I have watched a lot of QR scans come in from mail pieces, and the difference between a card that gets scanned and one that gets ignored is rarely luck. It is a handful of small, boring decisions made before the file ever reaches the printer. Here are the nine that matter most.
Get the physical details right
1. Print at three quarters of an inch or larger. Phone cameras are forgiving, but arm's length scanning of a small code is not a test you want to run at 5,000 households. I treat 0.75 inches as the floor and one inch as the comfortable default.
2. Protect the quiet zone. The white margin around the code is part of the code. Busy backgrounds and tight crops are the most common reason a technically valid QR fails in the wild.
3. Keep it clear of the address panel. On a postcard, the carrier owns part of the back for the address block and postage. A code that drifts into that zone can be overprinted or obscured. This is why Relay's editor blocks placements in the keep-out area instead of letting you find out after printing.
4. Test on paper, not on screen. Codes that scan beautifully from a monitor can fail on matte stock in dim kitchen light. Print a proof and scan it from three feet away. Every time.
Give people a reason to scan
5. Caption the code. "Scan for your free home valuation" outperforms a naked square. The code is a doorway; the caption is the sign on the door.
6. Land them somewhere specific. A QR that opens your generic homepage wastes the scan. Point it at a page that continues the exact offer on the card. Adoption is not the bottleneck anymore: roughly seven in ten consumers scanned a QR code in the past year. The bottleneck is giving them a destination worth arriving at.
7. Make it mobile-first. Every QR visitor is on a phone by definition. If the landing page has a tiny form or a desktop layout, you paid postage to show someone a pinch-to-zoom exercise.
Make every scan count as data
8. Use one code per recipient, not one per campaign. A shared code tells you that someone responded. A per-recipient code tells you who, which unlocks suppression, follow-up, and honest list evaluation. I covered why this changes the economics in my direct mail attribution piece.
9. Route scans through a redirect you control. A tracked short link logs the scan and forwards in about 50 milliseconds, invisibly to the scanner. That redirect is your measurement layer, and it means you can update the destination after printing if a page moves.
None of these nine is difficult. Together they are the difference between a pretty design element and the best response tracking direct mail has ever had.