ASO Case Study
Part I: CVR & A/B Testing
In order to improve the conversion rate from App Store page visits to install, we decided to update the CodeCheck App Store presents. The baseline CVR was already pretty good with 49% (yes, we are ambitious – but also: we crafted the current screenshots, that’s why the baseline CVR was so high) but there is always room for improvement. So, time for a little refresher!
So, here are the screenshots they were using...
In order to improve the CVR, optimizing the screenshots is a powerful leverage factor. We at Customlytics like to combine creative genius with facts and science. That’s why we started off with some market research before kicking off the project. This started with an analysis of the app’s brand, its benefits for the users, the users’ favorite or most engaging features, the competing apps that are also fighting for the users attention, and current trends in the market. This provided us with a good understanding of the opportunities for ASO as well as the relevant insights into how an ideal screenshots concept for CodeCheck may look like.
Subsequently, we found out most of the app’s users are female, they care equally about food and drink products and cosmetics, as long as they are sustainable (which aligns well with CodeCheck’s rather unique value proposition of “shop better, live better”). In addition, we found that users love personalization (e.g. with a customizable profile), and are paying great attention to the topics of microplastics, palm oil and water consumption. In the end, we applied these insights in the conceptualization and storyboarding of the new screenshots and began the production.
...and here are the newly updated screenshots!
After submitting the new screenshots, we could observe a positive impact on the CVR. As seen below, the app’s CVR was at 49% with the original version, then it went up to 52% in the month where the new screenshots were live. However, we would never be convinced without a test, so we conducted an A/B test to see if we could validate the new screenshots.
The solution (Part II):
Another variant of Codecheck app store screenshots was created based on the new concept yet with a different creative direction. This is what would have been the original version (AKA control variant) should it had been completed earlier. So, we used our version as the control variant, and this new version (not created by the Customlytics design team) as a reverse-test variant. Let’s see what we found:
CVR Results (Part II)
The overall concept is very similar to our version, but the colors, footage and device designs are different. The result? We validated our version:
Here’s what we observed: The app’s CVR dropped by an amount matching the test result! And this makes sense: 50% of real users saw the losing variant, its weaker conversion power means fewer of these users got converted compared to the winning one. So, now we are clear: our screenshots were an absolute win!