Guide · Any business that holds stock

How to run a cycle count without shutting down

Replace the dreaded annual stock-take with short, regular counts: what to count, how often, how to count well, and what to do with the differences.

A full stock-take means closing the stockroom, counting everything and then spending weeks arguing about the differences. Cycle counting does the same job in small pieces: count a few items every week, fix what you find, and your numbers stay right all year. Here is how to start.

1. Decide what to count, and how often

Not every item deserves the same attention. A simple rule many businesses use:

• High-value or fast-moving items: count monthly.
• Everyday items: count every quarter.
• Slow, low-value items: count twice a year.

Also count any item the moment something looks wrong: a negative balance, a job that couldn't find a part that should have been there, or a big adjustment last time.

2. Keep each count small

Twenty to forty items a day or a week is plenty. A count that takes half an hour gets done; one that takes all afternoon gets postponed. Spread counts across locations so the same shelf isn't always first.

3. Count blind

Give the counter the item and location, not the expected quantity. People who can see the expected number tend to find it. Count, write the number down, and only then compare.

4. Investigate before you adjust

When a count doesn't match, check before changing anything: is there unrecorded receiving, a job that used parts without recording them, stock in a second location, or units of measure mixed up (boxes vs each)? Recount if the difference is large. Then adjust, and write down why. The reason is what helps you fix the cause.

5. Fix the cause, not just the number

Look at your adjustment reasons every month. If most differences come from unrecorded job use, the fix is recording parts at the job, not more counting. Cycle counting works best as an early-warning system for the process.

The checklist

• Items grouped by value and movement, with a count frequency for each.
• Small, regular counts spread across locations.
• Blind counts: the counter doesn't see the expected quantity.
• Differences investigated and recounted before adjusting.
• Every adjustment has a written reason; reasons reviewed monthly.