Food cost percentage calculator

A restaurant started the month with $14,200 of food, bought $27,600 more and finished with $13,300. It used $28,500 of food to make $86,000 of net food sales, an actual food cost of 33.1%. The menu says what it sold should have cost $25,786, or 30.0%. The gap is $2,714 a month, 3.2 points of sales, and $32,568 a year. Recorded waste and staff meals explain $1,630 of it. The other $1,084 is unexplained.

The calculator computes both numbers from your own figures, shows the variance in dollars, in points and in relative terms, and separates what you can account for from what you cannot. It also breaks food cost down by menu item and shows what one point is worth.

Who reaches for this

Two questions sit behind every food cost figure. What did the kitchen use, and what should it have used? The first is answered by counting inventory and adding up purchases. The second is answered by taking the menu items sold and multiplying each by its recipe cost. The difference is food that left the kitchen without being sold, and the calculator is built around finding out how much of it there is and where it went.

You enter the inventory and purchases for the period, net food sales, and the menu items with their units sold, menu prices and recipe costs. You can also enter the waste you record and the staff meals you serve. The calculator then reports both food cost percentages, the variance three ways, the part of it you have explained, and a table by item. Food cost sits inside prime cost, which the restaurant break-even calculator uses to work out the covers a restaurant needs.

Actual food cost starts with a simple identity: food you had at the start, plus food you bought, minus food you still have at the end, is food you used. With $14,200 of beginning inventory, $27,600 of purchases and $13,300 of ending inventory, the kitchen used $28,500. Divide by net food sales of $86,000 and the actual food cost percentage is 33.1%.

Which sales?

The denominator matters as much as the numerator. Use food sales, not total sales, so that beverage does not dilute the figure. Use net sales, after comps and discounts, because that is the revenue the food produced. Leave out sales tax and tips, which are not revenue. If a POS reports gross sales with discounts as a separate line, subtract them first. A percentage calculated on gross sales will read lower than one on net, and a mismatch between periods makes a trend meaningless.

Dollars, not units

The inventory and purchase figures are in dollars. One trade publication gives the formula as (beginning inventory + purchases − ending inventory) times the product cost. That works when the inventory is counted in units of a single product, and it double-counts if the figures are already in dollars. For a whole kitchen, value the inventory at what you paid, using the most recent invoice cost, and add it up.

For one dish

Food cost for a single item is its recipe cost divided by its menu price. A burger that costs $4.60 in ingredients and sells for $15.50 has a food cost of 29.7%, and it earns $10.90 per plate. The two views are related, since the period figure is a sales-weighted average of the item figures, but the period figure includes everything that happened in the kitchen and the item figures include only what the recipe says.

Theoretical food cost is the cost of what was sold, calculated from recipes. For each item, multiply units sold by recipe cost, and add them up. The units come from the POS product-mix report, and the recipe costs from your costed recipes, using current purchase prices and the yield after trim. The recipe cost calculator costs each ingredient with trim and cooking yields.

Menu sales at full prices are $90,915, and $25,786 of cost against them is 28.4%. That is the food cost of the menu as priced. Net food sales were $86,000, $4,915 lower, because of comps, discounts and voids. Those plates were made and rung, so their food is in theoretical cost, and their revenue is not in net sales. On net sales, theoretical food cost is $25,786 ÷ $86,000 = 30.0%. That is the figure to compare with actual, since both use the same sales.

The mix moves the number

Food cost percentage depends on what people order. In the example, the steak has the highest food cost at 36.6% and the pasta the lowest at 22.9%. If guests shift toward steak, the theoretical figure rises with no change in the kitchen, and if they shift toward pasta it falls. Compare actual with a theoretical cost calculated from the mix that was actually sold, and not from the menu in the abstract.

A high percentage is not always a problem. The steak’s 36.6% is the highest on the menu, and it also earns $18.40 per plate, the most of any item. Judged on dollars per plate, it is the best seller on the menu. Food cost percentage tells you how much of each dollar goes to ingredients, and the dollars per plate tell you how much you keep.

When you cannot cost every item

A small restaurant may not have a product-mix report or costed recipes for the whole menu. Start with the items that matter. In the example, four items, the burger, the steak, the sides and the fish tacos, are 69.7% of theoretical cost. Costing those precisely, and estimating the rest at a typical percentage, gives a theoretical figure that is close enough to find a gap. The calculator warns you when the menu table covers less than your sales, so you know how much of the picture is missing.

Costing a recipe well means weighing the ingredients for one plate, dividing each price by the usable yield, and adding the garnish, oil and trim that people forget. Revisit the costs whenever a major supplier price changes. The number is only as good as the cards behind it, and a stale card produces a variance that looks like a kitchen problem and is really a paperwork one.

The variance is actual minus theoretical, and there are three ways to express it. In dollars it is $28,500 − $25,786 = $2,714. In percentage points of sales it is 33.14% − 29.98% = 3.16 points, which is the same $2,714 divided by $86,000. Relative to theoretical it is $2,714 ÷ $25,786 = 10.5%.

Sources mix these up, and it changes what a benchmark means. One guide describes a 4-point variance between a 27% theoretical and a 31% actual food cost. Another shows an ingredient-level example in which 200 burgers should have used 1,200 ounces of beef and actually used 1,550, a variance of 350 ounces or 29%. The first is points of sales and the second is relative to theoretical. When a guide says that a gap of 2% to 3% is normal and 5% is a problem, ask which it means. A 3-point gap on a 30% theoretical cost is 10% relative, and a 3% relative gap on the same cost is under 1 point.

Use points for the whole restaurant, since they are comparable across weeks and sales levels, and use relative variance for individual ingredients, where the question is how much more you used than the recipe says. Always look at the dollars as well. A gap that is small as a percentage of a large sales figure is still real money.

Read the trend, not the month

A monthly average can hide a change. Suppose the month’s $2,714 variance came in four weeks of $610, $540, $790 and $774, on weekly net food sales of about $21,500. That is 2.8, 2.5, 3.7 and 3.6 points. The month reads 3.2 points, but the last two weeks are a point higher than the first two, and something changed in the second half. A new hire on the line, a supplier switch or a missed count could each do it.

Weekly figures are noisier, since one late delivery can swing a week, so look at a rolling average of three or four weeks and at the direction. A steady drift up is a reason to look. A single spike that reverses the next week is usually a counting or timing effect.

A variance is a total, and the value is in breaking it into causes. Start with what you already record. Kitchen waste logged at cost was $1,150 and staff meals were $480 in the example, so $1,630 of the $2,714 is accounted for. The $1,084 left, 1.26 points of sales, is the part that needs an investigation.

The workbook’s investigation sheet keeps a running list. In the example, waste, staff meals, over-portioning on the steak and short deliveries come to $2,120, or 78% of the variance, and $594 remains unexplained. The exercise matters more than the exact figures: each cause has an owner and a fix, and the unexplained part shrinks as you find it.

Work from the largest dollars. The items with the highest cost of goods, the burger at $5,290 and the steak at $5,088 in the example, carry 40% of theoretical food cost between them. A small percentage error on those moves more money than a large error on the salad. Pull the POS report, sort by dollars of cost, and check the top items first.

The actual figure is only as good as the counts behind it. Count on the same day and at the same time each period, ideally after close, so that deliveries and service do not move stock during the count. Use the same count sheet, in the same storage order, and value everything at the latest invoice cost. Include every location: walk-in, dry storage, freezer, line and bar-side food.

Small counting errors are not small in money. A 1% miscount of the $13,300 ending inventory is $133, or 0.15 points of sales. That sounds minor until you see it as a swing in the variance that has nothing to do with the kitchen. Two consecutive periods with opposite errors can show a “bad” month followed by a “good” one. Weekly counts, and daily counts of proteins and other expensive items, make each period’s error smaller relative to the whole.

Cut-off and invoices

Match purchases to the period. An invoice for food delivered on the last day of the month belongs in that month, and one for the first day of the next belongs in the next. Late invoices and credits should be booked to the right period, or the actual cost is distorted. Transfers between locations, and food used for catering or events, need their own lines.

A weekly routine

A workable routine takes about an hour. Count after close on the same day each week, enter the inventory, add the week’s invoices, and pull the POS product-mix and net food sales for the same seven days. Run the calculator, note the variance in dollars and points, and list the top five items by cost. Compare each with the last few weeks, and investigate any that moved. Share the result with the kitchen. A number that only the owner sees does not change behavior.

Beverage, catering and transfers

Beverage deserves the same calculation with its own inventory, purchases and sales, because its losses come from pouring, comps and spillage and not from trim and yield. Catering and events need their own lines, or the food that left the building without a rung sale will appear as a variance. Transfers between locations, and food given away for promotions, should be recorded so that they can be taken out of the gap and not be mistaken for waste.

Guides give targets. One says most restaurants aim for a food cost of 28% to 35%. Another puts the industry target at 28% to 32%, and adds that the number to watch is prime cost, food plus labor, staying under 60% to 65%. I could not trace these to a primary source, and they describe a wide field of concepts.

A target has to fit the concept. A steakhouse serving prime cuts, a seafood restaurant and a pasta place have very different plate costs and price points, and the food cost that supports a healthy profit differs with them. What matters is what remains after labor, rent and everything else. Food cost also trades against other lines: a higher food cost with lower labor can be a better restaurant than the reverse.

Use published targets as a rough check, and set your own from your menu, your prices and your other costs. The restaurant break-even calculator shows what a given food and labor cost implies for the covers you need. Your own trend is more informative than any benchmark: a food cost that has crept up two points over a year is a signal whatever the industry average is.

One point is 1% of net food sales. On $86,000 a month, that is $860 a month or $10,320 a year. Against a 30% target, the example’s 33.1% actual food cost is 3.1 points over, which is $2,700 a month. Turning points into dollars makes it easier to decide how much effort a fix deserves.

None of these is dramatic on its own, and together they are a large share of a restaurant’s profit. The point is to work in order of dollars, and to count a fix as done only when the next period’s variance shows it.

The formula runs backward from a target. Menu price = recipe cost ÷ target food cost percentage. To run the burger at 30%, its $4.60 cost needs a price of $4.60 ÷ 0.30 = $15.33. The steak at $10.60 needs $35.33, against the $29.00 on the menu, and that gap is why it runs at 36.6%. The salad at $2.80 needs $9.33, and at $11.50 it is priced well above that.

Do not price every plate to the same percentage. Guests judge prices against alternatives and expectations, and a steak at $35 may sell much less than one at $29. Some items are priced for the traffic they bring. The check is the mix: the weighted food cost of what you sell should land near the target, and the dollars per plate should support labor and overhead. The menu price calculator prices a dish four ways and tests a price rise. The markup vs. margin calculator converts between a cost multiple and a margin, and the contribution margin calculator shows what each plate adds after variable costs.

See what your food and labor costs mean for covers with the restaurant break-even calculator, or work out plate economics with the cost per unit calculator.