off: oil, butter, garnish, sauce cups, takeout containers, lids, bags, cutlery kits, bread that goes out free, and the dressing nobody wrote down.

2. Portioning

The most common single cause, and the easiest to verify.

Watch the arithmetic on one ounce. A dish speccing six ounces of protein at eight dollars a pound uses three dollars of product. Sent out at seven ounces, it uses three dollars and fifty cents. That fifty cents does not show anywhere except your variance.

Three hundred plates a month makes it a hundred and fifty dollars. Across four or five popular dishes with the same drift, it is most of what you were looking for.

Verify it by weighing ten random portions during a busy service. Not at prep. During service, when hands are moving fast.

Watch the fryer too. Oil usage that nobody tracks, baskets filled past capacity and product cooked in oil that should have been changed all cost money in ways that never appear on a recipe card.

3. Yields

Recipes are usually costed on what you buy, not on what survives prep.

A case of romaine does not become a case of usable romaine. A whole fish does not become its weight in fillets. Trimmed protein loses weight before it ever hits a spec.

If your recipe costs assume raw purchase weight, every yield loss shows up as variance you will never find by watching the line, because nothing is going wrong on the line.

The fix is to cost from yielded weight. Trim ten of something, weigh what is usable, and use the real ratio.

4. Purchase price changes

Prices move. Compare the last three months of invoices for your top ten items against the same period last year.

You are looking for two things. Increases that arrived quietly, and pack sizes that shrank while the price held. The second one is harder to spot and just as expensive.

5. Waste and spoilage

Waste is product ruined during production. Spoilage is product that expired before it sold. Both are variance, and neither is visible unless somebody writes it down.

Put a clipboard by the trash for two weeks. Every discarded item gets a line. Two weeks is enough to reveal the pattern, and the pattern is usually one over-prepped item, one item that never sells fast enough, and one station producing most of the errors.

6. Incorrect POS recipes

An underrated cause with a very specific signature.

If your POS thinks a dish uses four ounces of cheese and the kitchen actually uses six, your theoretical cost is wrong. Variance will show a leak that is not there, and the real leak sits in the data rather than the kitchen.

This happens constantly with dishes that changed since opening. Somebody improved the recipe. Nobody updated the system.

Before chasing a variance, confirm the recipe in the POS matches what the kitchen genuinely makes today.

7. Employee meals

Worth having, worth defining. Without a written policy, the shift meal tends toward the most expensive protein in the walk-in, and it is never recorded.

Decide what staff meal is, price it into your numbers, and ring it in so it lands somewhere countable.

8. Receiving

Product you were billed for and never received is pure variance.

If nobody checks deliveries against the invoice, weighs cases and confirms counts before signing, you are trusting the accuracy of a driver in a hurry. Most are accurate. Occasional errors are inevitable, and unchecked errors are permanent.

9. Theft

It is on the list because it is real. It belongs at the bottom because it is the least common cause and the most frequently assumed one.

Work through everything above first. If variance persists in one specific category after portioning, yields, waste, receiving and POS recipes have all been verified, then it is worth investigating. Starting here damages trust with a team that is usually not the problem.

10. Counting

Inventory errors create variance that looks exactly like every other cause.

Count the same way, in the same order, at the same time, using the same units, ideally with the same people. A count that uses cases one month and pounds the next produces a variance number with no meaning at all.

A practical order of work

Confirm recipe costs are current. Confirm POS recipes match reality. Count inventory properly. Calculate variance. Then weigh portions during service, check yields on your top items, chart twelve months of invoice prices, and log waste for two weeks.

That sequence matters. Each step removes a category of noise, so the variance you are left with points somewhere real.

When to raise prices

After the diagnosis, not before.

If your recipes are current, portions are controlled, yields are costed correctly, waste is tracked and your variance is small, and food cost is still higher than the business can carry, then your prices genuinely are too low. Now you can raise them with confidence, and you will know exactly which items need it rather than moving everything by a dollar.

That is a much better conversation to have with your regulars than a blanket increase covering a leak you never found.

The related question is what your food cost percentage should even be, and the honest answer is that there is no universal number.

Get the variance calculated properly

Working out a real theoretical-versus-actual number is the part most operators never get to, because it requires current recipes, a clean count and someone with time. That calculation is the core of our restaurant consulting audit, along with what the variance is pointing at.