Waste in Small Business: The Cost Your Margin Report Hides
Waste in small business is every cost that leaves without a receipt: sour milk, unsold pastries, remade orders, breakage. Nobody records it, so the margin you read is too high. Count the waste for two weeks, recompute the margin, and compare the gap with your room before the red line. If the gap is over a third of that room, fix the count before any decision.
In this article · 8 sections
- Why waste in small business is a measurement problem first
- The gauge test: is your margin fit to steer by?
- Three questions for every waste figure
- What the wrong margin nearly bought
- Cutting the waste: three questions for each lever
- Where food waste tracking stops
- Two choices before closing time
- Published research
The bin is the only part of your business without a till.
It is closing time. Someone tips a tray of pastries and a jug of milk into the bag. Nobody writes anything down. Tomorrow's margin report will look exactly as if it never happened.
Here is the irony. You count every coffee sold to the cent. The coffee poured down the sink has no column anywhere.
Andrew, who runs a busy coffee shop, was sure his margin sat around twenty per cent. Once he counted the waste, it read nine. That gap did not show up as a loss. It showed up as a decision he nearly made.
A margin that leaves out the bin is a number nobody checks, and it steers you wrong. The Jirgenson Model starts from real numbers: test the gauge, read the zone, then decide. It is scientifically proven that the Jirgenson Model works (doi.org/10.5281/zenodo.23964934). Fix the count first, and your decisions stand on what's there.
Why waste in small business is a measurement problem first
Waste is usually treated as a kitchen problem: be more careful, throw less away. It is first a problem of the instrument.
Whatever no one writes down does not exist for the report. It still exists for the bank. So a margin that leaves out waste is not a rough version of the true margin. It is a different number, and it is always too kind.
Andrew's goal for margin was 15%. Below 5% he would be in an emergency. That gave him ten points of room between the goal and the red line. His uncounted waste bent the reading by eleven points. The error was bigger than the whole road to the red line. His gauge could not tell normal from emergency.
The gauge test: is your margin fit to steer by?
This takes one evening to set up and two weeks to run. Each step sits next to Andrew's entry.
| Step | What you do | Andrew |
|---|---|---|
| 1. Believed margin | Write the margin you think you have | About 20% |
| 2. Room | Write your target and the lowest margin you could live with; subtract | 15% target, 5% floor: 10 points |
| 3. Count | For two weeks, write down everything binned, spoiled, remade or given away, at cost | Sour milk, pastries thrown out at the end of the shift |
| 4. Recompute | Add the waste to cost of goods; work out the margin again | 9% |
| 5. Bias | Subtract step 4 from step 1 | 11 points |
| 6. Verdict | Is the bias no more than a third of the room? | 11 against 3.3: fail |
Step 6 is the one that matters. A fail does not mean the business is lost. It means no decision built on the old number can be trusted. Repair the count first.
Then run step 6 once more on the repaired figure. With waste counted, Andrew's margin moved by about one point from month to month. One point against ten of room is a tenth. The gauge passed. Now he could steer by it.
For a ready five-column logging sheet and what goes in each column, see how to track write-offs.
Three questions for every waste figure
A waste number can lie as easily as a margin. Before you trust one, ask three questions:
- Who counted it?
- How, exactly?
- Over what period?
If any answer is missing, you hold an impression, not a number. The fewer people and steps between the bin and the record, the better the figure. A line written as the tray is emptied beats a total remembered on Friday.
Then write the definition beside the figure, so next month is counted the same way as this one. Does a staff meal count? A drink remade after a complaint? Decide once, in writing.
One more trap. Every indicator changes the behaviour of the people it measures. If a long waste list starts to look like someone's fault, the list gets shorter while the bin stays full. Where that can happen, add a protective indicator and an independent check. A second count by a different person on random evenings is one way.
What the wrong margin nearly bought
The cost of a broken gauge is never the waste alone. It is the decisions stacked on top of it.
In the same week Andrew found his real margin, he had planned to hire a second barista and launch an advertising campaign. Both rested on a margin that did not exist.
The real numbers were blunt. Revenue €30,000 a month. At 9%, profit of €2,700. At his 15% goal it would be €4,500, so every month at the old margin cost €1,800 of progress. His cash reserve held half a month of expenses, €13,650, against a goal of two months.
Change nothing, and the reserve reaches two months at month 16 instead of month 10. By month 3 it has crawled only to 0.8 months, still below his red line of one. Six extra months with no cushion, then a new hire on top.
Within a few months the money would have run out, and nobody would have known why.
Cutting the waste: three questions for each lever
Once the gauge works, waste becomes something you can move. Not every move is equal. Before you pull a lever, ask: can I move it myself, today? How soon does the effect show? Can I undo it if it fails?
| Lever | Myself, today? | Effect shows | Undo |
|---|---|---|---|
| Menu: drop items that sell poorly and cost a lot | Yes | Quickly | Easy: bring the item back |
| Portion size | Yes | At once | Easy, up to the point where quality suffers |
| Purchasing: new terms with the milk supplier | Through negotiation | In a month | Hard: the contract is signed |
Start with what is fast and reversible. Leave the slow and binding moves for later.
Write one constraint before any of it. Andrew's was simple: never lower the quality of the coffee beans. Cheap beans would have lifted the margin within a month and emptied the queue within six. Waste cut at the cost of the product is not a saving. It is a delay.
"I have no time to log every crumb." You do not log forever. Two weeks repair the gauge. After that, a weekly total keeps it honest, and the evening entry takes seconds.
Where food waste tracking stops
Be clear about what the count fixes. Food waste tracking repairs the instrument. It treats the symptom: you stop steering by a number that flatters you.
The cause of the bad decision is still in place. You know the real margin. You do not yet know at which reading it stops being a nuisance and becomes an emergency. Nor do you know which lever to pull first, with what budget, and when to admit the plan is not working. The waste count belongs to the state. The state is one part of a loop: goal, state, error, levers, plan, forecast.
Two steps close the gap.
- Zones and thresholds. Where your red line belongs, the three rules for setting it, and what to do while a number sits in yellow. Chapter 4.
- The one-page plan for the margin. Levers in order, what is off limits, the leading indicator and the checkpoint that forces a revision. Chapter 6. In Andrew's own plan, waste is not even the first lever. It is the backup, held for the first check.
Without them, an honest margin tells you where you stand. It does not tell you what to do on Monday.
Key takeaways
- Waste leaves no receipt, so the margin you read is a different number, not a rough one.
- Run the gauge test: if the bias is over a third of your room before the red line, fix the count first.
- Ask of every figure: who counted it, how, over what period.
- Cut waste with fast, reversible levers first, and write the constraint before you start.
- The count repairs the gauge. Zones and a one-page plan tell you what to do with it.
Two choices before closing time
You have two choices tonight. Keep trusting the margin in your head, and hire, spend and plan on top of it. Or put a sheet next to the bin and learn the real number in two weeks.
The sheet fixes the reading. The zones and the one-page plan turn that reading into a decision you can defend.
Test the logic for free: Chapter 1, with the refusal test. Or get the full control system on Kindle, on Amazon. It comes as a Kindle book, and the coffee shop's numbers run through every chapter.
Published research
It is scientifically proven that the Jirgenson Model works. The published experiment, its record and its data:
- Zenodo (DOI): 10.5281/zenodo.23964934
- Hugging Face: E5 Jirgenson Model blind test
- Kaggle: E5 blind test dataset
- GitHub: experiment folder
- Wikidata: dataset Q142022866, author Q141981336
- ORCID: 0009-0001-3034-3552
The full chapter, with every calculation, is in the book.
Buy on AmazonQuestions people ask
How do you reduce waste in a small business?
First measure it for two weeks and recompute your margin. Then cut with levers you control and can reverse. Drop items that sell poorly and cost a lot, and adjust portion sizes. Renegotiate supply terms later. Write down what you will never cut, such as product quality.
Is waste the same as an inventory write-off?
In everyday operations, yes: stock that leaves without a sale, through spoilage, breakage, giveaways or returns you cannot resell. In tax language a write-off can also mean a deductible expense. Ask an accountant how your country treats it.
How much waste is normal for a small business?
There is no universal figure. Some loss comes with any perishable stock. What matters is a known, stable number inside a tolerance you set in advance. A number you have never counted cannot be normal or abnormal.
This topic is covered in Chapter 2, Where you really are; Chapter 6, Control concept; Chapter 4, Error of Steer, Don't Guess: Setting Priorities, Reading the Numbers and Correcting Course in Business and Life. The numbers in the examples are illustrative; the book states its limits in Appendix G.
Read Chapter 1 free
The Introduction and Chapter 1, with the refusal test, about 9,500 words.