Back to Blog
Guides

QR Codes for Inventory Management: What a Scan Can and Can’t Count

Nov 12, 2026•9 min read
QR Codes for Inventory Management: What a Scan Can and Can’t Count

The sentence every QR inventory pitch skips: a QR code does not count anything. It is a printed address. The label on shelf A-03 does not know there are fourteen boxes on it, does not notice when someone takes two, and will cheerfully open the same page the day the shelf is empty. The count lives in a system, whether that is a spreadsheet, an inventory app or the stock module of your shop platform. The code is how a person standing at the shelf gets to the right line of that system without typing.

That is still worth a lot. The slow part of a stock count is rarely the counting. It is finding the right row, the right SKU, the right tab, with a box in one hand. One scan removes that. But it means the decisions that matter are about what the code opens, and they are worth taking before you print three hundred labels.

First decision: a phone or a scanner?

This one changes what you encode, and getting it wrong is the most expensive mistake on this page, because it is printed on every label.

If the person scanning uses an ordinary phone camera, the code should contain a web address. The phone opens it. That address can be a page about the bin, a row in a shared sheet, a form for logging a count, anything that loads in a browser. No app to install and no hardware to buy.

If your inventory software reads codes with its own scanner, a handheld scanner gun or the app's built-in camera, it usually expects the code to contain an identifier: the SKU, the bin ID, a serial. A handheld imager works like a keyboard. It types whatever the code contains into whichever field has focus. Point one at a code that holds a full web address and your SKU field fills with https://…. Check what your software expects before generating anything. It is a two-minute test with one printed label.

One hardware fact to check at the same time: a laser scanner of the kind built for ordinary barcodes cannot read a QR code at all. QR is a two-dimensional symbol and needs an imager, a camera-type scanner. If the guns on your floor are laser units, a QR label is a phone-only label there. QR code vs barcode covers what each symbol holds and when a plain barcode is still the better choice.

What to tag: the shelf, the product, or the unit

Three different things can carry a code, and they cost very different amounts to keep honest:

  • Locations: the bin, shelf, rack or pallet position. A location label rarely changes, there are fewer of them than products, and it answers the question people ask most in a stockroom: what is supposed to be here? This is where most small operations should start.
  • Products (SKUs): every item of the same kind shares one code. Useful when the same product lives in several places, or on the item itself for reordering. One address per SKU.
  • Units: each individual item has its own code, for serial-tracked or high-value stock. This is a real register with one record per item, and it is the most work by a distance. Do not start here unless your stock genuinely needs it.

The trap is mixing them on one label without saying which it is. A code on a shelf that opens a product page is a product code wearing a location's sticker, and the day that product moves to another shelf, the label lies.

A count that works from a phone

The workflow that holds up with phone-only labels is short. Scan the location, see what is meant to be there, record what is actually there, move on. Concretely, the location's code opens one small page or sheet row that shows:

  • the location ID in large type, so the person knows they scanned the right label;
  • the SKUs expected there, with a photo if packaging looks alike;
  • one obvious way to record the count, a form or an editable row in whatever tool already holds your stock numbers.

Two rules keep it honest. Never put a quantity on the label: a printed number is wrong from the first sale. And never make the count depend on the page loading, for the reason in the next section but one.

Static, dynamic, and the bulk-generated label

Shelves get renamed, stockrooms get reorganised, the spreadsheet gets replaced by an app. A bin label has to keep working through all of it, so the question is where the flexibility lives.

A static code has the address printed into the pattern itself, permanently. There is nothing in the middle to pay for or to break. It also cannot be changed. A dynamic code holds a short address that forwards, so the destination can be changed after printing. In exchange, it depends on that forwarding service continuing to run. Static vs dynamic QR codes sets out the general trade.

The approach that survives longest for static labels is the one in QR codes on equipment: encode an address on a domain you own, one stable path per location, such as yourcompany.example/bin/A-03-2. You decide what that path serves, so the label never needs reprinting even when the system behind it changes.

That matters here specifically, because inventory labels come in hundreds and nobody types hundreds of codes by hand. The Business plan's bulk generator takes a CSV of up to 1,000 rows per batch, one URL per row, and makes one code per row. Those codes are static. That suits bin labels well, provided each row's URL is a stable address you control and not today's link into a tool you might leave. It also means the next section's scan counts do not apply to them. A static code sends the phone straight to the address, so no scan passes through anything that could count it.

Warehouse labels fail in predictable places

The code is the robust part. What breaks is the label, and a stockroom breaks labels in its own ways:

  • Height. A top-shelf label is read from the floor, at a distance and an angle. Size it for that, and remember the lever that helps most: a shorter encoded address means fewer, larger modules, which read from further away.
  • Glare. Shrink-wrap, laminated pouches and glossy label stock throw light straight back at the camera. Matte stock scans more reliably than shiny.
  • Cold. Ordinary adhesives let go in freezers and condensation. Freezer-grade labels exist for a reason.
  • Abrasion. Shelf edges get hit by boxes, trolleys and forks all day. Put the label where the traffic is not, or use something tougher than paper.

Keep the code plain: dark on light, a full quiet zone, no logo in the middle. QR code design that still scans covers the limits. A bin label does not need branding.

Stockrooms have dead spots

The back of a warehouse, a basement store or a walk-in cooler is often exactly where phones lose signal, and a code that opens a web page does nothing there. Three defences, cheapest first:

  • Print the location or SKU in human-readable text beside every code. Somebody can write it down and look it up later. It costs nothing, and it is the fallback that survives every failure on this page.
  • Keep the destination light. A small page loads on one bar. A heavy dashboard does not load at all.
  • Do not build a count that cannot finish without a page load. If it can't, the count gets done on paper and typed in later, or not at all.

What the scan data shows, and what it is not

On a dynamic code, scan analytics record that a label was scanned: when, roughly where (country, from the network), on what kind of device. There is no name and no stored IP address on the other side. On inventory labels that is useful in narrow, practical ways. It shows which locations get scanned during counts and which labels have not been scanned since the day they went up. It also shows whether a cycle count you scheduled for one aisle actually touched it.

The line that matters: a scan is not a count, not a pick, and not a stock movement. It records that a phone read a square. It cannot say how many boxes were on the shelf, whether anything was taken, or who was holding the phone. Your stock figures come from what people record in your inventory system. The scan only tells you somebody arrived at the right place to record it. QR code tracking, explained goes through what each number supports.

When a plain barcode is the better tool

If you already run software built around barcodes, with laser scanners on the floor and SKUs printed as barcodes on every product, a QR label adds nothing for picking and packing. The barcode already does that job at speed. QR earns its place where a code needs to open something on any phone, with no dedicated hardware: a location page, a photo of what belongs on the shelf, a count form, a supplier's reorder page. Plenty of stockrooms end up with both, a barcode for the scanner and a QR code for the person.

Start with one aisle

Do not label the building first. Label one aisle, or the twenty locations people ask about most, and run one count with it. If people use the labels and the count goes faster, you have learned the system works for you. If they ignore the labels, you have learned that for the price of twenty stickers and found out why before printing the other three hundred.

The short version

  • The code does not count stock. The count lives in your system, and the code is how someone reaches the right line of it.
  • Decide phone or scanner before printing. Phones open web addresses. Scanner guns type whatever the code holds into a field.
  • Laser barcode scanners cannot read QR codes. Check your hardware.
  • Start with location labels. Per-unit codes need a real register behind them.
  • Never print a quantity on a label.
  • Bulk-generated codes are static. Give each row a stable address on a domain you own.
  • Size for the top shelf, avoid glare, use freezer-grade adhesive where it is cold, and print the ID in text.
  • A scan shows someone reached the label. It is never a count.

Ready to create beautiful QR codes?

Make your own in 30 seconds — no signup required to try.