A cafe stocktake that takes twenty minutes

A cafe owner who has tried a full stocktake remembers it as a Sunday that ran to midnight. Very few of them do it twice.

The count that changes a business is smaller and more frequent, and it covers the handful of lines where the money actually sits.

Count what moves and what costs

Twenty lines usually carry most of the value: coffee, milk, the main proteins, alcohol where it applies, and whatever the menu uses most.

The rest can wait for a quarterly count. Weighing the sugar sachets every Sunday buys nothing and is the reason the whole habit collapses.

The same person, the same day, the same order

A stocktake is a comparison, so consistency matters more than precision. A count whose method changes each week produces a number nobody can trend.

Sunday evening after close, walking the same route through the store, counting in the same units. That repetition is what turns a count into information.

Count in units the kitchen uses

A supplier sells a twenty kilogram box and a kitchen uses portions. Counting boxes and costing portions means somebody does arithmetic under pressure.

One unit per item holds everywhere: the order, the count and the recipe. Mixed units cause more stocktake errors than miscounting ever does.

Theoretical against actual is the whole point

Opening stock plus purchases minus closing stock gives what the cafe used. Sales multiplied by recipes give what it should have used.

The gap is waste, over portioning, giveaways or theft. Without the count the gap hides inside food cost, where it looks like a supplier price rise.

Chase the variance, not the total

A cafe with a perfect total can still have two items badly wrong in opposite directions. The item level variance is the useful output.

One or two lines usually explain most of it, and they repeat month after month until somebody changes a portion, a supplier or a habit.

The count comes before the delivery, not after it

A count taken with a delivery half unpacked mixes two periods and produces a figure nobody can use.

Counting after close and before the first delivery of the week gives a clean line between one period and the next.

Deliveries and wastage feed the same picture

A stocktake with no record of what arrived is half a calculation. Supplier invoices give the purchases and the waste log gives one explanation for the gap.

Our guide to wastage covers collecting that in seconds a day, which is what makes the variance readable.

Act on it the same week

A variance noticed on Sunday and discussed the following month teaches nobody anything. The people who caused it have forgotten the week.

A short conversation on Monday, naming the item and the number, changes behaviour while the shift is still fresh.

A variance that repeats for three weeks has become a fact about the business rather than an accident. That is the point to change a recipe, a supplier or a portion.

Par levels and the count answer each other

A count showing three weeks of an ingredient on the shelf is an ordering problem rather than a usage one. Cash is sitting in the store room.

Our guide to par levels covers setting those properly, and the weekly count is what tells you whether they still fit the trade.

What to change first

• Twenty lines carry the count, and the rest waits for a quarter
• One person, one evening, one route through the store, every week
• Each item has a single unit, used by the order, the count and the recipe
• Theoretical usage meets actual usage, item by item
• The count happens after close and before the next delivery
• Monday brings a short conversation about the two worst lines