Unique scans

A redirect service never learns who scanned, so “unique” has to rest on a stand-in for a person. The common ones are the IP address (or a hash of it), the IP address combined with the user agent, or a cookie set by the redirect domain. Each errs in its own way:

• Undercounting: many people can share one public IP address, such as everyone on an office or venue WiFi network or behind a mobile carrier’s shared gateway.
• Overcounting: one person can show several addresses by moving from WiFi to mobile data, using a VPN or getting a new address from their carrier. Cookies are lost in private browsing and differ between browsers and in-app scanners.

The time window matters too. Unique scans per day, per month and for a whole campaign are different numbers, and adding up daily uniques overstates the campaign total.

Total scans measure activity; unique scans estimate reach. The gap between them shows repeat use. A loyalty card code scanned by regulars every week has many scans per person, while a poster at a train station mostly gets one scan per person.

There is no standard method, so unique counts from two providers are not comparable, and switching tools mid-campaign breaks the trend. Use unique scans to compare codes inside the same tool and period, and treat the absolute number as an estimate.

Say a café’s table code shows 1,200 total scans and 800 unique scans in a month: about 1.5 scans per scanner, which suggests many guests come back or scan more than once per visit. Guests on the café’s WiFi usually share one public address, though, so the real number of people is probably higher than 800.