Convert more visits to installs with store listing experiments
Get more installs by using store listing experiments to find the best app icon, images, video, and descriptions for your app on Google Play.
Why it works
Store listing experiments let you A/B test changes to your store listing on subsets of your users so that you can choose the changes, backed by data, that bring you the most installs.
- Test variants of your store listing's graphics and text.
- Run global tests for graphics alone in your default language, or tests including text and graphics for any localization in up to five languages simultaneously.
- Experiment with the order of your screenshots.
- Select the percentage of visitors Google Play will display an experiment to, up to 50 percent.
- View results in the Play Console.
- Make an experiment's winning content the default store listing.
- Store listing experiment performance can be measured using either first time installer or one day retained users metrics. These metrics are reported hourly, with the option to get email notifications when the experiment completes, so you can quickly apply the best variant.
- Determine a clear objective. Make sure that you've a question or objective in mind when designing your experiments.
- Test one attribute at a time. Get the clearest results by testing one variation of listing content at a time, unless testing completely different designs for a store listing page against one another. Also, test similar changes in icons and feature graphics together, for example, when comparing the use of characters against objects for a game.
- Start testing with your app icon. The app icon can have a huge impact on whether users pick your app from search results and category lists, so test it early.
- Test all of your listing's attributes. After icons, test screenshots and short descriptions as they are seen by users from search results and category lists. Test the feature graphic as a priority, as it has such a presence on your store listing's page.
- Run your experiment for at least 7 days. Run tests for a week to capture both weekday and weekend behavior, even if you have enough traffic to get meaningful results in a few days.
- Run your experiments with 50 percent of your audience. Using the largest allowed test audience will get results faster; particularly for localized tests that may be visible to only a small percentage of your users.
- Create at least three alternatives for testing graphics and text. For your hi-res icon, feature graphic, screenshots, promo video, short description, and full description, test a few different versions to see which performs better, and experiment with the order of screenshots.
- Iterate. Be bold about testing different artwork, but iterate on the results to confirm and improve your findings.