Guide
Setting up a barcode inventory system
Barcodes are cheap, the hardware is cheap, and the payoff is immediate — but only if scanning is wired into the moments where stock actually moves. Here is how to set it up without buying anything you do not need.
What barcoding actually buys you
Not speed, primarily. Accuracy. A person typing a SKU makes an error every few hundred entries; a scanner does not. Over a year of stock movements that difference is the gap between records you trust and records you check against the shelf.
Speed follows as a second-order effect, and it matters most in the places where you currently avoid recording things because it takes too long — which are exactly the places your data is worst.
The hardware you need
| Situation | What to use | Rough cost |
|---|---|---|
| Shop floor, warehouse aisle, van | The phone already in your pocket | ₹0 |
| Fixed counter or receiving desk | USB keyboard-wedge scanner | ₹1,000–₹2,500 |
| Moving around near a desktop | Bluetooth keyboard-wedge scanner | ₹2,500–₹6,000 |
| High-volume warehouse picking | Rugged handheld terminal | ₹25,000+ |
Start at the top of that table. A phone camera is a perfectly good scanner for anything under a few hundred scans a day, and the marginal cost is zero. Buy a wired scanner when you find yourself standing at the same desk scanning repeatedly.
"Keyboard wedge" means the scanner presents itself to the computer as a keyboard: it types the digits and presses Enter. There is no driver, no configuration and no integration — it works with any software that has a text field, which is why the cheap ones are fine. Avoid anything requiring proprietary software unless you have a specific reason.
SKU vs barcode — keep them separate
These are different identifiers doing different jobs, and merging them causes problems that only surface months later:
- SKU — yours. You choose the format, it means something to your team, and it survives a supplier changing their packaging or their part number.
- Barcode (EAN/UPC) — the manufacturer's, printed on the box. You do not control it. Two of your SKUs can share one if you repack, and a supplier can change it without telling you.
Store both on every product. The practical payoff is bulk updates: a supplier price list keyed to their barcodes can be imported directly, and your own team's count keyed to your SKUs can be imported directly, with no manual matching in between.
Products with no barcode
Loose goods, repacks, own-brand and anything you make yourself. Two workable approaches:
- Print your own. Use the SKU as the barcode value and print Code 128 labels. A basic thermal label printer is ₹6,000–₹12,000 and any label software will encode plain text as Code 128. Do not buy into a proprietary label ecosystem for this.
- Do not barcode them. Every screen that accepts a scan also accepts a search by name, SKU or category. Barcodes make things faster; they are not a prerequisite. Barcode the fast movers, search for the rest.
Option 2 is underrated. Barcoding a 400-line catalogue where 40 lines are 80% of the movement is mostly wasted effort — ABC analysis will tell you which 40.
Where scanning actually pays off
A scanner attached to nothing is a toy. The value is in wiring it into specific moments:
- Receiving. Scan each line as it comes off the vehicle, against the purchase order. Discrepancies surface while the driver is still there, which is the only time they are easy to resolve.
- Stock takes. The biggest single win. Walking a shelf scanning is several times faster than reading and typing, and vastly more accurate. Cycle counting becomes practical rather than aspirational.
- Counter sales. Scanning at point of sale keeps stock live without anyone doing a second data-entry job.
- Picking and despatch. Scanning against the order catches the wrong-item error before it reaches the customer, which is where it becomes expensive.
- Lookups. Scanning to see current stock, price and history is a small thing that gets used dozens of times a day.
A one-week rollout
- Day 1. Audit the catalogue. Which products already have barcodes on the packaging? Capture them — scan each product once into the barcode field.
- Day 2. Decide which unbarcoded products get printed labels. Use ABC: the A items, and anything counted often.
- Day 3. Print and apply labels. Put them where they can be scanned without moving the product.
- Day 4. Wire scanning into receiving. This is the highest-value workflow and the easiest to train.
- Day 5. Run a cycle count on one category, scanning. Compare the time taken with your last manual count; this is the number that convinces everyone else.
- Week 2 onwards. Add point of sale and picking once receiving is habitual. Do not roll out five workflows simultaneously.
In SmartShelfKart the camera works on Android and iOS, keyboard-wedge scanners work on the web, and a scan lands directly in the movement screen you were headed for — so the scan produces a recorded transaction rather than just a lookup.