Actual vs theoretical food cost: where the gap is

Actual vs theoretical food cost is the comparison that turns a food cost figure from a score into a diagnosis. The actual figure tells you what the kitchen really used, measured from stock. The theoretical figure tells you what it should have used if every dish sold had gone out exactly to its recipe, with nothing wasted, dropped, over-portioned or taken home. The difference between the two is not a rounding error. It is a list of specific things that went wrong, and most of them can be found.

Plenty of venues calculate only the actual figure. They see it move, they feel the pressure, and they respond by tightening everything at once: smaller portions, cheaper suppliers, a stern word at the pass. That is expensive guesswork. The theoretical figure tells you whether the movement came from the kitchen at all, or from the menu mix, or from a price change nobody noticed.

This guide explains how each number is built, what a normal gap looks like for your venue, how to break the gap down into its causes, and the order in which to chase them. The examples use illustrative figures rather than benchmarks, because the only gap that matters is your own.

Actual food cost is calculated from stock. You take what you had at the start of the period, add what was delivered, take away what is left at the end, and divide the result by food sales. That number is the cost of goods sold, and it captures everything that left the shelves for any reason: the plates that went to guests, the trim in the bin, the soup that burnt, the case of lemons that turned, the staff lunch, and anything that walked out of the back door.

That breadth is both its strength and its weakness. Because actual food cost includes everything, it cannot miss anything. But because it includes everything, it cannot tell you which of those things happened. A food cost percentage of 34 might be a kitchen running perfectly on an expensive menu, or a kitchen running badly on a cheap one. On its own, the number cannot say. Our step-by-step guide to calculating food cost percentage covers how to get the actual figure right before you compare it with anything.

Theoretical food cost starts from the other end: from sales. For every dish sold in the period, you take its costed recipe, multiply by the number sold, and add the results together. That gives you the cost of the food that should have been used. Divide it by food sales and you have the theoretical food cost percentage.

The theoretical food cost assumes perfection. Every portion is weighed, every recipe followed, every delivery used before it spoils, and nothing goes out without being rung through the till. No kitchen achieves that, which is exactly why the figure is useful. It is the floor. The actual figure can approach it but will always sit above it, and the distance between them is the cost of everything that was not perfect.

Two things have to be true for the theoretical figure to mean anything. Every dish on the menu needs a costed recipe, including the sauces, garnishes and sides that are easy to forget. And the costs need to be current. A recipe costed at last spring's prices will understate the theoretical figure and make the gap look worse than it is. Our guide to recipe costing walks through one dish in full.

The theoretical food cost moves even when the kitchen does nothing differently. If guests order more of your expensive-to-make dishes this month, the theoretical figure rises. If a cheap pasta becomes the bestseller, it falls. This is the most important reason to calculate it: without it, a change caused by what guests chose looks exactly like a change caused by what the kitchen did.

Here is an illustrative month for a restaurant with a mid-priced menu. These numbers show how the comparison works; they are not targets.

The gap is 3,600, or three percentage points. That is the food cost variance for the month, and it is the figure to investigate. Three points sounds small. On 120,000 of sales it is the cost of a part-time cook for most of a month, every month, until somebody finds it.

The next step is to break that 3,600 into pieces. The rest of this guide does exactly that, cause by cause.

Almost every gap between actual and theoretical food cost comes from some combination of five causes. They are listed here roughly in the order they are cheapest to check.

Before assuming anything happened in the kitchen, check that the actual figure is right. A missed shelf in the closing count, a delivery dated into the wrong month, a credit note never applied, or a transfer to another venue that was never recorded will all create a gap that does not exist. This is an inventory variance in the narrowest sense: the difference between what the records say you have and what is really on the shelves, and it is often a records problem rather than a food problem.

The discipline of keeping records and shelves in agreement is a whole field of its own, usually called inventory control, and a restaurant needs only a small, steady piece of it: one count, one cut-off, one way of valuing stock.

If a supplier raised the price of your main protein by eight percent in the middle of the month, your actual cost reflects it immediately and your theoretical cost does not, because the recipe still carries the old price. Compare this period's invoice prices with the prices in your recipe costings for your twenty highest-value lines. Where they differ, update the recipe, then recalculate the theoretical figure before going any further.

Food that is bought, prepared and thrown away shows up in actual food cost and nowhere in theoretical. This includes spoilage in storage, over-production on the line, dropped plates, burnt batches and trim that could have been used. A short waste log, kept for even two weeks, will tell you how much of the gap sits here. Our guide on how to reduce food waste in a restaurant kitchen covers where to start.

If the recipe says 180 grams and the line serves 200, every plate costs eleven percent more than the theoretical figure assumes. Recipes also drift: a garnish gets more generous, a sauce gets richer, a side gets bigger because a guest once complained. None of it is malicious and all of it is expensive. Weigh ten portions of your top five dishes on a busy night and compare them with the recipe card.

The last cause is food that was used but never paid for. Some of it is legitimate and simply unrecorded: staff meals, tastings, comps for a guest who waited too long. Some of it is not. The way to separate them is to record the legitimate kind properly. Once staff food, comps and tastings are logged at cost, whatever gap remains is much easier to see clearly.

The point of the five causes is to turn one gap into several smaller, explained ones. Continuing the illustrative example, a fortnight of investigation might produce something like this:

That accounts for 3,600, and every line now has an owner and a fix. The paperwork errors are fixed by changing the count routine. The price change is fixed by updating recipes. The waste and portioning are fixed in the kitchen. The staff food is fixed by recording it. None of those fixes involves cutting the quality of a single dish.

In practice you will rarely account for every last unit of the gap, and that is fine. The aim is to explain most of it, fix the largest pieces, and keep watching the remainder.

A single gap for the whole kitchen tells you how much is missing. A gap for each of your most valuable ingredients tells you where it went. This is the most useful refinement of the comparison, and it does not need any special tools, only a little patience in the first month.

For each of your top twenty ingredients by spend, work out two quantities for the period. The first is theoretical usage: for every dish that contains the ingredient, multiply the quantity in the recipe by the number of that dish sold, and add the results together. The second is actual usage: opening stock of that ingredient, plus what was delivered, minus what is left. Put the two side by side and the ingredient with the widest gap names itself.

Suppose the recipes say the kitchen should have used 61 kilos of beef mince over the month and the stock movement says it used 70. That nine kilo difference is a specific question with a short list of answers. Were the burgers heavier than spec? Was a batch of ragù thrown out? Was mince used in a staff meal or a special that never had a recipe? A question that specific can usually be answered in one conversation with the sous chef, which is far more than a vague complaint about food cost will ever achieve.

Twenty ingredients usually cover most of the money in a kitchen, so this exercise captures most of the gap without costing every grain of salt. Once the top twenty are stable, extend it to the next ten if the remaining gap is still worth chasing.

A gap with no owner never closes. The comparison works best when the work is split cleanly between the people who can actually change each part of it.

The head chef owns the recipes and the portions: whether the costed recipe matches what really goes on the plate, and whether the line serves it consistently. The person who does the ordering owns prices and deliveries: whether the recipe prices are current and whether every delivery and credit is recorded in the right period. The manager running the floor owns food that leaves without a sale: comps, staff meals and anything given away, recorded properly every time. Whoever runs the numbers owns the comparison itself and brings it to a short weekly conversation with all three.

That conversation should be brief and specific. Which ingredient moved, what the owner thinks happened, and what will be different next week. It should never become a hunt for someone to blame, because the moment the gap becomes a weapon, people stop recording the things that make it accurate, and the number gets worse while looking better.

A gap that narrows is usually good news. Occasionally it is a warning, and it is worth knowing the signs before celebrating.

The theoretical figure can creep upward if recipes are re-costed with generous allowances, a little extra here for waste and a little there for luck. The gap closes, but only because the target moved toward the result. The actual figure can drop if a closing count is inflated, deliberately or by accident, which pushes this month's problem into next month's. And the gap can vanish for a period simply because a large delivery was dated into the following month.

The check is simple. When the gap narrows sharply, look at whether the recipes changed, whether the count was done the usual way, and whether next month's figure jumps back. A real improvement holds. An accounting improvement reverses within a period or two.

There is no universal figure for an acceptable gap. A kitchen with a short menu, few prepared components and a tight team will naturally run close to its theoretical figure. A kitchen with a long menu, lots of house-made sauces and a high-turnover team will run further from it. What matters is that you know your own gap, that it is stable, and that you notice when it widens.

A practical way to set your own threshold is to calculate the gap for three or four consecutive periods once your counting is solid. That gives you a normal range. Anything outside it deserves investigation; anything inside it is the ordinary cost of running a real kitchen. Over time, as you fix the largest causes, the normal range should narrow, and narrowing it is a far better goal than chasing a number borrowed from another venue.

The actual food cost percentage on its own rewards the wrong behaviour. A kitchen can lower it by pushing cheap dishes, shrinking portions or switching to lower-quality ingredients, and all of those can damage the business more than they help it. The gap does not reward any of that. It only gets smaller when the kitchen wastes less, portions accurately, records honestly and keeps its costings current.

That makes the gap a fairer conversation with the head chef. Instead of "food cost is too high", which invites defensiveness, you can say "we are three points above our recipes, and half of it seems to be waste". That is a problem the kitchen can own and solve.

The comparison is only as frequent as your stock counts. If you count monthly, you can compare monthly. If you count your highest-value lines weekly, you can run a narrower weekly comparison on just those lines, which often catches problems sooner than a full monthly comparison catches anything.

A sensible rhythm for most venues is a full comparison every month, a quick check on the top twenty lines every week, and an immediate investigation whenever the monthly gap moves outside its normal range. Our guide to building a food cost calculator in a spreadsheet shows how to lay out both the recipe costs and the comparison so the monthly check takes minutes rather than hours.

The comparison is powerful, but it is easy to undermine. These are the errors that most often make the gap look bigger or smaller than it is.

The actual figure comes from your stock counts, which happen in the kitchen. The theoretical figure needs two things from your sales: exactly what was sold, dish by dish, and what each dish costs to make.

In servearound, every dish lives once in a library and appears on every menu it is part of, and once its cost is set, the menu report in Analytics shows each dish's units sold, revenue and margin from what actually went out during service. That is the sales half of a theoretical food cost, taken from the same orders the kitchen cooked rather than from a separate export.

Two other parts of the system help with the causes above. The payments report shows void and comp rates and breaks voids and comps down by member of staff, which is where food that left without a sale becomes visible. And when the kitchen marks a dish sold out, it comes off every menu at once, so nobody keeps selling a dish the kitchen can no longer make well.

Not in a real kitchen. Some trim, some spillage and some testing are part of cooking. A gap at or near zero more often means the theoretical figure is built on prices or portions that are too generous.

No. Set a normal range from a few periods of your own data and investigate when the gap moves outside it. Chasing every small change wastes time and wears out the team.

It cannot. Theoretical food cost tells you what should have been used; only a count tells you what was. The comparison needs both.

Start with your twenty best-selling dishes. They usually account for most of your sales, so a theoretical figure built on them, even before the rest are done, is already far more useful than none.

Actual food cost tells you what was used. Theoretical food cost tells you what should have been used. The difference is your food cost variance, and it can almost always be broken into five causes: counting errors, stale prices, waste, over-portioning and food that left without a sale. Check them in that order, put a number to each, and fix the largest first.

Measured this way, the gap becomes the most honest number in the kitchen. It does not reward cheap ingredients or small portions. It only shrinks when the kitchen gets better at what it already does.