Most App Store A/B testing advice assumes one market. But if your app is localized, your product page is really several product pages, one per language, and each one can win or lose separately. A headline that works in English can fall flat in Japanese, and a busy screenshot that Korean users like may look cluttered in Germany.
This guide covers how Product Page Optimization handles localizations, and a simple way to test market by market without multiplying your workload.
How Product Page Optimization Handles Languages #
When you create a test in App Store Connect, you choose which localizations it includes: all of them, or only some. Each treatment then needs screenshots for every localization in the test.
Two things follow from that:
- Traffic only counts from the included localizations. Include one small market and the test runs on that market's traffic alone.
- Every localization multiplies the work. Three variants, five languages and two device sizes is 30 screenshot sets. Apple also notes that tests with more localizations can take longer to show results.
Pool Markets or Test Them Separately? #
Pool them (one test, several localizations) when:
- The variant changes something visual that doesn't depend on language: layout, color, device frame, order.
- The markets are similar, for example English, German and Dutch.
- Each market alone doesn't have enough traffic to finish a test.
Test separately when:
- The variant changes wording. Captions don't translate one to one.
- The markets are culturally far apart, for example the US and Japan.
- One market is big enough to finish a test on its own.
Why it matters: if a variant wins by 15% in one market and loses by 15% in another, a pooled test reports roughly nothing, and you throw away a winner. App Store Connect lets you run only one test at a time, so separate tests have to run one after another. Choose the order carefully.
Not sure a market has enough traffic? Put its daily impressions into the A/B test calculator.
A Market-by-Market Plan #
1. Start in your biggest market #
It finishes fastest, and it's where a win is worth the most. Test a bold idea there first. Our 25 screenshot A/B test ideas is a good place to find one.
2. Treat the winner as a hypothesis elsewhere #
A winning visual change (a new first screenshot, a different layout) is usually safe to roll out to similar markets. A winning caption has to be translated, and a translated caption is a new caption. Test it again in your next-biggest market before you apply it everywhere.
3. Localize more than the words #
Translated text over English example content still looks foreign. Try a variant for one market that shows local names, local currency and local places in the UI. That is often a bigger win than any headline change, and it is a very clean test: same design, only the content differs.
4. Keep a results log per market #
Write down what you tested, where, and what happened. After a few tests you'll know which markets behave alike, and you can pool those in the future to finish tests faster.
Don't Let Languages Slow Down the Test #
The reason most developers test only in English is that making a variant in five languages takes five times as long. In Screenshot Studio, every language lives in one project. A design change applies to all of them, captions can be translated for you, and each language keeps its own screenshots where it needs them. Then upload the variant straight into the test: Screenshot Studio adds any localization the variant is missing and fills every language and device size in one step.
For more on localization in general, see how to create localized App Store screenshots.