A cycle count is a rolling stock count covering a small subset of lines frequently, rather than a full physical count of everything twice a year. It suits small stores because it never requires stopping, it finds errors while the cause is still traceable, and it spreads the work into portions that actually get done.
In more detail
The full annual count has one advantage, completeness, and several disadvantages. It stops operations, it is exhausting, and by the time it finds a discrepancy the movement that caused it happened months ago and cannot be investigated.
Cycle counting trades completeness for traceability. Counting twenty lines a week means a discrepancy surfaces within days of the error, when the delivery, the picker and the paperwork are all still identifiable.
Prioritise by value and by movement. Fast-moving, high-value lines are where drift costs most and compounds quickest. A slow-moving low-value line can wait months without anyone being worse off.
How often to count what
| Line type | Suggested frequency | Reasoning |
|---|---|---|
| Fast-moving, high value | Monthly | Drift costs most and compounds quickest |
| Fast-moving, low value | Quarterly | Errors are frequent but individually cheap |
| Slow-moving, high value | Quarterly | Rare movement, expensive to be wrong about |
| Slow-moving, low value | Once or twice a year | Little movement, little cost |
Where people get it wrong
- Adjusting the number without recording the cause. That throws away the only information the count produced.
- Counting everything equally often. Value and movement should decide frequency.
- Counting while picking is happening in the same location, which produces discrepancies that are artefacts of the count itself.
- Treating a repeated shortfall as a counting problem. The same line short three times is telling you about receiving, picking or shrinkage.
Counting without stopping
The practical objection to counting is that it interrupts picking. Two things mostly remove that: count a small number of lines rather than an area, and count before opening or after closing rather than mid-shift.
Counting a location while picking is happening in it produces discrepancies that are artefacts of the count itself, which is worse than not counting, because you then investigate a problem that does not exist.
Common questions
How often should I cycle count?
Weekly for a handful of lines, prioritised by value and movement. Fast-moving high-value lines might be counted monthly, slow low-value lines once or twice a year.
What does a discrepancy mean?
It is information about a process rather than a number to correct. Record what you found and which cause it was, so a repeat has something to compare against.
Does cycle counting replace a full stock count?
For many small stores it effectively does, though your accountant may still want a full count at year end. The main benefit is finding errors while the cause is still traceable.