Barcode inventory management that works on a phone and at a desk

Two ways to scan, depending on where you are standing

Barcode hardware in a small business is rarely uniform. Somebody has a phone in their pocket on the floor, and somebody else has a ₹1,200 USB scanner wired to a desktop in the office. Both work, and they write the same record.

• On Android and iOS, the camera is the scanner. Open the barcode screen, point it at the label, and the product resolves. No extra hardware, no pairing, nothing to charge.
• On the web, a USB or Bluetooth barcode scanner behaves as a keyboard: it types the code and presses Enter. Focus the field, scan, and the product is found. This is why cheap scanners work — there is no driver to install and no integration to configure.

The same screen accepts a typed code, which matters more than it sounds. Labels get scuffed, and a system that only accepts a clean scan stops the person using it.

SKU and barcode are different fields, on purpose

Every product carries both. They are not the same thing and conflating them causes real problems later:

• The SKU is yours. You choose the format, it means something to your team, and it stays stable even when a supplier changes packaging.
• The barcode is usually the manufacturer's — an EAN or UPC printed on the box. You do not control it, and two of your SKUs can legitimately share one if you repack.

Keeping them separate means Excel import can match on either. When a supplier sends a price list keyed by their barcode, you can bulk-update against that. When your own team sends a count keyed by your SKU, you can bulk-update against that instead.

Where a scan actually lands

Scanning is not a feature in isolation — it is an entry point into the movement screens:

Counting a shelf with a phone

The stock take flow is the one that most repays a scanner. You create a count — the whole catalogue, a category, or a zone — and then walk the shelf scanning and entering what you actually see. The system holds its own figure back until you finish, so nobody anchors on the expected number, and at the end you get a variance list: what was over, what was short, and by how much.

You then post the count, which writes adjustment transactions with the stock take as the reason. That is the important part. The quantity does not simply change; there is a record saying it changed because of a count on a particular date, approved by a particular person.

If your products have no barcodes

Plenty of businesses sell loose, repacked, or own-brand goods with nothing printed on them. Two things work:

• Use your SKU as the barcode value and print your own labels — any thermal label printer will encode a Code 128 from plain text.
• Or skip barcodes entirely. Every screen that accepts a scan also accepts a search by name, SKU or category. Barcodes make it faster; they are not a prerequisite.

Every scan is a transaction, not an edit

This is the part that separates a stock system from a spreadsheet with a scanner attached. A scan never overwrites a number. It appends a typed movement — a receipt, an issue, a transfer, an adjustment — carrying a quantity, a timestamp, a reason and the person who did it. The current quantity is the sum of those movements.

Because of that, the reports can be trusted, the audit trail is complete by construction, and a disputed number can always be walked backwards until you find the movement that caused it.