Why Do QR Codes Still Work When Damaged?
A scratched code is not necessarily a broken code
You have probably seen a QR code on a worn menu or a slightly creased label still open the right page. That can look like the scanner is guessing. In fact, the code carries extra information that helps recover damaged data. Understanding the difference matters when you design a label, add a logo or troubleshoot a code that refuses to scan.
What error correction actually does
QR codes use Reed–Solomon error correction: the original data is accompanied by mathematical redundancy. Within the decoder's limits, that redundancy can correct errors rather than merely notice them. Think of it as built-in recovery information, not a backup photo of the QR code. There are four levels, L, M, Q and H; choosing a stronger level provides more recovery capacity but also increases the data that must fit into the symbol.
Why “30% damage” is an easy claim to misread
DENSO WAVE describes recovery relative to codewords—the data units used by the encoding—not a promise that any chosen portion of the printed square can be erased. Damage location matters, and the scanner must first find and interpret the symbol. A logo that happened to work on one sample is therefore not evidence that every similarly sized logo is safe. A successful scan is a test result for that sample and setup, not a universal design rule.
The blank border is doing real work
A standard QR code needs an unprinted quiet zone at least four modules wide on every side. A module is one small square in the pattern. Putting a decorative frame or caption against the edge uses space the reader needs to distinguish the code from its surroundings. Keep that border when exporting, cropping and placing the image in a poster. The final printed layout is what matters, not just the clean preview in your editor.
More correction is not always the best first fix
Suppose a café menu code works on a laptop screen but fails on a small printed card. Increasing error correction alone may not solve it. More encoded material can create a denser pattern, leaving smaller squares when the overall print size stays fixed. DENSO WAVE recommends considering module size and the reader's capabilities together. First try giving the code more physical space and preserving sharp edges. Test the actual paper, print method and expected scanning distance.
A practical checklist before you publish
Use a clear dark-on-light design. Preserve the four-module border. Avoid stretching the square or shrinking it until the small squares blur together. Print one sample at its intended size, then try it with several phones under the lighting where people will use it. If it fails, change one variable at a time—size, border, print quality or surface—so the next test tells you something useful. This workflow is a practical recommendation, not a guarantee for every scanner.
Readable does not mean trustworthy
Error correction protects readability; it does not prove that a destination is safe. A perfectly readable code can still point to the wrong page. Before sharing a code, open it yourself and confirm the destination. When scanning someone else's code, review the address before entering credentials or payment details. For a café or public notice, also make the intended destination understandable in nearby text so users have context.
The takeaway
A damaged QR code may survive because recovery information was included at creation. But dependable scanning is a whole-system problem: encoding, border, print size, image quality and reading conditions work together. Spend less time chasing a damage percentage and more time testing the finished design where it will actually be used.
Sources and reading notes
Original explanatory writing, with the technical points checked against the QR code developer's documentation. The checklist and café example are practical applications, not claims of a measured test.