Airbnb A/B Tests And Rejects Full Height Photos On Their Landing Page
Airbnb ran an experiment where they increased the size of the header photo on their Host signup landing page. And interestingly it was rejected a month later.
A - Aug 1, 2019 Screenshot
Highlighted UI Changes From This Leak
In the B variation we can clearly see an increased image height of the hero / header image. From what I remember, the image height was either equal to or more than the height of my screen. This might have created an intentional false bottom - whereas in the control, it was clearer that the page contains additional content below.
0.5 Repeatability has been assigned to Pattern #58: Full Height False Bottom as evidence that it's getting better
Repeatability is a net count of evidence for or against a pattern. It’s how we can predict which patterns are better than others. :)
For some reason, the experiment also introduced a more transparent navigation with changed link colors. The B variant also lost the duplicate "Get Started" button in the top right - although I believe it reappered on scroll.
I think it's too bad that these additional changes were introduced, as the photo change would have been a super nice isolated experiment on its own without any additional confounding variables.
Ankit 3 years ago ↑0↓0
Aren't you guys missing the Header? It must have played an important role in conversions also i agree on the image part.
Christian Hamann 4 years ago ↑1↓0
Thats just bad art direction. Variation A shows two happy people and variation B is someone selling a carpet.)
Jakub Linowski 4 years ago ↑0↓0
Agreed. This extra piece of carpet could raise more questions than it provides answers. :)
Ivan Burmistrov 4 years ago ↑1↓0
Thank you for this leak! For me, this experiment shows that a smaller staged photo is less harmful than a larger staged photo. I agree with Vern B that replacing a staging photo with something more neutral can help.
Vern B 4 years ago ↑1↓0
I can see exactly why this failed. What did the extra image provide me? Did it give me a better sense of happiness or drive me to want to rent the location? No - all it did was take the focus off the people in the image and deemphasis the happiness the two people had. Now if they do the same test again with an image that doesn't loose focus - they might have a different result.