Every shop has some gap between system stock and physical stock. The problem is not that a gap exists — it is not knowing why. An unexplained variance means you are making purchasing and pricing decisions on numbers that are wrong.
Common causes of variance
- Items sold without being recorded (quick sales outside the system)
- Purchases physically received before the purchase invoice is entered
- Returns put back on the shelf without a recorded return
- Damage and waste not recorded as a stock deduction
- Misconfigured units of measure (case versus piece)
- Undocumented transfers between warehouses
Seven practical measures
- 1Stop off-system selling with one clear rule: no goods leave without an invoice.
- 2Tie goods receipt to entering the purchase invoice the same day.
- 3Create a formal return procedure instead of putting stock straight back on the shelf.
- 4Record damage and waste as a separate stock movement with a written reason.
- 5Configure units of measure and conversion factors accurately once per item.
- 6Document every inter-warehouse transfer as a stock movement, not a verbal message.
- 7Run rolling partial counts on your top 20% fastest-moving items instead of relying only on an annual full count.
Why rolling partial counts work better
An annual full count reveals variance months after it happened, making the cause almost impossible to trace. Rolling partial counts — auditing a group of items each week — shorten the gap between an error occurring and being found, which makes diagnosing and fixing the cause realistic.
The system’s role in reducing variance
A good system does not prevent human error, but it makes it traceable: permissions that block unauthorised edits, an audit log showing who performed each movement and when, and reports that surface items with recurring variance. With those in place, variance shifts from a mystery to a solvable problem.
All articles