“We throw too much away” is the most misleading sentence in a kitchen. Loss takes five forms with different causes and different fixes, and the most expensive one never reaches the bin: it is the over-generous portion, served and paid for, invisible to everyone. Here is how to separate the five, and the fix specific to each.
In short
- The biggest loss is served, not binned. Twenty grams too much, a hundred times a week.
- The second walks in the delivery door unchecked: you pay for what you did not receive.
- Breakage and spoilage are visible, so they get overestimated: they are rarely the biggest line.
- Theft exists and it is small, but it only shows once everything else is clean.
- Nothing gets fixed without measurement. Record the loss before trying to reduce it.
The five forms of loss
Treated together they stay unsolvable. Separated, four of the five are fixed with a habit rather than a purchase.
| The form | Where it happens | The fix |
|---|---|---|
| Over-portioning | At plating, continuously, without anyone noticing. | Weigh two random portions per service, mid-rush. |
| Unchecked goods-in | At delivery: stated weight, real weight, quality. | Weigh and tick against the purchase order, before signing. |
| Spoilage | In the store and the cold room, on dated products. | First in first out, and a weekly date check. |
| Breakage and error | During service: returned dish, broken plate, wrong order. | A recorded stock-out, with a reason, every time. |
| Staff and giveaways | Staff meals, comped to a customer, taken home. | A dedicated reason: comped, staff, breakage. |
Note that three of those five lines cost nothing to fix. They need a daily habit rather than a budget, which is exactly why they get put off.
Why over-portioning dominates
It has three properties that make it formidable, and none of them is shared by the other forms.
- It is invisible plate by plate: Nobody sees twenty grams. The customer is happy, so is the cook, and nothing in the day flags anything. It only becomes visible at stock take, a month later, as a variance that will be blamed on something else.
- It lands on your most expensive products: Nobody over-serves rice; they over-serve meat and fish. Loss therefore follows cost rather than volume, which makes it disproportionate to how it feels.
- It is fixed without buying anything: A scale and two checks a service. It is the best effort-to-recovery ratio on the whole list, and it does not decay while the checking continues.
The link to the menu is direct: a drifting portion makes the food cost in your https://blog.bellocommerce.com/en/restaurant-food-cost/ wrong, and you then take pricing decisions from a card that no longer matches what leaves the kitchen.
The method, over six weeks
One thing a week. That is the rhythm that lasts, not the grand Sunday-night plan.
- Week 1: measure. Create three stock-out reasons, breakage, comped and staff, and record everything. Correct nothing yet.
- Week 2: goods-in. Weigh every delivery against the purchase order before signing, and note the gaps.
- Week 3: portions. Two random weigh-ins a service, on your three most expensive products.
- Week 4: dates. A weekly check of dated products, old at the front, new behind.
- Week 5: a rolling count. One shelf or one cold room a week, not the full stock take.
- Week 6: read it. Compare the recorded loss against the stock-take variance. The difference between the two is what you are not yet measuring.
Step 6 is the only one that really counts. Two hundred dirhams of recorded loss against a thousand dirhams of stock-take variance tells you there are eight hundred dirhams of unmeasured causes, and that is where to look, not anywhere else.
Do not announce a loss figure to your team in week one
Until you are recording, you do not have a figure, you have an impression. And an impression announced as a figure does exactly what suspicion does to cash variance: it stops the reporting. Measure quietly for a week first, then show the real numbers, and the team will declare breakage instead of hiding it.
The Moroccan context
Market buying, first. Some fresh produce is bought with no usable document, at the weight the seller states, and without weighing on arrival nobody will ever know whether the five kilos paid for were five kilos. It is the easiest line to recover because it needs nothing but a scale at the delivery door.
Then heat, and Ramadan. Summer shortens the life of fresh produce well before the printed date, and Ramadan concentrates demand into a few hours with volumes prepared in advance. Both mechanically increase spoilage, and both are predictable: last year’s history, product by product, is what lets you order precisely rather than generously.
What recovers most, per unit of effort
Best ratio first.
- Weigh two portions a service. Almost no effort, a lot recovered.
- Weigh deliveries before signing. A scale and five minutes.
- Record breakage with a reason. Free, and it is what makes the rest readable.
- First in, first out. A storage habit, not an investment.
- Rolling counts. One quiet hour a week, without closing.
Mistakes to avoid
- Treating all loss as waste. The biggest share is served, not thrown away.
- Fixing before measuring. You will never know whether it worked.
- Signing a delivery note without weighing. That is agreeing to pay for a weight you never checked.
- One stock take a year. A ten-month-old variance cannot be explained.
- Starting by suspecting the team. Theft is the smallest of the five and the last to diagnose.
Frequently asked questions
What is the main cause of stock loss in a restaurant?
Over-portioning: portions slightly more generous than the recipe card, served and paid for. It never reaches the bin, it lands on the most expensive products, and it only appears at stock take. It is also the cheapest to fix, with a scale and two checks a service.
How do I measure loss without complicated software?
Create three stock-out reasons, breakage, comped and staff, and record every one for a week without correcting anything. Then compare that total against the variance found at the count. The difference tells you what you are not yet measuring, and that is where the money is.
Should I weigh deliveries?
Yes, and it is the most profitable check after portions. Without weighing at goods-in you pay the stated weight rather than the received weight, and the gap shows up nowhere afterwards. A scale at the delivery door and five minutes per delivery is enough.
What loss rate is normal in catering?
It varies too much by cuisine, menu and season for a general figure to mean anything, and we publish none. Measure yours for a month, then track how it moves: the trend of your own number is what is useful, not a comparison against an average.
Is theft a significant cause?
It is the smallest of the five in most places, and above all the last that can be diagnosed. While portions, goods-in, dates and breakage go unmeasured, a variance proves nothing. Once all of that is clean, a regular variance in one direction becomes serious information.
What to take away
Stop talking about waste and start talking about five distinct lines. Measure for a week without correcting, weigh deliveries, weigh two portions a service, record breakage with a reason, and count one shelf a week. After six weeks you will know which of the five costs you most, and in most kitchens it will not be the one you and the team would have named on day one.
Record the loss where it happens
BelloPOS tracks stock, records stock-outs with their reason and keeps the purchase prices entered at goods-in, so the gap between theoretical and real becomes readable. It all runs offline, and the Lite licence is free for life.
Read next
Other practical guides on the same subject: