You cannot improve a placement you cannot measure
A QR code on a poster, a package, and a table tent may all point to the same page, but they do not perform the same. Without tracking, every scan looks identical and the budget keeps flowing to placements that never earned it. The good news is that useful tracking does not require a paid platform. A static code plus a tagged link gives you per-placement numbers in the analytics you already have. This guide shows the three practical methods, what each one can and cannot tell you, and how to read the results without fooling yourself.Setting up scan tracking with tagged links
The tagged-link method works with any static code and any ordinary website analytics. It takes ten minutes per placement.- Give every placement its own URL Start from the same destination page, then add tracking parameters for each location: one version for the window poster, one for the counter card, one for the packaging. The page is identical; the tags differ. - Build the tagged URLs carefully Use a consistent naming scheme for source, medium, and campaign, such as source=window-poster, medium=qr, campaign=spring-menu. Write the scheme down, because a tag typed two ways becomes two placements in your reports. - Encode each tagged URL in its own code Generate a separate static code per placement and label your files clearly. The codes look identical, so the file name and a note on the printed piece are the only record of which is which. - Read scans as sessions in analytics Each tagged arrival is a scan that led somewhere. Compare placements by arrivals, then by what those visitors did next: time on page, clicks, orders, or sign-ups. - Review monthly and move codes Once a month, rank placements by results, not by scans alone. Move or remove the bottom performers and give their budget to the placements that produce actions.
Reading scan numbers honestly
Scan counts flatter easily. These habits keep the numbers tied to reality.- Judge a placement by completed actions per scan, not raw scans. A busy corridor produces scans; a table card produces orders. - Compare like with like: same offer, same season, similar footfall. A December number and an August number measure the calendar as much as the code. - Expect some scans to bounce instantly. Accidental scans and curious scans are normal; a placement is only failing when its action rate, not its bounce rate, trails the others. - Count offline conversions separately where you can, for example a till code word shown on the landing page, so web analytics is not your only witness. - Keep the tags stable across redesigns. Changing the naming scheme mid-year breaks every trend line you have built.