A stocktake goes wrong in one of two ways: the sheet is missing a column you needed, or two people counted the same shelf and nobody counted the next one. Both are layout problems, and both are fixed before anyone picks up a pen.
Last updated
A count sheet has to be usable by someone standing on a step ladder holding a torch. That constraint decides the layout more than anything else: few columns, wide rows, and the item name where the eye lands first.
If a sheet says "beer — 14", nobody a week later knows whether that is fourteen bottles, fourteen cases or fourteen kegs. This one ambiguity produces more wrong stock valuations than arithmetic errors do.
Decide the unit per line, print it on the sheet, and count in that unit only. Where something is genuinely counted two ways — full cases in the store, loose bottles in the fridge — give it two lines, not one line and a note.
The reason a stocktake takes all night is rarely the counting. It is the recount, because two people worked the same aisle from opposite ends and the totals do not reconcile.
Split the sheet by physical location before you start, and give each person a section outright. Count in the order the stock is physically arranged, not alphabetically — an alphabetical sheet makes a counter walk the room once per item.
The count gives you quantities. The value is quantity × unit cost, and the only real decision is which cost you use.
| Method | What it means | Good for |
|---|---|---|
| Latest cost | The most recent price you paid for that item. | Fast-moving stock and a quick figure. Easiest to actually do. |
| Weighted average | Total spent on the item, divided by total units bought. | Items whose price moves around. Smooths out a single odd invoice. |
| FIFO | Assumes the oldest stock sold first, so what remains is valued at recent prices. | Perishables, and where you need a defensible figure for the accountant. |
A stocktake in isolation gives you a number. A stocktake compared to what should be there gives you an answer.
Expected closing stock is: opening stock, plus everything delivered, minus everything sold or written off. The gap between that and what you actually counted is your variance — and variance is the entire point of the exercise. A count that is never compared to anything is a night's work for a spreadsheet nobody opens.
Stock that moves during a count is stock that gets counted twice or not at all. Count before opening or after close, not during service, and freeze deliveries for the duration.
If a delivery genuinely cannot wait, put it aside unopened and in one place, and note it separately. A pallet counted into the shelves halfway through a count is the reason the variance report makes no sense the next morning.
Regentflow's Stocktake does this with your whole team at once — one link, everyone counts on their own phone, and the sheet fills in live. No transcription at the end, because there was never any paper. Build the list by pasting it from a spreadsheet or reading it off a photo of a supplier invoice; walk the room saying the numbers instead of typing them; save the list so next month's count starts already written. Excel and PDF export on every count.