QR Codes on Billboards

The dwell-time test

Before the size question, answer this one: how many seconds is a phone pointed at this board, from a fixed position, within range?

You need about four. One to notice, one to get the phone out, two for the camera to lock on.

• Highway billboard, moving traffic. Zero usable seconds. Do not print a code. Print a short, sayable URL.
• Roadside board at a signalled junction. Traffic stops, but the board is thirty metres away and scanning behind the wheel is illegal in most places. Still no.
• Transit shelter, platform, station concourse. Ninety seconds of standing and waiting, from 1.5 to 3 m. This is where large-format codes earn their keep.
• Shopping centre and airport large format. Walking pace, but people stop for a board that promises something specific.
• Building wrap facing a plaza or terrace. People sit there. Works better than most media planners expect.

The size maths, once the dwell test passes

Same one-tenth rule as every other surface, and at billboard scale it produces numbers that surprise people:

• Scanned from 3 m needs a 30 cm code
• Scanned from 5 m needs a 50 cm code
• Scanned from 10 m needs a full metre

On a six-sheet transit poster of roughly 1.2 by 1.8 m, a correctly sized code takes about a quarter of the width. That is not a design failure. That is the correct proportion, and shrinking it to make the layout breathe is what kills the campaign.

Height matters more here than anywhere else

A code mounted above 2.5 m is photographed from below at a steep angle. The pattern foreshortens, the finder squares stop reading as squares, and decoding fails. Keep the code in the bottom band of the artwork, at standing eye level, even when that fights the layout.

What the scan should open

Not the homepage. A billboard scan is a person on a platform with sixty spare seconds, and the page has to reward exactly that:

• A coupon page with something they will still have when they arrive
• A short video that finishes before the train does
• A multi-link page when the campaign has several destinations

Use a dynamic code so an eight-week booking can change its offer in week five, and so you learn which sites actually produce scans. Out-of-home is normally the hardest medium to measure. One code per site turns it into one of the easiest.

What to do when the dwell test fails

For moving traffic the answer is a short URL a passenger can remember at the next red light. Some campaigns print both: a code for the pedestrians who pass the same board on foot, and a spoken URL for the cars. That is a deliberate design. Printing only the code and hoping is not.