How to A/B Test a QR Code on a Poster or Flyer

To A/B test what is printed, such as the design, the call to action or where the poster hangs, print two different dynamic QR codes and compare their scans. To A/B test the page people land on, use one dynamic code that splits scans between two or more destinations, then compare what visitors do after they arrive.

November 12, 20266 min read

Key takeaways

  • Testing the print needs two codes. Testing the landing page needs one code with a split.
  • Both need dynamic codes, which need a free Qravio account. Static codes record no scans.
  • Two separate codes work on every plan, including Free. Splitting one code is on Pro and Agency.
  • Qravio splits website codes by weight across up to five destinations, and a phone keeps its version for 30 days.
  • Scans tell you which print gets noticed. Only what happens after the scan tells you which page converts.

Decide what you are testing

Most QR A/B tests fail because they mix up two questions. "Which poster gets more people to scan?" is about the print. "Which page turns more scanners into customers?" is about the destination. They need different setups, and running the wrong one gives you a number that answers neither.

You want to compareSetupWhat you measureQravio plan
Two poster designs or calls to actionTwo codes, one on each versionScans per codeAny, including Free
Two placements (counter vs window)Two codes, one per spotScans per codeAny, including Free
Two landing pages or offersOne code split between destinationsConversions per destinationPro or Agency

Test the print: one code per version

When the thing you are changing is physical, each version needs its own code, because a single code cannot tell which poster it was printed on. Make two dynamic codes that point to the same page, put one on each version, and compare scans in the dashboard.

Keep everything else equal. Hang both versions for the same days, in places with similar foot traffic, at the same size. If one sits by the entrance and the other by the toilets, you are testing the wall, not the design.

  1. 1Sign in to Qravio (dynamic codes need a free account) and create a website QR code for your page.
  2. 2Duplicate it, or create a second code with the same destination, and name each one after its version, for example "Poster A, red headline".
  3. 3Print each code on its own version and put them up at the same time.
  4. 4Compare total and unique scans per code after a week or two, then keep the winner and reuse the code that lost for your next test.

Test the page: one code, split destinations

When the print stays the same and you want to know which page or offer works better, use one code that sends a share of scans to each destination. On a Qravio website code this is A/B split testing, in the code’s settings on Pro and Agency.

You add up to five destinations and give each a weight. Equal weights split scans evenly; weights of 80 and 20 send most traffic to your current page and a small share to the new one, which limits the risk while you test. Each phone is assigned a version on its first scan and keeps it for 30 days through a cookie, so someone who scans twice sees the same page both times.

You can pause a destination or change its weight at any time without reprinting, since the printed code never changes.

Start with two tracked codes

Create a free account to make dynamic codes and compare their scans. Split testing one code is on Pro.

Open the dynamic generator

How to tell which version won

For a print test, scans are the result: the version with more unique scans got more people to act. Judge it on unique scans rather than totals, so one curious person scanning five times does not decide the test.

For a page test, scans are not the result. The split decides how many scans each page gets, so a 50/50 split gives you roughly equal scans whatever the pages say. Qravio’s A/B results show scans, unique visitors and traffic share for each destination, which confirms the split is working. The winner is the page that turns more of those visitors into sign-ups, bookings or sales, and you measure that where the conversion happens.

The simplest way is to make each destination a different URL, such as /offer-a and /offer-b, and compare conversions per landing page in your own analytics. Adding a campaign tag to the code labels all of its traffic in GA4 too.

How long to run it

Small tests are noisy. Twenty scans against fifteen is not a result; it is chance. Let the test run until each version has at least a few hundred scans, or for a page test a few dozen conversions each, and cover full weeks so weekday and weekend traffic are both in it.

Change one thing at a time. If version B has a new headline, a new colour and a new offer, a win tells you B is better but not why, and you cannot carry the lesson into the next poster.

Frequently asked questions

Can I A/B test a QR code for free?

Yes, for print tests. Qravio’s Free plan includes five dynamic codes with scan tracking, so you can put a different code on each version and compare scans. Dynamic codes need a free account. Splitting one code between destinations is on Pro and Agency.

Can I A/B test with a static QR code?

No. A static code sends the phone straight to the destination, so there is nothing in between to count the scan or choose a version. A/B testing needs a dynamic code.

Will someone see a different page each time they scan?

No. On a Qravio split test, a phone keeps the version it was first given for 30 days. A different phone, or a browser that clears cookies, may be assigned again.

Which QR code types support split testing?

On Qravio, splitting one code between destinations works on website codes, the type that redirects to a URL. For other types, such as a menu or a business page, run the print-style test with two codes instead.

Does Qravio pick the winner for me?

It highlights the destination with the most scans, which is a useful check on the split but not a verdict on which page converts. Compare sign-ups or sales per destination in your own analytics to call the winner.

Start with two tracked codes

Create a free account to make dynamic codes and compare their scans. Split testing one code is on Pro.