Barcode scanning in a small warehouse, without buying handheld terminals
4 min read · Updated
The short answer
Barcode scanning in a warehouse exists to replace a judgement with a check. A picker reading a shelf label and taking the right tub is a judgement that is correct almost every time; a scan that confirms the tub matches the order line is a check that is correct every time. You do not need rugged handheld terminals to do it — a $25 USB barcode gun behaves like a keyboard and needs no drivers, and a modern phone camera reads Code 128 and GS1-128 reliably at bench distance. The hardware question is much less important than where in the process you put the scans, and there are four places worth having them: receiving, picking, packing and shipping.
The four scans that matter
Adding scanning everywhere is how a warehouse ends up slower than it was on paper. These four each prevent a specific, expensive error, and the rest are optional.
| Point | What is scanned | What it prevents |
|---|---|---|
| Receiving | The product, then the bin it goes into | Stock that exists in the count and cannot be found, because nobody recorded where it was put |
| Picking | The product at the bin, before it goes in the tote | The wrong variant — the 500g instead of the 250g, the unscented instead of the lavender |
| Packing | Every unit into the box | The short ship: three of four items in the carton, discovered by the customer |
| Shipping | The order number on the label against the box | The swapped label, which sends two customers each other's parcel and costs four shipments to fix |
The last one is worth dwelling on because it is the cheapest scan to add and the most expensive error to have. A label swap is invisible at the bench — both boxes are correct, both labels are correct, they are simply on the wrong boxes — and it is discovered by two customers at once.
Do you need a handheld terminal?
Usually not, and the assumption that you do is often imposed by the software rather than by the job.
| Option | Cost | Good for | Watch out for |
|---|---|---|---|
| Phone camera | Nothing extra | Picking and receiving on the move; anywhere you are already carrying a phone | Needs decent light; slower per scan than a laser; battery over a full shift |
| USB barcode gun | $25–$40 | A fixed packing bench, where the same station scans all day | Tethered — it is a bench tool, not a walking one |
| Bluetooth pocket scanner | $60–$150 | Walking picks at high volume, paired to a phone | One more thing to charge and to lose |
| Rugged handheld terminal | $400–$1,500 | Cold storage, wet or heavy environments, very high pick volumes | Lead times, a mobile device management story, and a fleet to maintain |
The honest rule: start with the phone and the USB gun, and buy handhelds when a specific person tells you a specific thing is too slow. Buying a fleet first is a five-figure decision made before you know your own volumes.
Which barcode to print
For internal warehouse labels — bins, totes, order numbers — use Code 128. It encodes the full ASCII set, it is compact, and every scanner reads it. There is no meaningful competitor for this job.
- Code 128 — internal labels: bins, totes, order numbers, licence plates. The default, and correct almost always.
- GS1-128 — when you need to encode structured data alongside the identifier, such as a lot number and an expiry date in one symbol. Worth it if you ship into retail or need lot data readable by a customer's system.
- EAN-13 / UPC-A — the retail barcode already on your product packaging. You do not print these; you record them so a scan of the manufacturer's own barcode identifies the product.
- QR / DataMatrix — when the label is tiny or the data is long. Reads well from a phone, less universally from older laser guns.
Record the manufacturer's existing retail barcode against each product even if you also print your own. It means a picker can scan whatever is facing them rather than hunting for your label.
Printing labels that survive a warehouse
Two practical things decide whether scanning works on the floor, and neither is about software.
The first is quiet zone — the blank margin either side of the bars. A barcode printed edge-to-edge on a label will fail to read intermittently, which is worse than failing consistently because people conclude the scanner is unreliable rather than the label. Leave a clear margin of about ten times the narrowest bar width.
The second is size and placement on bin labels. A bin label is read from a metre away by somebody holding a tote, so print it large and put it at eye level on the bay, not on the shelf lip where it is obscured by whatever is stored there. Thermal labels fade in direct sunlight and under some warehouse lighting; if a label lives near a window, laminate it or reprint it annually.
What happens when a scan fails
This is the design question that separates a scanning system people use from one they work around. Scans fail routinely — a damaged label, a product without one, a unit repackaged by the supplier — and if the only options are 'scan it' or 'abandon the job', staff will find a third one, which is usually a photograph of a barcode kept on somebody's phone.
A workable system lets a supervisor override with a reason recorded, lets a picker report an empty or short bin without losing the rest of the job, and keeps a log of rejected scans so you can see which products keep failing. That log is genuinely useful: a product that fails scanning twice a week has a label problem, and you cannot know that without the record.
Common questions
Can I use a phone camera for warehouse barcode scanning?
- Yes. Modern phone cameras read Code 128, GS1-128, EAN-13 and QR codes reliably at bench and shelf distance, and phone-based scanning avoids buying and maintaining a fleet of handheld terminals. The practical limits are lighting, per-scan speed compared with a laser gun, and battery life across a full shift. For most small warehouses a phone for walking work plus a cheap USB gun at the packing bench covers everything.
What barcode format should a warehouse use?
- Code 128 for internal labels — bins, totes and order numbers. It encodes the full ASCII character set, is compact, and is read by every scanner. Use GS1-128 when you need to encode a lot number or expiry date in the same symbol as the identifier. Record the manufacturer's existing EAN-13 or UPC-A retail barcode against each product as well, so staff can scan whichever barcode is facing them.
How much does it cost to set up barcode scanning in a small warehouse?
- Under $100 in hardware for a small operation: a USB barcode gun at $25 to $40 for the packing bench, and phones staff already carry for everything else. A thermal label printer, if you need to print your own bin and product labels, is $150 to $300. The significant cost is the software subscription and the time spent labelling shelves — not the scanners.
Why do my barcodes scan intermittently?
- The most common cause is an insufficient quiet zone — the blank margin either side of the bars, which should be about ten times the narrowest bar's width. A barcode printed right to the edge of a label reads most of the time and fails unpredictably. Other frequent causes are print density too low on a worn thermal ribbon, labels faded by sunlight, and curved surfaces distorting the bars.
Should every pick be scanned?
- Scanning at the pick catches the wrong-variant error, which is the most common and most expensive picking mistake — the 250g instead of the 500g, the wrong scent. That makes it worth having in almost every operation. What matters as much is what happens when a scan cannot be done: if there is no way to record an exception with a reason, staff will work around the scan rather than through it, and you lose the control entirely.
Kinetel does the things described on this page.
Inventory and lot tracking, barcode scanning, guided and batch picking, pack verification and shipping — for growing product companies, not for enterprises with an implementation budget.