Take the QR code off your own product packaging and decode it, but don't follow the link yet. Just read the address that comes out. If it sits on a domain your company owns, that code keeps working no matter what happens to the tool that produced it. If it sits on the generator's domain, the printed code is exactly as durable as that company.
That one distinction accounts for most of the vendor risk in QR codes, and you can check it in about a minute per code.
Static carries the destination, dynamic borrows it
A static code is the URL. The pattern of modules encodes your address directly, so nothing sits between the phone and your site. The generator's involvement ended when the image was exported. It can shut down, get acquired, or delete your account, and the flyer in the field is unaffected. What can still break a static code is your own destination: rename the page and the code points at a 404. That is the honest version of the do QR codes expire question.
A dynamic code encodes a short address on the vendor's redirect domain. A phone hits that address, the vendor's server answers with a redirect, and the visitor lands wherever you configured. The indirection is the whole feature, since it buys editable destinations and scan analytics. It also means three things have to stay alive for the code to resolve: the vendor's domain registration, their redirect service, and your account in good standing. The full trade is laid out on the dynamic code page.
Shutdown is not the only way this goes wrong
A trial ends and the codes stop resolving. A downgrade freezes every code above the new plan limit. A card expires, the account suspends, and redirects start returning an error page. An acquisition retires the old short domain and migrates only the customers who filled out a form. Or the service stays perfectly healthy and adds an interstitial page between the scan and the destination.
To a customer standing in front of your poster, all of these look identical: nothing happens, or a stranger's error page appears.
Reading what your codes actually contain
Point a scanner at the code and read the URL before opening it. Most scanner apps preview the destination first; if yours opens links immediately, use a different one. For codes that exist only as artwork files, upload the image to an image reader and read the decoded string.
Three outcomes:
- Your own domain. Either static, or dynamic through a custom domain you control. Both are in decent shape.
- A short third-party domain you don't recognize. Dynamic and vendor-bound. Write it down.
- Your domain, but a path nobody remembers creating. Someone routed through a shortener. Find out whose account it is before that becomes a surprise.
Do this on an actual printed piece, not the file sitting in the design folder. Reprints get pulled from stale artwork more often than anyone likes to admit.
The audit is one afternoon and one spreadsheet
One row per printed asset. Columns:
- what the asset is and where it lives: packaging, vehicle wrap, signage, business cards, manuals
- how many are in circulation and how long they stay there
- the decoded URL, copied verbatim
- static or dynamic
- which vendor, whose account, and who owns the billing
- cost and lead time to reprint
The rows that deserve attention are the ones that combine a long asset life, a high reprint cost and a vendor-owned redirect domain. Vehicle wraps. Engraved plates. Packaging with two years of inventory sitting in a warehouse. A menu insert or an event badge scores badly on none of those, so it barely registers as risk.
Migration means reprinting, and there is no trick around it
If a redirect domain stops resolving, you cannot repoint those codes. You don't control DNS for a domain you never owned, and there is no editing ink that dried in 2024. Recovery means new codes, a new print run, and physically swapping every deployed asset. The generator is the cheap part; the labor and the paper are the bill.
What you can do while the vendor is still running: export the mapping between every short code and its destination and keep that file somewhere outside their dashboard. Store the encoded URL alongside the source artwork for each design. If a shutdown notice ever arrives, you will already know which assets are affected and what each one is supposed to point to, which turns a scramble into a scheduling problem.
Hedge one: run redirects on a domain you own
If your generator supports custom domains, use it. Stand up something like go.yourbrand.com, point it at the vendor with a CNAME, and issue every dynamic code on that hostname. The printed code now contains an address you control. Change providers later and you re-point DNS, recreate the same paths at the new vendor, and the codes in the field carry on working.
Two conditions make it real. The domain has to stay registered, which means auto-renew enabled in a registrar account owned by the company rather than by whoever set it up three jobs ago. And you have to keep that export of paths and destinations current, because recreating the paths is the part that actually saves the codes.
Hedge two: static for anything permanent and printed at volume
Where the destination genuinely will not change, static encoding removes the vendor from the equation entirely. Product manuals, a wifi placard, machine instructions, a link to your homepage. A static code has no service behind it and nothing to lapse, which is the point of picking a generator with no expiration in the first place.
You can keep some flexibility without taking on a dependency. Encode a static code that points at a path on your own site, say yourbrand.com/go/manual, and let your own server redirect that path onward. Both ends belong to you. Change the target by editing a redirect rule, no reprint required. You give up the vendor's analytics dashboard and get server logs instead, which for most teams is a fair swap.
Sizing and quiet-zone specs live on the print guidance page. For anything heading to a printer this quarter, settle static versus dynamic before artwork approval rather than after, and put the answer in the file name so the next person to open it doesn't have to guess.
