Food cost percentage is the first number anyone learns in this business and the one most often calculated wrong. The formula is simple — plate cost divided by menu price — but the plate cost almost always comes out too low, because of two things nobody puts on the spreadsheet: yield and Q factor.
Here is the whole calculation on one dish, with arithmetic you can check, then what to do with the answer.
You do not cook what you buy. You cook what survives trimming, boning, peeling and shrink. Costing a dish at the invoice price understates it every time.
A whole chicken, 3.5 lb at $2.20/lb, costs $7.70. After bones and trim you have roughly 1.9 lb of usable meat. So your real cost per usable pound is:
$7.70 / 1.9 = $4.05 per usable pound
Nearly double the invoice price. For vegetables the convention is a yield percentage instead. Potatoes at $0.94/lb with an 80% yield cost $0.94 / 0.80 = $1.18 per usable pound. Green beans at $2.40/lb with an 88% yield cost $2.40 / 0.88 = $2.73.
You do not need published yield tables. Weigh one case before prep and the usable product after, twice, and use your own number — it will be more accurate than any table, because it reflects your suppliers and your cooks.
A roast chicken plate, portioned as it actually goes out:
Plate cost: $3.14
The ounces-to-pounds conversion is where careless cost cards fall apart. Six ounces is 6/16 of a pound, or 0.375 — not 0.6.
Q factor is everything that reaches the table without being on the recipe: bread and butter, the condiments, the lemon wedge, the free refill of tap water with ice. It is small per cover and large per year.
Add it as a flat amount per plate. If you serve bread and butter to every table, $0.30 a cover is a defensible starting figure — measure yours by dividing a month of bread, butter and condiment invoices by the month's cover count.
True plate cost: $3.14 + $0.30 = $3.44
At a $18.00 menu price:
$3.44 / $18.00 = 0.1911 = 19.1% food cost
And the reverse formula, for pricing a new dish to a target: price = plate cost / target food cost percentage. At a 30% target, $3.44 / 0.30 = $11.47.
Which is where a lot of operators do something expensive: they see 19.1% against a 30% target and conclude the chicken is underpriced at $18 — or worse, they price the next dish at $11.47 because the formula said so.
Percentages do not go in the bank. Dollars do. The number that pays your rent is contribution margin: menu price minus plate cost — the figure the menu engineering matrix is built on.
Chicken plate: $18.00 - $3.44 = $14.56 contribution at 19.1% food cost.
Now a side salad: $9.00 menu price, $1.10 plate cost. That is $1.10 / $9.00 = 12.2% food cost — much better-looking than the chicken. But the contribution is $9.00 - $1.10 = $7.90.
The salad wins on percentage and loses by $6.66 per sale (14.56 - 7.90). Sell 40 of them instead of 40 chicken plates on a Friday and you are $266.40 lighter, with a healthier-looking food cost report. This is exactly how a restaurant hits its food cost target and still cannot make payroll.
Three real jobs:
You do not have to cost 60 items to get value from this. Pull your top 10 sellers by volume — they are typically most of your food spend — and cost those properly, with yield and Q factor. That is an afternoon, and it tells you more than a full-menu exercise you abandon halfway through.
Then keep one habit: every time you write a new dish, fill in the cost card before you decide the price. Pricing first and costing later is how dishes end up on the menu that were never going to work.
MenuGenie runs this kind of analysis across your whole menu from a photo or PDF of it, free — upload yours at menu-genie.com.
Upload your menu and get the same analysis across every item — free, in about 30 seconds.
Analyze my menu