Counting every line equally often
Most of the effort lands on slow items where nothing has changed.
Stock audit glossary · Counting methods
noun · also "cycle counting"
A cycle count checks part of the stock at a time on a rolling schedule, so that every item is counted over a period while the site keeps working. Groups of locations or items are counted in turn, usually with high-value and fast-moving lines counted more often. It gives continuous checking through the year, but no single date on which the whole stock was verified.
Meaning
A business that cannot stop trading has no day on which everything stands still, so a single full count is costly or impossible. A rolling programme solves this: the stock is split into groups, each group is counted in turn, and every item is reached over a period rather than at one moment. The schedule is weighted so that high-value and fast-moving lines come round far more often than slow ones.
The result is continuous checking of how well stock is recorded. Differences show up soon after they arise, while the documents that caused them are still at hand, instead of a year of errors appearing at once with no way to trace them. The limit follows from the same feature: different areas are counted on different days, so no single date exists on which the whole site was verified. Where a dated position is needed, a full count supplies it.
At the audit
Where a business counts in rotation, the programme itself is checked.
In practice
By location, category or value band, and give each group a counting frequency, tighter for high-value and fast-moving lines.
No picking or replenishment in that group while the count is on.
The counter is not shown the expected quantity, so the count is independent.
At once, while the goods are still there, then trace each one through the movement history.
Record the cause, and keep a running record of which groups were counted and when.
Worked example
Sample results from the schedule above.
| Result | Counts | What it means |
|---|---|---|
| Agreed with the system | 576 | No action needed |
| Receipt or issue not yet posted | 15 | Timing: corrected once posted |
| Goods in the wrong location | 6 | Found nearby, moved and re-recorded |
| Real shortage | 3 | Traced through the movement history |
576 of the 600 counts agreed, an accuracy of 96%. The value is in the 24 that did not: most were recording delays, which point to a process to fix, and three were real losses that need explaining.
Because the counts are spread across the month, these causes surface while the documents are still at hand. An annual count would show the same differences all at once, with the trail gone cold.
Watch out
A rolling programme fails quietly, which is why these are costly.
Most of the effort lands on slow items where nothing has changed.
The same difference comes back next cycle.
The result cannot be reconciled.
No single day covered the whole site.
Activity looks high while part of the range is never reached.
Tidy, reachable racks are counted again and again; awkward ones wait.
Questions
Group the stock by value and movement, give each group a counting frequency, and draw up a calendar that reaches every item within the period. On the day, freeze the group being counted, count without being shown the system quantity, recount any difference at once, and trace each difference through the movement history before posting the correction.
Two figures are usually worked out. The counts needed in a year are the items in each group multiplied by how often that group is counted, added across the groups. Count accuracy is the number of counts that agreed with the system divided by the number of counts made; 576 agreeing out of 600, for example, is 96%.
There is no fixed figure. Each business sets its own schedule, usually counting high-value and fast-moving items most often and slow, low-value items less often but still within the period. What matters is that every item is reached and that the coverage can be shown.
Patron Accounting LLP. This page makes no regulatory statement.
Sources: general stock-count and audit practice; no statutory statement is made on this page. General information, not professional advice.
Need help with cycle counts? See our stock audit service. A stock audit tests a rolling programme as well as the stock.