Cutting Waste in a Coffee Shop — Where It Actually Hides
- Café waste is expensive mainly because it is invisible — it happens in seconds, in small quantities, and nobody prices it at the moment it happens.
- It hides in three places: the milk jug, the grinder and dial-in routine, and the display case at closing. Each has a different fix, and each is a habit rather than a purchase.
- Measure with a reason attached. Nine litres of waste means nothing; five point eight litres of over-poured milk plus a dropped jug plus expired stock is a to-do list.
- In the worked example here, a two-week count found 9.5 litres of wasted milk out of 210 used — about 4.5 per cent — and one habit change brought the next fortnight down to 3.1 litres.
- The display case needs a written day-end policy — markdown, staff, or donation — decided when you are calm, not at closing time by whoever is tiredest.
- Stock software counts quantities, not money: litres and pieces recorded as waste, kept separate from sales, with the costing half left to your spreadsheet.
Waste in a coffee shop rarely looks like waste. It looks like a barista being careful — a fresh jug for the next order, a re-tuned grind, a display that deserves to look full at four in the afternoon. Each event costs less than the attention it would take to prevent, which is exactly why the total goes unmeasured for years. The cafés that cut it do not become scolds about milk. They make the waste visible for two weeks, let the numbers choose the two habits worth changing, and then stop thinking about it.
Where Waste Actually Happens
| Place | What it looks like | Why it persists | First move |
|---|---|---|---|
| Milk | Over-poured jugs, steamed milk poured away, expired cartons | Every event is small and happens mid-rush | Jug sizes matched to drink counts; pour for the round, not the hour |
| Beans and dial-in | Shots pulled to re-tune the grinder, stale purge, stale ground at day's end | Read as quality practice rather than consumption | Dial in early and once; weigh the week's tuning waste once |
| Display case | Unsold slices, day-old pastries, fill-level pride | No written end-of-day policy | Bake to the sales curve; decide the day-end rule in advance |
Write It Down at the Moment
The mechanism is deliberately crude: a record made when the waste happens, carrying what, how much, and why. The why is the load-bearing column. A fortnight of honest reasons sorts itself into a short list of habits — poured too much, dropped, expired, unsold, dial-in — and habits can be changed. Quantities without reasons just produce a total, and a total only tells you the size of the problem you still cannot see. Keep the recording where the waste is: the sheet by the bin, the entry on the stock screen, the notebook on the pass. A record made later, from memory, is half fiction.
The Milk Jug
Milk waste is a pouring practice before it is anything else. The habits that hold it down are unglamorous and few: steam only what the round in front of you needs, in a jug sized for it — a small jug for one drink, never a big jug half-filled for three; never re-steam milk that has already been heated, and never top a cooled jug back into the bottle; and treat the last splash left in every jug as a design feature rather than a moral failing, because the alternative to a small remainder is a larger over-pour. Then watch the fridge discipline, because milk that expires on the shelf never had the chance to be anybody's pour.
Beans and the Morning Dial-In
Every grinder drifts overnight with humidity and wear, and the morning routine that brings it back is genuine quality practice — the waste is not in tuning, but in tuning carelessly. The habits: dial in once, early, with weighed shots, rather than iterating through the first ten orders of the day; purge to a consistent routine instead of a mood; and at close, empty or use what the hopper holds rather than discovering it next Tuesday. Weigh your dial-in waste for one week — the shots, the milk they take, the purge. The number is usually smaller than feared and larger than assumed, and either way it is now a fact.
The Display Case
The display wastes food by succeeding at its job — a full-looking case sells more, and a full-looking case at 16:00 is partly a forecast that did not come true. The fix is to bake and stock to the sales curve you actually have, which in practice means fewer trays later in the day and acceptance that the last hour sells from a thinner shelf. Then write the day-end policy and post it: what gets marked down in the last hour and by how much, what staff may take, what goes to donation if you have a partner for it, and what genuinely gets binned. Food-safety rules around markdown and donation differ by country, so check the rules where you trade — but the policy itself, written and dated, is the difference between a decision and a shrug.
Worked example
A café records waste for two weeks with reasons attached, on a sheet by the bin and in the stock screen. Milk: 9.5 litres wasted out of 210 litres used — about 4.5 per cent — split into 5.8 litres over-poured, 1.2 litres dropped, 1.5 litres expired, and 1.0 litre from display lattes never collected. Beans: dial-in and purge came to 0.9 kg over the week. Display: 18 unsold items across the fortnight. The fixes were three habits and no spending: jugs matched to the round, dial-in moved before opening with weighed shots, and the afternoon bake cut by one tray. The next fortnight's count: 3.1 litres of milk — under 1.5 per cent of a similar 205 litres used — and 0.4 kg of beans. The display held at 9 unsold, and the marked-down last-hour shelf sold most of what would once have been binned.
Turning Records Into Corrections
Read the record weekly, at a set time, and change one habit at a time — a café that rewrites five practices on a Monday changes none of them by Thursday. Give each fix a fortnight, then read the same columns again: if the reason-category shrank, the habit took; if it did not, the habit did not either, and the next guess should be different rather than louder. Over a quarter this loop — count, one change, recount — is the whole of waste reduction, and it is the rare corner of café management where the spreadsheet evening genuinely replaces the consultant.
The counting and the fixing are different jobs, and mixing them is the classic failure: the barista who measures the over-pour while being told off about it will stop recording before the habit changes. Separate them cleanly — waste recorded without commentary during service, habits discussed afterwards, in the quiet, with the numbers on the table.
Where the System Carries the Count
Paper works, and works well, but the count lands harder when it lives beside the stock it depletes. In Kafe360, stock control records Waste as its own movement type — separate from stock in, stock out and adjustments — so the fortnight's list adds itself up instead of waiting for a spreadsheet evening, and recipe deductions at the point of order keep the sales side honest underneath it. How recipes deduct ingredients as drinks are ordered is covered in recipe-based stock control, with a worked example — this article is about what happens around the recipes, not the mechanics inside them.
Be clear-eyed about what the panel does with those numbers: it counts quantities, and nothing else. There is no cost field, so no screen shows waste in money, and there are no automatic alerts when waste climbs — only the low-stock badge and your own weekly read of the movement list. The conversion from litres to lost margin happens once, in your spreadsheet, with your supplier prices. That is enough to make the decision; it is just not automatic.
The Waste Checklist
- Two weeks of records, made at the moment, with a reason on every line.
- Read the reasons, not the total — pick the one or two categories that repeat.
- Jugs matched to the round; no re-steaming; no topping cooled milk back.
- Dial-in before opening, weighed, once — and the week's tuning waste weighed once too.
- Afternoon display trimmed to the real curve, with a written day-end policy behind it.
- Re-count the same fortnight columns after the change, and keep only what shrank.
The end state is not a café obsessed with waste — it is a café where nobody thinks about it, because the pouring habits fit the drinks and the display fits the day. The two-week count is how you get there; the weekly read is how you stay.
Record a real fortnight of waste as its own movement type and read what the reasons say. Kafe360 is $249 a year, with a 30-day free trial and no card to start.
Start your 30-day free trialFrequently asked questions
Do I need a POS to track waste?
How much milk waste is normal for a café?
What should happen to unsold food at closing?
Is dial-in waste really worth counting?
Can Kafe360 tell me what my waste costs?
When is something too small to record?
Opening a venue? Start with a free QR menu, move to the full system when you are ready.