Google Play has native A/B testing built into the Play Console. Free, on by default once you have a published app, and surprisingly capable. It is the closest thing to Apple's PPO on the Android side.
Two kinds of experiment
- Default graphics experiments. Tests global graphics — feature graphic, icon, and screenshots — across all languages.
- Localized experiments. Tests per language. Lets you isolate graphic or text changes for a specific locale.
What you can test
Graphics, descriptions, or both. That includes the short and long descriptions, which is one place Play beats App Store PPO — Apple forces metadata changes through review, Google does not.
Concurrent limits
You can run up to five localized experiments at once. Default graphics experiments are limited to one at a time. An ambitious month is one global hero test plus five language-specific copy tests.
Custom store listings on Play
The CPP equivalent on Play is Custom Store Listings (CSLs). Up to 50 per app. Each CSL can be triggered by country, install state, keyword, Google Ads campaign, or a unique URL. Experiments can run on top of CSLs, so you can A/B-test a campaign-specific listing without touching the default.
How to read results
Play reports an "applied installs" metric and a confidence range. Wait until the lower bound of the winning variant is above zero before you call it. If your traffic is low, lean on the longer- running tests and resist the urge to stop early.
Bottom line
Compared to App Store PPO: more flexible (text counts), broader (50 listings vs 35 CPPs), faster turnarounds (no review for most changes). Less flexible on icon variants, but Android does not push icon changes the same way iOS does.