Part one: cost the three cases before deciding how many units
The first is labor.In a mid-sized supermarket staff are asked about prices dozens of times a day, and each interruption costs 1 to 2 minutes. Let the machine answer the repeat price questions: what you save is not a salary, but the hours that go back into replenishment, display work and real customer service.
The second is price errors.When the shelf price differs from the system price, the outcome ranges from complaints to regulatory fines. A price checker shows the price held in the system, so a pricing error surfaces before the customer reaches the checkout.
The third is retrofit cost.An electronic shelf label covers the whole store and is priced per label; a price checker is priced per location. For stores where labels are not yet rolled out, or where the shelf structure does not suit them, price checkers fill the gap.
Conclusion:Map the spots where customers stop to look for a price first, then size the order by the number of locations rather than by store floor area.
Part two: size the model by location, not one model for the whole store
Inside one store, three kinds of location should not share one configuration:
| Location | Screen | Models | Reason |
|---|---|---|---|
| Shelf endcaps, freezer sides, narrow pillars | 5″ / 8″ | XT802C / XT802D | Limited width, shopper stands close, a small screen is enough and costs less |
| Main aisle, category entrance, beside the pharmacy counter | 10.1″ | XT802E / XT802E-X86 | Fits product images and several lines of specifications, readable from 1m |
| Store entrance, escalator landing, outside fitting rooms | 15.6″ | XT802F / XT802F-X86 | Readable from a distance, suits promotions and wayfinding |
| Service desk, where a printed slip is needed | 21.5″ / 32″ | XTCM215 / XTCM320 | Large screen plus printer or ID module, handles combined self-service tasks |
The system choice comes down to one question: where does the software run. A web page or Android app means the Android model; a Windows client means the X86 model. Mounting holes are identical on both, and changing system later means replacing the whole unit, so decide at order time.
Part three: five bench tests, skip none
Spec sheets do not show how a unit actually behaves. Test these five things:
- Scan real products: take the 20 best-selling items in the store, scan each 20 times and record the failures. Watch for soft produce packaging, metal cans and laminated labels, which are the hard cases.
- Check the screen under trading lighting: check once by day and once at night. A spotlight aimed at the scan window lowers the read rate, so adjust the angle or add a shield.
- Run 8 hours of continuous power: feel the back of the housing and watch for thermal throttling or crashes. Passive cooling models need this test most.
- Disconnect the network: unplug the network cable and check whether local cache still answers and whether the interface clearly shows offline status.
- Try different users: have colleagues around 155cm and 180cm tall each try it once, then check whether they have to bend or look up and whether the scan window falls naturally to hand.
Part four: deployment, six decisions before go-live
| Decision | Options | How to decide |
|---|---|---|
| Power | POE / DC 12V / built-in 220V | Use POE where a POE switch exists, otherwise DC; large-screen models take 220V built in |
| Network | Wired / WiFi | Wired first, since WiFi drops in public areas; plan IP addresses for wired units |
| Data source | Local price list / store system API | Local list if the unit only shows prices; connect to the system for member prices or stock |
| System lockdown | Dedicated launcher / restricted account with UWF | Launcher on Android, restricted account on Windows |
| Scheduled restart | Once a day, overnight | Set in the OS to limit memory fragmentation from long run times |
| Monitoring | Device online status / query log | Know which unit dropped offline, rather than hearing it from a customer complaint |
Part five: five common pitfalls
- Buying on price and ending up with too small a screen: if the text is too small, customers still ask staff and the spend is wasted. Set the size from the viewing distance first.
- Mounting directly under a spotlight: reflections on the scan window visibly reduce the read rate. Test once under the store lighting after installation.
- Enclosing the unit in a sealed light box: heat cannot escape and the unit crashes in summer. Leave 3 to 5cm behind the housing.
- Leaving the system unlocked: customers can exit the app, reach the desktop or even change settings. Configure kiosk mode before going live.
- No spare units: one failure breaks the whole service line. For multi-location rollouts, keep 1 to 2 spare units.
Part six: acceptance checklist
| Item | Pass criteria |
|---|---|
| Housing and mounting | No scratches on the housing, stand firmly fixed, no wobble when pushed lightly, no exposed cables |
| Screen | No bright or dark spots, touch responds at corners and edges, brightness suits the site |
| Scanning | 20 consecutive reads of real store products all succeed, and codes on phone screens also read |
| Software | Results match the checkout, offline state is clearly flagged, and exiting returns to the query page automatically |
| Network | IP fixed or reserved in DHCP, and the unit reconnects after a restart |
| Documentation | Manual, warranty terms, system recovery method and interface notes all supplied |
Part seven: common questions
How exactly do I run the five bench tests from the buying guide?
Work through five steps in order: first, scan real store products 20 times each and log the read rate; second, check screen glare under trading lighting; third, run 8 hours of continuous power and watch for throttling or a hang; fourth, disconnect the network and check whether the local cache answers; fifth, have three staff of different heights operate it and note whether they must bend or look up.
What does one price checker cost roughly?
Price depends on screen size, system, power method, optional modules and quantity, so it varies widely. Send us your requirements and we issue a formal quotation against the configuration. We do not post a single blanket price.
Can I buy one unit to try first?
A sample order can be arranged. Confirm the sample before volume, and the sample can serve as the first installed unit. Confirm the process with sales.