How to run a warehouse stock count, step by step

As the year end approaches, many warehouses go through the same routine: count sheets are printed, teams are set, receiving and shipping are put on hold. The count finishes, the results go into the system, and the first question is usually: “Where did these differences come from?” This article covers how to plan a warehouse stock count from start to finish, how to pin down the differences, and how to stop the count from being a job crammed into the last days of the year.

Why counts don’t match

When a count doesn’t match the records, the cause is usually not the count itself but the months before it. Every move that was never recorded turns up as a difference on count day. The count can create errors of its own too: a rack is counted twice or not at all, a counter sees the expected quantity on the sheet and ticks it off as “matches”, or an item is counted in boxes but recorded in units.

A good count therefore does two things at once: it knows for certain which locations have been counted, and it narrows differences down to an area small enough to find their cause.

Types of stock counts

Not every count has to cover the whole warehouse. Depending on the goal, there are four kinds:

Before the count

The quality of a count is largely decided before count day:

• Close open work. Put away goods that were received but not shelved, and ship prepared orders or move them to a separate area. Every half-finished job becomes a difference in the count.
• Stop the movement. For a full count, hold receiving and shipping for the duration; for a zone count, make sure nothing is taken from the zone being counted.
• Check locations and labels. Every bin needs a scannable location label; reprint missing or worn ones, and label items that have no barcode.
• Separate set-aside stock. Damaged, returned or quarantined goods should sit in their own locations so they don’t mix with sellable stock.
• Assign people and areas. Give each counter a storage area or a range of locations, so no rack is counted twice or skipped.

Count day, step by step

1. First count The counter scans each location, counts the items there and enters the quantity, without seeing the expected quantity. Because locations are counted by scanning, you know exactly which ones have been counted.
2. Check count Locations that came out short or over go to a check count and can be assigned to a different counter from the first one. This time the counter sees the details of the discrepancy and counts with pinpoint focus. Locations that matched aren’t counted again.
3. Until the difference is zero Locations that still differ get another check count, and this can go on until nothing differs. A supervisor can count too when needed.
4. Review and approval Counted, WMS and ERP quantities are reviewed side by side, the causes of differences are investigated, and approving the result updates location stock.

Blind counts and pinpoint check counts

If the count sheet says “24 pcs should be on this shelf”, the counter will, without meaning to, try to confirm that number. That’s why the first count should be blind: the counter never sees the expected quantity and simply enters what they count, and the system then tells them whether the result matches, is short or is over. That way the counted quantity really is counted.

In a check count, the goal changes: now you need to find where the difference comes from. At this stage the counter sees the details of the discrepancy and counts with pinpoint focus. The location can also go to someone other than the person who counted it first, so the same eyes don’t repeat the same mistake.

Where differences come from

A difference only stops coming back once its cause is found. The most common causes are:

• Put away in the wrong location: the item is in the warehouse, just on a different shelf. If one location is short and a nearby one is over by the same amount, this is usually why.
• Unrecorded moves: goods handed out without a document, returns or transfers that were never recorded.
• Shipping errors: an item that wasn’t on the order, or too many units, went out; it comes back as a shortage in the count.
• Unit mix-ups: items with different quantities per box counted in boxes and recorded in units, or the other way round.
• Damaged and set-aside goods: items taken out of sellable stock without updating the record.

Most of these can be prevented well before count day by recording every move with a scan.

Check with your accountant about your legal year-end inventory requirements and how count differences should be posted in your books.

Don’t wait for the year end: random counts

Counting the whole warehouse once a year means seeing all the differences once a year, when it’s too late to find their cause. With random counts, a group of randomly selected locations is counted throughout the year, every week for example. As a result:

• Differences are caught close to the time they happen, while the cause can still be found.
• The warehouse doesn’t stop; counting happens in small batches between daily tasks.
• You go into the year-end count with cleaner records, and the differences are smaller.

You can keep an eye on records between counts too: a report comparing WMS and ERP stock shows any gap opening up between the two systems without waiting for count day.

Stock counts in Buyasoft WMS

In Buyasoft WMS, counts are planned the way this article describes and done on handhelds:

• Full, initial and random count plans, with counters assigned by storage area or location range
• First count, check count and supervisor count, with recounts limited to locations that came out short or over and check counts repeated until the difference is zero
• Check count locations reassigned to other counters, who see the details of the discrepancy and count with pinpoint focus
• In the first count, the handheld never shows the expected quantity; it only tells the counter whether the result matches, is short or is over
• Counted, WMS and ERP quantities side by side; approving updates location stock
• Location lists uploaded from Excel and location labels printed in bulk, plus reports on stock by location and WMS vs. ERP stock

Count accuracy doesn’t stop at the warehouse. Your dealers see stock warehouse by warehouse in the Buyasoft B2B portal, which takes stock from your ERP, so the more accurate your records, the more accurate the stock your dealers see. Your sales team works on the same platform with Buyasoft CRM. Buyasoft WMS works with Logo, Netsis, Mikro, SAP and other ERPs.

Frequently asked questions

Does the warehouse have to stop completely during a count?

For a full count, usually yes: receiving and shipping are put on hold for the duration. For a storage area or location range count, stopping movement in the section being counted is often enough, and random counts fit in between daily tasks.

How many times should a location be counted?

Once, if it matches. Locations that came out short or over go to a check count, and check counts can repeat until the difference is zero. That way the effort goes where the problems actually are.

How do we run the first count when moving to a WMS?

You upload your location list from Excel and print location labels in bulk; products and barcodes come from your ERP. Your team then counts every location on handhelds under an initial count plan; counted quantities appear next to the ERP quantities, and approving creates your location stock.

Do we need handhelds for stock counts?

Yes. In Buyasoft WMS, counts are done with the handheld app, which runs on Android devices, by scanning location and item barcodes. That’s what makes it possible to track counted locations, keep the first count blind and run check counts.