Every kitchen throws away money, and almost none of them know how much. It leaves in small amounts, in a bin nobody looks into, spread across a hundred moments that each felt reasonable — a tray of prep that did not get used, a steak that came back, a container at the back of the shelf. Nothing about it appears in a report. A week of writing it down usually finds more money than a month of negotiating with suppliers, which is why this is the topic with the best return in the course.
🗑️ A month, written down
This is what a real month looked like in the example kitchen once somebody started keeping a log.
| Where it went | Cost | What it usually means |
| Prep and trim | $260 | Yield: normal up to a point, then a technique or a spec problem |
| Overproduction | $180 | Prepping for a Friday that turned out to be a Tuesday |
| Remakes and returns | $150 | Wrong order, wrong cooking, dropped plate |
| Spoiled and out of date | $210 | Too much stock, or first in, first out not happening |
| Comps and mistakes given away | $200 | Service recovery — sometimes money well spent, always worth knowing |
| Total | $1,000 | 6.25% of the food bill, 2% of sales |
A thousand pesos a month, in a restaurant whose entire budgeted profit is $46,200 a year. Roughly half of it is recoverable without anybody working harder — that is $6,000 a year — and the other half is the cost of running a kitchen, which is fine, but you want to be the one who decided which half is which.
Read the causes column and notice that four of the five rows point at something covered elsewhere in this course. Prep waste is a yield and specification question from topic 3. Spoilage is stock levels and rotation from topic 4. Overproduction is a forecasting question, and the sales analytics course is where next week's number comes from. Only remakes are their own thing. Waste is rarely a waste problem: it is the place where every other problem becomes visible.
📔 A week of logging, four times a year
You do not need a permanent system. You need a bucket, a clipboard and one honest week per quarter.
Put a container by the pass and a sheet above it with three columns: what it was, roughly how much, and why. Everything that does not reach a customer goes in — trim, remakes, the tray that turned, the portion that came back. Price it at the end of the week using the recipe sheets from topic 2, which is the only reason the numbers above exist at all. Then read it once, out loud, with the kitchen.
The whole thing depends on one rule, and it is not about accounting: nobody gets in trouble for what is written on the sheet. A log that produces consequences produces beautiful, fictional numbers within two days. What it should produce instead is a couple of decisions — prep less of that on Mondays, change the order of the shelves, hold a spare of the thing we drop — and those decisions are worth more than the precision would have been.
🥄 Portions in the middle of service
Topic 2 put a scale on the line and found $0.27 a burger in the beef. The rest of the portion money is in everything that is served by hand, and it goes the same way: upwards, quietly, whenever the kitchen is busy.
- Give every portion a tool. A scoop for the fries, a ladle for the sauce, a jug for the dressing, pre-weighed patties. The fries alone are $0.19 a plate when 150 grams becomes 180.
- Portion in advance where you can. Sauces, garnishes and pickles portioned before service are consistent by construction, and they take pressure off the line at exactly the moment when accuracy dies.
- Photograph the plate. One picture per dish where the line can see it does more for consistency than any amount of explanation, and it makes new staff productive faster.
- Check at the pass, not on the sheet. A weekly plate-up of two or three dishes, weighed, takes five minutes and tells you whether the recipe and the reality still agree.
- Watch the free things. Bread, chips, refills, the extra sauce. They are given away with no ceremony and they are food you paid for; the answer is rarely to stop, it is to portion them.
The reason to keep saying this out loud in the kitchen is that portioning sounds like meanness and is not. An over-portioned dish is an inconsistent dish, and a customer who gets a generous plate once and an ordinary one next time has been let down by the generosity, not by the standard.
🚪 The part nobody likes to name
Some of the gap between what your recipes say you should have used and what the count says you did use is not waste. It is food and drink leaving the building. Nobody enjoys this section, and skipping it is how small restaurants lose serious money for years.
Treat it as a systems question rather than a character question, because the honest truth is that opportunity does most of the work. Everything that leaves the kitchen gets rung in, including staff meals and comps, recorded as what they are. Comps and voids require a manager, and the report gets read — a pattern by person, terminal or hour is information long before it is an accusation. Deliveries are received by someone who is not the person who ordered them, if you are big enough to split that. High-value stock is counted weekly. Staff meals are a defined thing with a time and a menu, not an open kitchen.
Most of what these controls catch is not theft at all. It is a dish that never got rung in on a busy Friday, drinks poured heavy, a manager's generosity nobody was tracking. The purpose is not to catch people; it is to make the invisible visible, which — as topic 1 promised — turns out to be the whole of cost control.