Skip to content

Warehouse operations5 min read

From an Annual Stocktake to Cycle Counting: Keeping Inventory Accurate All Year

A full annual stocktake stops operations and finds errors late. Learn how cycle counting works: what to count and when, why blind counts matter, and how to approve variances and record adjustments with clear reasons.

MIRSAD Editorial Team

Practical content on inventory and warehouse management

Abstract illustration in MIRSAD colours of a count sheet with check marks in front of rows of shelving

In many businesses the stocktake happens once a year. Receiving and issuing stop for a day or more, the whole team is called in to count, and large variances appear that nobody can date or explain. The figures are corrected in a single day, and then the same errors slowly build up again until the next stocktake.

The problem with a single annual stocktake

A full stocktake is not wrong in itself, but on its own it is not enough:

  • It finds problems late: a variance found today may have started months ago, which makes its cause hard to trace.
  • It stops operations: pausing receiving and issuing delays customers and suppliers, sometimes at a sensitive time of year.
  • It exhausts the team: long, continuous counting increases the chance of counting errors.
  • It fixes the number, not the cause: quantities are corrected, but the working practice that created the variance stays the same.

What is cycle counting?

Cycle counting means counting a small part of your stock on a regular schedule, daily or weekly, so that every item is counted within a set period and the important items are counted more than once. The warehouse keeps running, the whole team is not needed, and variances surface while they are small and close to their cause, so they are easier to explain and fix.

What to count, and when

Not every item deserves the same frequency. Set priorities by the impact an error would have:

  1. High-value items: a small variance means a large amount of money.
  2. Fast-moving items: frequent receiving and issuing create more opportunities for error.
  3. Items with past variances: repeated variances signal a problem in the process.
  4. Items tracked by batch or serial number: errors here affect expiry and warranty.

A practical approach is to count by location rather than by product: count one rack or one aisle at a time. The work stays organised, and it is easy to confirm that no bin was missed.

Timing matters just as much. Avoid counting during heavy receiving or peak issuing, because stock in motion creates false variances. Counting at the start of the day, before picking begins, or in an agreed quiet period works better, with movement paused only for the location being counted.

Blind counts: why the counter should not see the expected quantity

If counters see the system quantity before they count, they will unconsciously aim for it and may stop counting when they get close. Blind counting avoids this: the counter sees only the product and its location, records what is actually there, and the system compares the result with the recorded balance after the count is submitted. This simple step makes results far more reliable, because it separates the person counting from the person who knows the expected number.

To keep the count itself accurate, agree on the counting unit with the team before starting: are you counting pieces or cartons? If a product is counted in cartons but recorded as pieces, a large variance appears that does not exist in reality. Record the quantity in the location where it was actually found, even if that is not its usual place. That in itself is valuable information: it shows stock was moved without being recorded, and helps you correct course before it happens again.

Approving variances and recording adjustments

Not every variance should become an automatic change to the balance. Set approval limits by quantity and value: a small variance on a low-cost item might be approved by a warehouse supervisor, while a higher-value variance needs the inventory manager or the owner. Before approving, recount items with large variances to make sure the difference is real and not a counting mistake.

When a variance is approved, record the adjustment with a clear reason: damage, loss, an earlier receiving error, or an unexplained surplus. Over time these reasons become a valuable source of improvement. If "receiving error" keeps coming up, for example, the problem lies in the receiving process, not in the count.

How to start in the first few weeks

Do not try to roll out cycle counting across the whole warehouse at once. Start with one zone or with the highest-value group of products, set a fixed day and a short counting window, and let the same team handle it until the routine settles. After each count, spend a few minutes reviewing variances with the supervisor: what keeps recurring, and which location needs tidying or a clearer label? Once variances in the first zone are few and explainable, extend counting to the next zone. This way counting turns from an exhausting annual event into a light weekly habit, and the team sees its effect in more accurate balances and faster picking.

Checklist to get started

  • Have you set priority items and a counting frequency for each group?
  • Are counts done by location, in quiet periods, away from peak receiving and issuing?
  • Are counts blind, so counters cannot see the expected balance?
  • Are there clear approval limits for variances by quantity and value?
  • Is every adjustment recorded with a reason, and are recurring reasons reviewed regularly?

In MIRSAD you can create a count for any warehouse and record counted quantities by product and storage location, make it blind for the counter, set variance approval limits per role by quantity and value, and record adjustments with their reasons in a movement history and audit log that cannot be edited. You can see the rest of the platform on the platform page.