- Phone cameras are good enough for occasional scanning, such as checking in tickets or looking up a product; teams scanning all day usually need dedicated scanners.
- Google's ML Kit scans most standard barcode formats on the device without a network connection, and Apple offers a built-in data scanner on iOS 16 and later.
- Rugged scanning devices and Bluetooth scanners send scans to the app as keystrokes or through vendor tools, so plan the input method early.
- Browser-based scanning is still uneven: the Barcode Detection API is experimental and not available in some widely used browsers.
- Design the app around scanning: instant feedback, duplicate checks, manual entry as a fallback and an offline queue.
Should your business app scan with the phone camera or a dedicated scanner?
Use the phone camera when scanning is occasional: checking in tickets, looking up a product, logging an asset or confirming a delivery. Use dedicated scanners, either rugged scanning devices or Bluetooth scanners paired with phones, when staff scan all day, work with damaged labels or need to scan quickly at a distance. Many apps support both, so office staff use phones and warehouse or counter staff use scanners.
Our mobile app team builds scanning into stock, POS, field service and event apps. This guide covers the choices we work through with clients before development starts.
How good is phone camera scanning now?
Much better than it used to be. Both major platforms now offer solid scanning tools:
- Android and iOS with ML Kit. Google's ML Kit barcode scanning reads linear formats such as Code 128, EAN-13 and UPC-A, and 2D formats such as QR Code, Data Matrix and PDF417. Scanning happens on the device and does not need a network connection.
- Android without camera permission. Google's code scanner lets an app scan codes without asking for camera permission, by handing the scan to Google Play services and returning only the result.
- Apple's built-in scanner. Apple's DataScannerViewController, available from iOS 16, scans the live camera feed for text and machine-readable codes such as QR codes. The app must explain why it needs the camera, and the user must agree when the system asks.
For light use this is enough, and it costs nothing in hardware. The limits show up with volume: a phone camera needs good light, a steady hand and the right distance, and aiming a phone hundreds of times a shift is slow and tiring.
When do dedicated scanners pay for themselves?
Dedicated hardware earns its cost when scanning is the job rather than an occasional step. Consider scanners when:
- Volume is high. Pickers, receivers and cashiers who scan constantly work faster with a trigger and a purpose-built scan engine.
- Labels are poor. Wrinkled, faded, curved or shrink-wrapped labels read more reliably with a proper imager.
- Distance matters. Scanning labels on high racking is impractical with a phone camera.
- Conditions are rough. Cold stores, gloves, drops and dust are hard on consumer phones.
- Battery has to last a shift. Continuous camera use drains a phone quickly.
There are two common forms. Rugged scanning devices run Android with a built-in scanner. Zebra, for example, says its DataWedge service is pre-installed on every Zebra Android device and lets any app on the device collect data from barcode scanners and other inputs. The cheaper option is a Bluetooth handheld or ring scanner paired with a normal phone or tablet, which usually behaves like a keyboard.
Before buying in bulk, run a short pilot. Give two or three staff each option for a normal week of work, time the tasks that matter most, and count missed scans and rescans. Ask them which they would choose. Then add up the full cost per worker: devices, cases, chargers, spare batteries, a device management licence if you need one, and replacements for breakages. The pilot usually settles the question faster than any spec sheet.
Can a web app or PWA scan barcodes?
Yes, but with more caveats than a native app. The browser Barcode Detection API would make it simple, but MDN lists it as experimental, with limited availability, because it does not work in some of the most widely used browsers, and it only runs on secure HTTPS pages. Web apps therefore usually rely on JavaScript scanning libraries, and results vary with the device's camera and browser.
A web app is fine for occasional scans by office staff. When scanning is central and must work offline, a native or cross-platform app is safer. Our guide to PWAs and native apps covers that wider decision.
How does scanned data get into the app?
Decide this early, because it shapes the app's design:
| Input method | How it works | Watch out for |
|---|---|---|
| Camera with an SDK | The app opens a scanner view and receives each result directly | Camera permission, lighting, slower for high volume |
| Keyboard-style scanner | The scanner types the code into the focused field, often followed by Enter | The right field must have focus; on-screen keyboard behaviour |
| Device vendor tools | Software on rugged devices sends scans as keystrokes or as Android intents | Configuration profiles per app, tied to one hardware vendor |
Zebra's documentation describes DataWedge profiles that can be configured manually or programmatically and support mass deployment, which matters once you run more than a handful of devices.
Which barcode types will you be reading?
List the codes your staff will actually meet before choosing devices:
- Product barcodes. GS1 explains that linear barcodes such as UPC and EAN encode the GTIN used to identify products. Most retail stock uses these.
- 2D codes. According to GS1, 2D barcodes hold more data than linear ones while appearing physically smaller, and GS1 is leading a move towards 2D barcodes in retail. Expect more products to carry them.
- Your own labels. Bin locations, assets, job cards and tickets, where you choose the format.
- Supplier and shipping labels, which may contain several pieces of data in one code that the app needs to split.
How should the app be designed around scanning?
Scanning works well when the whole flow is designed for it, not when a scan button is added to a form. Build in:
- Instant feedback: a sound, vibration and clear on-screen result for every scan, good or bad.
- Validation: check the code against what the task expects, such as the right product for this order.
- Duplicate handling: decide whether a second scan adds quantity or is an error.
- Manual entry: a fallback for damaged labels.
- An offline queue: scans saved on the device and synced later. Our guide to planning offline mode covers how.
If the app updates stock in an online store, connect it to the same stock source as your other channels. Our POS development team builds these links so scans, sales and online orders stay in step.
Frequently asked questions
Can staff use their own phones for scanning?
They can, and it keeps costs down for light use. Think about battery drain, privacy and what happens when a phone is lost. For daily stock work, company devices managed through mobile device management are easier to secure, replace and keep on the same app version.
Which barcode type should we print on our own labels?
For internal labels such as bins, assets or job sheets, Code 128 and QR codes are common choices because most scanners and phone libraries read them. Products sold through retailers usually need GS1 barcodes with a registered GTIN, so check your retailers' requirements first.
Do we need a native app for camera scanning?
Not always. Web apps can scan with JavaScript libraries, but browser support and camera behaviour vary between devices. A native or cross-platform app gives more reliable camera control and offline storage, which matters when scanning is the main job.
Can one app support both phone cameras and scanners?
Yes, and many do. Build scanning as one input step that accepts data from the camera, a keyboard-style scanner or a device SDK. Office staff can then use phones while warehouse staff use rugged devices with the same app.
How do we test scanning before rollout?
Test with your real labels in your real conditions: damaged and wrinkled labels, glare, low light, gloves and the distances staff actually work at. Time a typical task with each device and watch for missed or double scans before you buy hardware in bulk.
Last updated