Wrong item in the box: where warehouse mistakes are really made

The stock count shows three units missing from one product and three extra of the product stored right next to it. Same model, different size. Nothing was stolen and nobody miscounted. The goods left three times under the wrong code, and the stock count simply added it up for the first time.

Mistakes like this usually get pinned on the packer. Look at how they actually happen, and the picker is most often just the place where the mistake became visible.

Every mistake has an address

There are three places in a warehouse where the wrong item gets into a parcel. Each needs a different fix, so it matters which one is yours.

The code

Two products differ only by size, colour or pack quantity, and their codes differ by one character, or only a note in the product name tells them apart. On screen it looks obvious. On a printed list, in a hurry, next to a shelf of ten nearly identical boxes, it does not. When the difference between products has to be read rather than checked, sooner or later it will be misread.

The same problem shows up later when a business tries to bring in AI, as we explained in the piece on why clean product data comes first.

The location

Stock has no fixed place; everyone “just knows where it is”. The person who put it away knows. A new picker, holiday cover or a colleague from another shift searches using someone else’s memory of the layout. Returned goods are the worst: they come back, go on the first free spot, and from then on sit where nobody expects them.

Closing the parcel

The last moment a mistake can be seen is before the box is sealed. If the contents are not compared with the order at that point, the next person to compare them is the customer, and that is the most expensive way for the business to find out.

Before any scanner: two weeks of an error log

Before buying equipment or software, spend two weeks writing down every delivery mistake, both those customers find and those colleagues catch. For each one, note two things:

• What was wrong: another variant of the same product, the wrong quantity, a different product altogether, damaged goods, or the wrong address.
• Where it was found: in the warehouse, by the courier, or by the customer.

The pattern tells you where to start. If swapped variants dominate, the problem is in codes and labels. If items “disappear” and turn up somewhere else, the problem is the layout. If the errors are in addresses and customer details, the warehouse is not at fault: the mistake came in when the order was entered, and that is where it has to be stopped.

What scanning changes, and what it does not

Scanning the barcode as the parcel is closed compares the item in hand with the item on the order. The wrong variant and the wrong quantity are stopped there, while the goods are still in the warehouse and the fix takes moments. You no longer need dedicated hardware for this: a phone camera can read barcodes, and the app should keep working in the parts of the warehouse where the signal drops.

Scanning does not fix what happened before it:

• If two products share a barcode, the scanner cannot tell them apart any better than a person. Every variant needs its own code and label, including goods without a manufacturer’s barcode, which get a label printed on arrival.
• If stock is in the wrong place, the scanner only catches it at the end, after the search. A tidy layout is faster.
• If the address was typed wrongly, the scanner never sees it.

An order of work that pays off

1. Codes and labels. Every variant gets its own code and a label on the item and on the shelf. It is the cheapest step, and nothing else works without it.
2. Fixed locations. Every product has its own place, and returned goods have a separate spot where they wait to be checked before going back on the shelf.
3. A check when the parcel is closed. Once codes and locations are in order, scanning does what it promises.
4. The log stays. Even after the check is in place, mistakes are still written down, there are just fewer of them. That shows whether the problem has moved somewhere else.

After the first step alone, some of the mistakes often simply stop, because no two products read the same any more. What remains is easier to find, and easier to measure.

A packing-check app on a phone, working offline wherever the warehouse has no signal, is part of our iOS and Android app development service. You need nobody’s help to keep the error log, and once it is filled in, send it to us and we will look together at where your mistakes start.