AP and EP — the bridge most operators don’t walk
The two letters that decide whether your dish prices are right are AP and EP. AP is “as purchased” — what you paid the vendor for the whole chicken, the case of tomatoes, the tin of saffron. EP is “edible portion” — what made it onto the plate after the cook trimmed, deboned, peeled, and reduced.
The bridge between them is yield, and yield is rarely 100%. A whole chicken is 70–75% bone-and-skin-out yield. A box of romaine is 88–92% after coring and washing. A pound of saffron is 100% yield because nobody throws saffron away. Every dish’s real cost is its AP cost divided by its yield percent — if you bought a chicken at $3.20 a pound at 72% yield, the chicken on the plate cost you $4.44 a pound, not $3.20.
Most operators price off AP and discover six months later that they have been losing margin to the bone bin. The recipe cost card forces you across the bridge: the line cost of every ingredient is the AP cost divided by yield, the way it actually leaves the walk-in.
Food-cost target is a policy decision, not a math one
The sheet asks for a target food-cost percent. Most operators type 30 because they read it somewhere, or because their accountant told them to, or because the previous chef said so. None of those are wrong, but none of them are answers either — the target is a decision your concept makes, not a number the formula returns.
A high-volume coffee bar can run 25% food cost because the labor and rent ratios cover the rest of the P&L. A neighborhood bistro with table service runs closer to 30%. A tasting-menu place with a small dining room and three cooks runs 38%, sometimes more, because their labor is concentrated and their average check carries the model. Each of those numbers is right for its restaurant and wrong for somebody else’s.
The right way to set the target is to back it out from the prime cost band you need. If your prime cost target is 60% and your concept supports a 33% labor cost, your food-cost target is 27%. If your prime cost target is 65% and your labor runs 35%, your food cost can be 30%. The food-cost target is the residual, not the input. Type the number that lets the rest of your model survive.
The four lies in dish-level costing
Recipe cost cards quietly produce wrong numbers in four common ways. None are intentional; all are the cost-card version of looking at your bank balance instead of your P&L.
Yields from a textbook. The 72% chicken yield in your old culinary textbook was for a 4-pound chicken in 1985. Your vendor’s chicken is a 3.6-pound bird with more breast and less leg, and your line cook hasn’t been re-trained on the new butchering pattern, so your real yield is 67%. The fix is one cutting test on a slow Tuesday: take five birds from this week’s delivery, weigh them whole, butcher them to your spec, weigh the EP. Divide. The number in the slider is the average; that is your real yield until your bird changes.
Plate weight that drifts. The dish was speced at six ounces of protein. The line cook plates eight, because the customer pays attention to portion. Your card shows six; your kitchen serves eight. The recipe cost card returns a price for a dish that no longer exists, and the actual food cost on the plate is 33% higher than the sheet says. Either reprice for the eight-ounce reality or train the line back to the six-ounce spec; do not pretend the card is current when it isn’t.
The forgotten ten percent. Salt, pepper, oil, butter, garnish, sauce drizzle. None of them line-item on the recipe cost card because each is a tiny fraction of plate cost — but a kitchen runs ten ingredients on every plate that don’t appear on any card. Industry rule of thumb is to add 8–12% to total plate cost as “Q-factor” for these. The honest plate cost is your card’s number times 1.10.
Vendor price that floats. The card was right when you wrote it. Six months later your produce vendor has moved their tomato price up nineteen percent and you didn’t notice because the change came in two cents at a time across eight invoices. The vendor-price-drift slider is the slider that surfaces this — if you slide it +10% across every ingredient, the price the sheet returns is the price you should be charging in three months when the drift compounds. Run that scenario quarterly.
The charm-price illusion
The sheet returns a target price — the math — and a charm price — the rounded-to-.95 number you actually print on the menu. The charm price exists because customers read “$21.95” as “twenty-one dollars something” and read “$22.00” as “twenty-two dollars.” That gap is real and worth a few cents in average check.
What the charm price is not is a license to round down. If the math says $21.73 and the charm rounds you to $21.95, you gained twenty-two cents. If the math says $22.04 and you charm down to $21.95, you lost nine cents on every plate, and over a thousand plates a month that is ninety dollars of margin you handed back. The charm round goes up to the next .95, not down to the previous one. The sheet does that for you; do not override it.
When to actually raise a price
The hardest decision on this sheet is when to push a menu price. Most operators wait too long, because they think a price increase is a customer-relationship event when it is actually a math event. The right cadence is to run the card every quarter and let the band tell you.
If the card’s charm price is more than $1.00 above the menu price for any dish that moves volume, the menu price moves at the next reprint. If the gap is over $2.00, the menu price moves at the next reasonable opportunity (a season change, a menu refresh, a Tuesday). Do not move every dish at once — move the three or four that have drifted most, and let the rest hold for the next quarterly cycle. Customers notice menus that change once a quarter; they don’t notice menus that move three dishes at a time.
The sheet does not make the decision for you, but it removes the guesswork. The yield slider tells you what the kitchen is actually doing. The vendor slider tells you what the market is doing. The target food cost tells you what your concept needs. The charm price tells you what to print. Everything else is a judgment call about your customers, your moment, and your nerve.
Open the sheet: Recipe Cost Card →
Run a cutting test, slide the yield, watch the price. Your numbers never leave the page.