App Store product page optimization: how Apple's A/B tests work, and how to read them.
Apple lets you test up to three alternate product pages against your original. The rules differ from Google Play's in ways that change how you plan a test.
If your app is on both stores, you can test on both. Just don't plan an Apple test the way you plan a Google one.
How it works
| Setting | Apple product page optimization |
|---|---|
| Treatments | Up to three, against your original page |
| What changes | App icon, screenshots, app previews |
| Duration | Up to 90 days, or until you stop it |
| Tests at once | One |
| Traffic | You choose the share; it is split equally across treatments |
| Consistency | A user sees the same treatment for the whole test |
| Localization | All your languages, or chosen ones |
| Review | New metadata must be approved by App Review |
Results in App Analytics show impressions, conversion rate, percent improvement and confidence for each treatment against the baseline.
Apple vs Google, in practice
- Text: Apple tests only visuals. Google's localized experiments can also test descriptions. See Google Play experiments.
- Icons: on Apple, every alternate icon must be included in the app binary, so an icon test waits for a release.
- Changes: an Apple test can't be edited once started; applying a treatment mid-test ends it.
- Metric: Apple reports conversion rate; Google reports unique Install, Open or Pre-register clicks.
Apple's own examples
Apple's page cites Peak Brain Training: a 44-day icon test with three treatments, where one icon gained 8% conversion at over 98% confidence while two others did worse. It also cites Simply Piano, where the original page beat a new app preview video. Tests that "lose" are useful too: they stop you shipping a change that hurts.
A sensible test plan
- Pick one hypothesis per test: "a screenshot that leads with the result converts better than one that leads with the interface".
- Use fewer localizations if traffic is low; more languages means longer to reach confidence.
- Wait for 90% confidence, as Apple recommends, then apply.
- Feed winners into your custom product pages, which reuse the same assets.
Questions people ask
How many treatments can I test in App Store product page optimization?
Up to three treatments against your original product page, using alternate icons, screenshots and app previews.
How long can a product page optimization test run?
Up to 90 days, or until you stop it. You can run one test at a time and can't change it once it starts.
Do I need an app update to test a new App Store icon?
Yes. Alternate icons must be included in the app binary. Tests without alternate icons can be submitted independently of an app version.
Want this mapped onto your product?
See how our ASO work runs, or send your store link and a founder will reply personally.
