Stocktakes have a sinister reputation they do not entirely deserve. Counting stock is not hard: it is boring, which is different. What turns boredom into disaster is attempting it without method inside a store that keeps living: customers buying what you just counted, colleagues moving boxes already tallied, and the well-founded suspicion, by day's end, of having counted something twice and something else never.

The remedy is not the heroic all-nighter: it is the logistics of small things, decided in advance.

Closed zones, not endless counts

The cornerstone principle of every stocktake that works: divide the store into small zones, and each zone gets opened, counted and closed. Closed means closed: tape, sign, mark on the map, and from that moment nobody touches anything in it until the end. The enemy is not the counting error, which can be fixed. It is the "almost finished" zone where someone pulled an item for a customer and did not say so.

The zone map with its progress states (to do, in progress, closed) pinned up in the back is the control tower of the whole operation. It beats any app if the alternative is collective memory.

Counting and selling are two jobs: separate them

The mistake that generates the most double counts: the same people counting are also serving customers, hopping back and forth. Every mid-shelf interruption is a potential error. The healthy split is sharp: counters count; floor staff sell and keep customers away from closed zones. With one explicit traffic rule: if a customer wants an item from a zone being counted, the floor person handles it and notes it on the zone sheet, rather than pulling it silently.

This does mean a stocktake day is a reinforced-staffing day, not a normal day with one extra chore. The extra hours get planned weeks ahead, when you also pick the right days: the quietest of the month, never under a big delivery.

Freeze first, count second

Half of stocktake chaos is born in the preceding days: unprocessed returns, goods in transit between back and floor, pending customer orders. Counting amid those flows is like photographing a moving subject. Hence the freeze rule: in the 24 to 48 hours before, clear all pending items and stop non-essential movements. The photo comes out sharp when the subject holds still.

Differences are information, not blame

Once the count is done the discrepancies arrive, and with them the temptation of a summary trial. Resist it: most differences have banal causes (receiving errors, returns never registered, breakage never logged) and only a cold analysis by product family reveals whether there is a pattern worth real attention. A stocktake that ends in a witch hunt produces exactly one guaranteed outcome: next time, the counts will be "adjusted" to avoid arguments. Goodbye, information.

Stocktake day is the perfect case for planning done properly: reinforced staffing, clear roles, counted hours. With Sked Solve you build it ahead of time, visible to everyone: the site shows how.