Last updated: 2026-11-15 · By the ZhiShuYun Research Team
Choose anti-counterfeit software by the strength of its record, not the look of its consumer page. A real system issues unique per-unit codes, captures where and when they were applied, records every scan as an event, and can prove afterwards that the record was not edited. If two vendors both show a QR code that opens a nice page, the difference is entirely in code issuance control, event capture and tamper-evident storage.
| Core mechanism | Unique per-unit identifiers, not one QR code per SKU |
| Minimum viable record | Issue event, apply event, scan event, each with time, actor and device |
| Tamper evidence | Hash chaining or on-chain anchoring so history cannot be silently rewritten |
| Operational dependency | Code application quality — an unreadable code is a system failure, not a printing issue |
| Typical failure | Systems that stop at the consumer verification page and never record production or channel events |
Authentication answers one question: is this unit genuine? Traceability answers a different one: where has this unit been? Many products sold as anti-counterfeit systems only do the first — a code that resolves to a web page saying the product is authentic.
That check is weak on its own. A counterfeiter who copies one valid code onto a thousand packages produces a thousand units that all pass the page check. What breaks the copy is the record: a code that was issued once, applied once, and scanned in a pattern that does not match a real distribution route.
First, code generation under your control: codes issued in ranges, bound to a product and a batch, with a documented issuance log. Second, application capture: a record of which code went onto which unit at which line, so a code cannot exist without a production event.
Third, scan event storage with device and location context. Fourth, exception handling: what the system does when a code is scanned the second time, in an unexpected region, or from an unexpected device fingerprint. Fifth, tamper-evident persistence — hash chaining or anchoring — so the record stands up if challenged. Sixth, export: you must be able to take your data out, which is the real test of whether you own it.
Ask what happens when the same code is scanned twice from two countries within an hour. Ask to see the issuance log for a specific code and the production event that created it. Ask whether the scan history can be edited by an administrator, and if so, whether edits are visible.
Ask how codes are ordered and delivered: a CSV download of codes with no association to a production run is a weak starting point, because the association has to be reconstructed later from spreadsheets. Ask what the system does when the printer drops out mid-run and the line keeps moving.
Pricing models differ in ways that matter more than the headline rate: per-code fees, per-scan fees, platform fees, and charges for line integration or label printing. A low per-code price with per-scan billing can invert the economics once consumers start scanning.
Separate the one-off costs — integration, printing equipment, packaging changes — from the recurring ones. The one-off costs are usually the larger number in year one and the recurring costs decide year five, so evaluate both against the volume you expect in the third year, not the first.
Only if the code is unique per unit and the platform behind it records events. A single QR code printed on every unit of a SKU is a marketing link, not an anti-counterfeit control.
Not necessarily, but you do need tamper-evident records. Hash chaining inside a conventional database gives much of the same guarantee at lower operational cost; on-chain anchoring matters when a third party must verify without trusting you.
Two components: per-code or per-unit platform fees, and one-off integration and printing costs. Ask for a three-year total including integration, labels and exception handling, not a per-code rate alone.
Yes — a standard camera-scan of a QR or Data Matrix code opening a web page is the usual model. Requiring an app dramatically reduces verification rates.
They offer different properties — no line-of-sight requirement, harder to copy — at higher unit cost. Many programmes use printed codes on consumer units and RFID at carton or pallet level.
Evaluate the record, not the demo
Ask for the issuance log, the production association, the exception process and the export. A vendor that can produce all four has a system; one that can only show a consumer page has a landing page.