A QR code on a printed card that people hold in their hands is often the fastest thing on the page. The same code on a flyer that gets photographed and forwarded is dead weight, because the person looking at it is holding the only camera they have and it is already pointing at the code.
That distinction does most of the work. Before the design questions, ask where the piece will be seen, and whether the reader will have a free hand and a second device.
When it earns its place
The test is whether the code removes real typing. A code that saves someone entering a long address, or a payment handle, or a form URL with a string of characters after it, is doing work. A code that points at a domain someone could type in four seconds is decoration.
The cases where it consistently pays off:
- Printed flyers on a wall. A school, church or studio notice that people walk past. They have their phone out already, the board is not going anywhere, and typing a URL while standing in a corridor is genuinely annoying.
- Reply links on a printed invitation. A form address is usually long and unmemorable, and a code turns it into one tap. This is the strongest single use.
- Anything with a payment or donation link. Nobody types those correctly.
- Day-of signage. A menu, a photo-sharing link, a schedule. People are standing, holding a drink, and will not type anything.
The cases where it is clutter:
- A formal invitation where the only destination is your wedding site. A short, readable address prints better and reads better.
- Any invitation delivered as an image to a phone. This is the big one, and it deserves its own section.
- A code that points at a page nobody needs before the day.
Why it fails on a photographed flyer
Print a flyer, pin it to a board, then take a photo of it and send it to the parents group chat. That is what happens to almost every school and church flyer, and it breaks the code twice over.
First, the person now looking at the flyer is looking at it on their phone. The code is on the screen in their hand, and the camera is on the back of that same hand. Some phones can scan a code inside a screenshot from the photo library, and plenty of people do not know that, and a fair number of the codes are too degraded by then anyway.
Which is the second problem. The photo was taken at an angle, in whatever light the corridor had, then compressed by the messaging app. A QR code is a grid of small squares and compression smears exactly that kind of edge. The pattern survives being made a bit soft; it does not survive being made soft and small and slightly skewed at once.
The fix is not to remove the code. It is to make sure the flyer works completely without it. Print the address in readable type, keep it short, and treat the code as the shortcut for the people standing in front of the real paper. If you are designing an event flyer, assume it will spend most of its life as a photograph in a group chat, because it will.
Size, quiet space and contrast
Three technical conditions, and codes fail for one of these three almost every time.
Size. The working rule of thumb is that a code should be about one tenth of the distance it will be scanned from. Held in the hand at ten inches, that is around one inch. On a wall read from six feet, it is closer to seven inches, which is why tiny codes on posters never work. For a printed card, I would not go below about three quarters of an inch square, and an inch is more comfortable.
Quiet space. The specification requires a clear margin around the code, four modules wide, a module being one of the small squares in the pattern. In practice that means a band of empty background roughly four squares deep on all sides, and it must be empty. No border, no illustration creeping in, no line of type tucked against it. This is the most commonly broken rule in invitation design, because designers hate an isolated square and try to tidy it by bringing something close.
Contrast. Dark code on a light background. Not because it looks better, but because plenty of scanners still expect that polarity, and an inverted or low-contrast code fails on a proportion of phones you cannot predict. A very dark ink on cream is fine. Mid-grey on blush is not, however much better it looks on your monitor.
One more thing about the data itself. Codes carry error correction, and higher correction levels tolerate more damage at the cost of a denser pattern. On a printed piece that will be handled, a higher correction level is cheap insurance. And keeping the destination URL short keeps the pattern coarse, which is what actually makes a code scan reliably. A long tracked link produces a dense grid that needs to be printed larger to work.
Always print the URL too
This is not optional. Under or beside every code, in small readable type, print the address it goes to.
Three reasons. Some people cannot or will not scan a code, and a card that only speaks to people who can is a card that loses guests. The printed address is what makes the piece work in a photograph, where the code has already failed. And a visible destination is a courtesy: people are reasonably wary of codes now, and showing where it goes removes a hesitation you never see.
Keep that address short enough to be typed. If the real link is a long form URL, use a shortened one for print. And test the code on the printed proof, not on screen, with at least two different phones, one of them not yours.