CaddieThe bottom line
Swapping the paywall image from a Greek building to immigrants arriving on a boat raised conversion rates by 13% for Neos Kosmos. The visual on a paywall can change how many people convert.
Key takeaways
- The paywall visual influenced conversion. The two images were the only change tested, and the results differed.
- The variant lifted conversion rates by 13%.
- The image of immigrants on a boat beat the image of a Greek building. It appears to have resonated more with users.
- The control showed a Greek building.
- The variant showed immigrants arriving on a boat.
The test
Only the paywall image changed between the two versions
| Control | Variant | |
|---|---|---|
| Paywall image | Greek building | Immigrants arriving on a boat |
| Everything else | Unchanged | Unchanged |
| Conversion | Baseline | +13% |
| Outcome | Lost | Won |
Neos Kosmos compared two paywall images to see which would convert better.
- The goal was to find which visual resonated better with users.
- The test sat in the paywall and free trial category.
- The control displayed a Greek building.
- The variant featured an image of immigrants arriving on a boat.
The result
The boat image converted 13% better than the building
Greek building (control)100
Boat arrival (variant)113
The boat image produced higher conversion rates than the building.
- Conversion rates rose by 13%.
- The result was first written up by Media Makers Meet in a piece on publishers A/B testing their paywalls.
In the growth library
Neos Kosmos, Paywall & free trial test. Neos Kosmos tested two different images on its paywall to see whether the visual alone changes conversion. Swapping a Greek building for an image of immigrants arriving on a boat lifted conversion by 13%.
Lessons that draw on this:
More from 
abtest.design · The whole library, and ask it anything