All posts
Analytics22 July 20266 min read

What QR scan analytics actually tell you (and what they don't)

The moment print becomes measurable, it gets over-interpreted. A spike gets treated as proof of something it does not prove.

Scan analytics are worth having. They are also narrower than they look, and knowing the edges makes them far more useful.

What you get

A scan is a request, so anything visible in a normal web request is available, and nothing else.

1

How many scans, and when

2

Country and city, approximately

3

Device, operating system and browser

4

Which page was served, if you use scheduling

Time of day is the most useful number

Totals feel important and rarely change a decision. The shape across the day usually does.

A restaurant that assumes lunch is its busy period and finds the menu gets opened twice as often at four in the afternoon has learned something it can act on.

  • Find the hours that actually matter
  • See which days carry the week
  • Spot a campaign peaking and fading
  • Compare a quiet site against a busy one

Location is rougher than it looks

Location comes from the network the phone is on, not from GPS. On mobile data it can resolve to wherever the network routes through, which is sometimes a city nobody in the story ever visited.

Treat it as a broad signal. Useful for confirming your customers are mostly local. Not useful for anything finer than that.

A scan is not a person

One person scanning three times is three scans. A code scanned by somebody testing it is a scan. A phone that reads the code without anybody opening the page may still count.

So scan counts are best read as a relative measure. This placement against that one, this week against last week. The absolute number matters less than the comparison.

The comparison is where the value is

One code with a big number tells you very little on its own. Two codes doing the same job in different places tell you which one to keep paying for.

That is the single habit worth building. Split by placement from the start, because you cannot separate the data afterwards.

  • One code per poster site
  • Separate codes for packaging and shelf edge
  • A different code for each event
  • Keep last year's code so you can compare

See how this applies to Marketing Teams.

Quick answers

No. You see counts and coarse technical details. There is no identity attached to a scan.