Cycle counting: the twenty-minutes-a-day method to keep book stock right

Why the annual stocktake is not enough

The annual stocktake serves the accounts. For out-of-stocks, it comes too late: a wrong stock record in February creates out-of-stocks until the next stocktake, because automatic ordering trusts book stock. The study by DeHoratius and Raman measured 65% of inventory records inaccurate at a retailer, with gaps all the larger when the reference sells fast. Those are precisely the references most often missing from the shelf.

Cycle counting corrects book stock where it matters, before the error costs sales.

Choosing the references to count

Counting everything every day is impossible. The selection makes the method:

1. The fifty best-selling references of the store, each counted once a week (ten a day).
2. Promoted references, counted the day before the operation and midway through.
3. References reported missing while book stock is positive: that is the list of probable phantom stocks, to count the same day.
4. References with recurring gaps: those where the last count revealed a gap are counted again the following week.

This selection changes every day; it fits on a sheet or a screen.

The method in twenty minutes

1. Print or display the day’s list: ten to fifteen references, with their book stock.
2. Count on the shelf then in the back room, including open cases and misplaced products nearby.
3. Correct book stock immediately for each gap, with a reason: unrecorded breakage, theft, receiving error, misplaced product.
4. Record the gap in a simple table (reference, date, book, actual, cause): that table reveals the causes to treat.

At ten references a day and two minutes per reference, that is twenty minutes, done by the same person, at the same time, preferably before opening or in the late-morning lull.

Reading the gap table

After a month, the gap table tells the story of the store:

• gaps concentrated in fresh point to unrecorded breakage;
• gaps on small expensive products point to theft;
• gaps at receiving on categories with many references point to checking by case instead of by reference;
• positive gaps (more actual than book) point to unrecorded returns or receipts.

Each cause has its fix, and the fix durably reduces out-of-stocks due to phantom stock, described in phantom stock: what to do when the book stock says the product is there.

What to avoid

• Counting without correcting: a count not followed by a book stock correction is useless.
• Counting at random: slow-moving references rarely have gaps that cost; targeting makes the return.
• Delegating without a fixed time: cycle counting survives if it is a ritual, not a task “when there is time”.

An example

2,200 m² supermarket. Cycle counting set up on fifty fast movers, ten a day. First month: 34 gaps found in 200 counts, including 21 overstated book stocks, that is 21 references whose automatic ordering no longer fired. After correction and treatment of the causes (recording breakage in fresh, checking by reference at receiving), the second month counts 12 gaps, and out-of-stock reports on those references fall by a third.

What shopper reporting changes

ShelfAlert provides the method’s third list, the most profitable one: products that shoppers report missing while book stock says they are there. Those are phantom stocks costing sales right now, to count the same day. 14-day trial, no payment card: see the plan for store managers.

Sources

• DeHoratius, Raman, “Inventory Record Inaccuracy: An Empirical Analysis”, Management Science, 2008: abstract.
• Corsten, Gruen, “On Shelf Availability”, 2004: summary.