Packaging is the longest-lived print medium most brands produce, and the gap between how long the pack lives and how long the average URL lives is where printed QR codes go to die. Designing the code for time is therefore not paranoia; it is matching the medium.
How long packs actually live
Production to shelf takes months; shelf life adds more; pantries, garages and gift cycles add years; resale and hand-me-downs add years again. A pantry SKU printed today will be scanned in 2029; a appliance label, well into the 2030s. Every one of those scans hits whatever the artwork froze at print time.
What decays first: the web, not the ink
Ink survives remarkably well. What dies is the digital side: the campaign microsite is retired next quarter, the agency lets the vanity domain lapse, the URL shortener changes terms or disappears, the platform the QR vendor ran folds. Every failure mode above has the same root cause, the artwork embedded a destination, and destinations are the most perishable object on the web.
The design that survives: print identity, resolve everything else
A Digital Link prints the product's identity, the GTIN, inside a resolver URL. Identity does not rot: the GTIN means the same thing in 2036. Everything perishable, destination, campaign, page, even recall status, lives behind the resolver as editable configuration. The pack becomes a stable question and the resolver a current answer, which is the only architecture where a decade-old scan lands somewhere correct.
De-risking the resolver itself
Ten-year thinking must also ask what if the resolver goes away. The mitigations are the exit criteria from the selection guide: standard URI syntax so any conformant resolver can answer the same identifiers, exportable configuration, and, for maximum control, your own hostname on the pack so re-pointing DNS re-homes every printed code at once. Onpack supports all three deliberately, and publishes the anti-lock-in commitment in its terms, because infrastructure for printed goods earns trust by making itself replaceable.
Common questions
How long does packaging stay in circulation?
Category-dependent: weeks for fresh food, two to five years for pantry and personal care, a decade or more for durables, appliances and anything resold secondhand.
What kills printed QR codes in practice?
Rarely the ink; almost always the web behind them: retired campaign pages, dead domains, folded vendors. The symbol outlives its destination.
How do I make a code safe to print for a decade?
Print identity, not destination: a resolver URL carrying the GTIN, so the answer stays editable forever, plus an exit-friendly resolver so even the operator is replaceable.
What about codes already printed with dead URLs?
If you control the domain, resurrect it with redirects. If not, the lesson is for the next print run; nothing patches ink.