How long should I wait before comparing sites?

Until each site has had its full booked period, or at least a matched number of days. Two days in is too early, and daily footfall varies a lot site to site.

Is a roadside poster worth putting a QR code on at all?

Usually still worth it for the few who scan later, once parked or at home, but don't judge a drive-past site on scan volume the way you would a pedestrian one; it's largely a brand-exposure booking.

Can I compare scans across sites booked for different lengths of time?

Not directly. Compare bookings of similar length, or normalise to scans per day in the ground before comparing.

Do I need Pro to run a multi-site poster test?

Free covers 3 dynamic codes; a test spanning more than three sites needs Pro (unlimited dynamic codes), and the device and region breakdown helps sanity-check scan quality on out-of-home placements.

QR Codes for Poster Sites: Which Location Actually Works

Create a QR code

QR codes on poster sites, one code per physical location

Book six poster sites for the same creative, a bus shelter, a railway underpass, two retail park panels, a community noticeboard and a gym corridor, and a single shared QR code can only tell you the campaign as a whole got scanned some number of times. It can't tell you the gym corridor outperformed the underpass three to one, because from the code's point of view they're the same code. Give each physical site its own dynamic code instead, same creative everywhere, and location becomes the thing you're testing rather than the thing you can't see.

The variable here is the site, not the design

This is deliberately narrower than testing two headlines against each other. Keep the poster artwork identical across every site and vary only which physical location it's in, so any difference in scans can be attributed to placement, not to a mix of placement and creative changing at once.

Naming codes by site, not by creative

Label each code by where it is, “Bus Shelter, High St”, or a panel reference from the media owner if one exists, rather than a generic batch number nobody will remember by the time results come in. A scan report that reads as a list of places is one you can act on immediately; a list of codes numbered 1 to 6 is not.

Footfall type: commuters, shoppers, gym-goers, or passing traffic that never stops
Dwell time: a bus shelter gets looked at for minutes, a roadside panel for seconds at most
Reading distance and angle: does the code need to be readable at speed, or only up close
Lighting: is the site lit at night, or does it go dark and effectively invisible after dusk
Booking length: a week-long site and a month-long site aren't comparable on raw scan counts

Code size has to survive the site, not the mock-up

A poster is read from a distance, not held in the hand, so size the code to how the site is actually approached rather than to how balanced it looks in the design file. The print size guide covers the underlying method, distance to minimum module size, but the practical point for poster sites is simpler: a code that looks fine on a laptop screen is routinely too small for a roadside panel seen from a moving car.

Site type	Typical reading distance	What that means for the code
Bus shelter, pedestrian panel	Under 2m, often stationary	A modest code is fine; keep contrast strong for after dark
Retail park or roadside panel	5-15m, often from a moving vehicle	Needs to be large and simple; most passers-by realistically won't scan a roadside panel at all, and that's fine, it's still doing a brand job
Community noticeboard, indoor corridor	Under 1m, stationary	A small code is fine; use the freed-up space for a stronger call to action instead

Give it time before judging a site

Bookings run days to weeks and footfall isn't even day to day, so comparing after two days will mislead. Wait for a site's full booking period, or a matched number of days across sites, before reading anything into a scan count. A site booked for a week and one booked for a month are not directly comparable on raw scans, only on scans per day booked.

A quiet site is not always a failing site

A roadside panel mostly exists for repeated brand exposure to drivers who will never realistically stop and scan. If that's the site's job, judge it on the media owner's impression data or a brand-recall measure, not on QR scans. Save the scan comparison for sites where scanning while looking at the poster is genuinely plausible, pedestrian, static, close.

Running a multi-site poster test

1
List every booked site
With its type (pedestrian, roadside, indoor) and its booking dates.
2
Generate one dynamic code per site
Labelled with the site name or panel ID, same destination and creative otherwise.
3
Size the code to the site's realistic reading distance
Not to how it looks in the poster mock-up.
4
Group the codes in one folder
So the whole booking stays together in the dashboard.
5
Compare only within the same site type
And only over a matched number of days in the ground.
6
Check device and region on Pro if volume looks odd
A spike of unlikely roadside scans is worth a second look, not an automatic celebration.

What plan a multi-site test needs

Free includes 3 dynamic codes across the account, which covers a small test; six booked sites need Pro (£9/month) from day one. Pro's device and connection-class breakdown is also worth having here specifically: it can help sanity-check whether scans on a roadside code look like they came from someone on foot or something else entirely, which matters more for out-of-home placements than most other uses on this site.

No. One dynamic code per site, same creative and destination otherwise, so scans can be attributed to a specific place rather than blended across the whole booking.

Part of Campaign & print attribution

QR codes on poster sites, one code per physical location

Naming codes by site, not by creative

Code size has to survive the site, not the mock-up

Give it time before judging a site

What plan a multi-site test needs

Related