Stock Audit Process for Small Factories: The Quarterly Checklist

Most SME factories audit stock once a year, panic when the variance is huge, and then promise themselves they'll "do it more often next year." They never do. The problem isn't discipline — it's that the annual audit is the wrong unit of work. It's too big, too disruptive, and too late to fix anything.

A quarterly cycle of small, structured counts catches problems while they're still cheap to investigate, and removes the need for a full shutdown audit almost entirely.

Pre-Audit: Freeze and Reconcile

A day before the count, freeze inward and outward movements for the SKUs you're auditing. Process every pending GRN, every pending issue, every pending dispatch — so the system stock represents an actual snapshot. Counting against a moving target is how variances get manufactured by the audit itself.

The Count: Two People, Blind Sheet, No System Number

The person counting should not see the system stock. If they do, the count becomes a search for confirmation, not a measurement. Use a blind count sheet — SKU and location only — and have a second person re-count any line where the variance exceeds 5%. This single rule eliminates more than half of "audit errors."

The Reconciliation: Variance With a Reason

For every line where physical doesn't match system, document the variance and a probable cause. Damaged in storage. Issued without entry. Received without entry. Mis-binned to a similar SKU. The reason matters more than the number — because the reason is what you fix.

The Quarterly Cadence

Divide your SKUs into four buckets — A items (high value or fast moving) audited every quarter, B items every six months, C items annually. Pick one day per month for a focused four-hour count. Over a quarter, you've audited every A item and a third of your B items, without ever shutting down production.

The Year-End Difference

Factories that run quarterly cycle counts walk into year-end with variance under 1% and an explainable trail for everything outside that. Factories that don't walk in blind, lose three production days to a full count, and discover variances they can no longer explain. The first version is cheaper, calmer, and audit-defensible. The second is what "we'll fix it next year" looks like in practice.