What your dish really costs

Everything in this course rests on one sheet of paper per dish: every ingredient, its weight, and what that weight costs. Most kitchens do not have it. They have a price that was worked out once when the menu was written, a rough idea of the food cost percentage, and a chef who knows the recipe by heart — which is exactly how a dish can cost thirty percent more than the owner thinks for two years without anybody lying to anybody.

📋 The standard recipe, in money

Here is the burger from the last course opened up. It sells for $16 and the menu says it costs $7; this is where the $7 actually is.

Two things become possible the moment this sheet exists. You can see which line is worth attacking — the beef is a third of the plate and the fries cost more than the cheese, which surprises most people — and you can recost the dish in ten seconds when a supplier price moves, instead of guessing. At $7 on a $16 price this burger runs at nearly 44% food cost, against 28% for the pasta in the same kitchen. That is not a scandal: it is what a workhorse looks like, and it is why one peso here matters more than five pesos on a dish nobody orders.

Do this for your ten best-selling dishes and you have covered most of what leaves your kitchen. Doing it for all forty is a project that never finishes; doing it for ten is an afternoon.

⚖️ What you pay per kilo is not what it costs you

The beef line hides the single most common costing error in restaurants. The invoice says $12.00 a kilo. The kitchen buys ten kilos, trims the fat and the sinew, and nine kilos reach the grill. So the meat that goes on a plate cost $13.33 a kilo, not $12.00, and a 180 gram portion costs $2.40 rather than $2.16. Every costing built on the invoice price is understated by exactly the amount that ends up in the bin.

This is called yield, and it applies to nearly everything you buy whole: lettuce that loses its outer leaves, a whole fish that comes back at half its buying weight as fillets, a case of tomatoes with two that are gone. The arithmetic is always the same — divide the price you paid by the fraction that survives — and doing it once per major ingredient is enough, because yields barely change while suppliers stay the same.

It also quietly answers a question that gets argued about with no evidence: whether to buy something prepared or do it yourself. Pre-trimmed beef at $13.50 a kilo looks more expensive than $12.00 and is not, because your own trimming produces meat at $13.33 plus the twenty minutes somebody spent doing it. Once yield is in the comparison, that decision stops being a matter of pride and becomes a number — and it can go either way depending on what your kitchen actually costs per hour, which is the labour course's territory.

📏 The twenty grams nobody weighs

Portions drift in one direction. A cook who is busy, or generous, or new, puts a little more on, nobody complains, and the new amount becomes normal. Six months later the burger carries 210 grams and the recipe still says 180, and the food cost percentage has moved a point and a half for a reason no report can show you.

The fix is unglamorous: a scale on the line for the expensive items, portioned tools for everything else — a scoop for fries, a ladle for sauce, a pre-weighed patty rather than one shaped by hand — and a plate photo where new staff can see it. What matters is that the correct portion is the easy portion. If getting it right requires care, it will be right on quiet Tuesdays and wrong every Friday, which is when you sell most of them.

And it is worth saying to the kitchen out loud what the exercise is for, because a scale arriving unannounced reads as an accusation. The message that works is the true one: the portion is the promise the customer already paid for, and the point is that everybody gets the same burger, not that anybody gets a smaller one.

🎯 Finding the peso

The previous course asked for one peso of contribution on this dish. With the sheet in front of you it is not hard, and no single line of it is something a customer could detect.

That is the dish at $6.00 instead of $7.00, a food cost of 37.5% instead of nearly 44%, and $500 a month — $6,000 a year — on one menu line, with the same beef, the same weight on the plate and the same bun. Notice how it was assembled: no item on that list is worth more than twenty-seven centavos, and nobody would have bothered chasing any of them individually. They only became worth doing because a sheet of paper made them addable.

Redo the costing whenever a supplier price moves more than a little, whenever the recipe changes — including the changes nobody announces, like a different cheese because the usual one was out — and otherwise once a quarter for your top ten. Two other courses depend on this number being current: the contribution figures that sort your menu into stars and workhorses, and the food cost percentage your budget is built on. A stale plate cost quietly makes both of those wrong.