If you want app reviews as a dataset, the two stores give you different things. Google Play organises reviews by language and shows them on one global listing page. The App Store organises reviews by country storefront and does not record the language at all. Both show developer replies publicly. Both hide star-only ratings that have no written text. Both show far more to the developer of the app than to anyone else.
The short version: Google Play is better for studying how an app is received across languages, the App Store is better for studying how an app is received across countries, and neither store gives an outsider the complete set of reviews. If the app is your own, the free official tools from Google and Apple are the right choice. If the app belongs to someone else, you are limited to what the public store page exposes, and this article explains exactly what that is.
What a logged-out visitor sees on Google Play
A Google Play listing is a single page per app, identified by its package name. The reviews shown on that page depend on two URL parameters: hl for language and gl for country. Change those and you get a different slice of the same review pool. Google filters reviews by the language the review was written in, as detected by Google, not by the country of the reviewer. That is why a Dutch visitor sees Dutch reviews first and an American visitor sees English ones.
For each review the public page shows a display name, a star rating, the review text, the date, a count of people who marked it helpful, and the developer reply with its own date if one exists. It does not show the device model, the Android version, the app version at the time of writing, or the reviewer's country. The developer sees all of those in the Play Console, and that is a real gap. From the outside you can tell a review is in German. You cannot tell whether it came from Germany, Austria or Switzerland.
Google documents the reply feature itself on its help page for replying to reviews. Two mechanics from that page matter for data work. A developer can edit a reply, and the public page shows only the current version. A user can update their review after a reply, and the store shows the updated review with the updated date. So a timeline reconstructed from the public page is a timeline of latest states, not of events.
Star-only ratings are the other blind spot. Google Play lets a user tap a star without writing anything. Those ratings count toward the average shown on the listing but never appear as reviews. On apps with a large install base the written reviews can be a small fraction of all ratings, and nothing on the public page tells you how small.
What a logged-out visitor sees on the App Store
The App Store works by storefront. Every country has its own storefront, and a review posted from a Dutch account lives in the Netherlands storefront only. The web version at apps.apple.com reflects this with the country code in the URL. Switch from /nl/ to /de/ and you get a different review set, a different rating average, and often a different rating count.
Each public review shows a nickname, a title, the body, a star rating, the date, and the developer response if there is one. Apple also lets the developer choose whether the displayed rating summary covers all versions or only the current version, and lets them reset the rating when a new version ships. Apple describes these controls in its overview of ratings and reviews. From the outside you cannot see whether a reset happened. A low review count with a recent-looking rating average is a hint, not proof.
The web page itself shows only a handful of reviews per storefront. There is no "see all" view that pages through everything. The nearest public route to a bigger set is Apple's long-standing RSS feed for customer reviews, which returns recent reviews per storefront in pages, sorted by most recent or most helpful. It is capped at a fixed number of recent reviews per country, so for popular apps it is a sample of the newest reviews, not an archive. Apple has never documented this feed as an official API and has changed its behaviour before, so treat it as convenient rather than guaranteed.
The App Store does not record or expose the language of a review. A review in the United States storefront can be in Spanish. A review in the Switzerland storefront can be in French, German, Italian or English. If language matters to your analysis, you have to detect it yourself from the text.
Developer replies: similar on the surface, different underneath
Both stores show one developer reply per review and let the developer edit it. Both show the reply date. That is where the similarity ends.
On Google Play, a reply triggers a notification to the reviewer, and the Play Console tracks whether the user changed their rating afterwards. That before-and-after number is only visible to the developer. On the public page you can see the reply exists and what it says, but you cannot see whether the review was a two-star review before the reply and a four-star review after it. The public date on the review is the date of the last edit, so a review that was updated after a reply looks like it was written after the reply.
On the App Store, a developer response also notifies the user, and the user can edit their review in response. Apple's public page shows the current review and the current response. Apple adds one detail Google does not: the response is shown with the developer name as a label, so it is unambiguous that the reply is from the publisher and not from another user.
For an analyst the practical difference is this. Google Play's public data lets you measure reply rate by language, since you can pull the same app in many hl values. The App Store's public data lets you measure reply rate by country, since you can pull many storefronts. Neither lets you measure the effect of a reply from the outside.
The free official routes, and when to use them
If the app is yours, do not scrape your own store page. Both platforms give you more and better data for free.
Google offers two routes. The Play Console lets you download monthly review exports as CSV files from a Google Cloud Storage bucket linked to your account. Those files include the device, the app version code, the reviewer language, and the full reply history. Google also offers the Reviews resource in the Google Play Developer API, documented at developers.google.com/android-publisher/reply-to-reviews. That page states the key limitation plainly: the API only returns reviews that were created or modified within the last week, and only reviews that include a comment. For anything older you need the monthly exports.
Apple offers the App Store Connect API, with endpoints for customer reviews and customer review responses, documented at developer.apple.com/documentation/appstoreconnectapi. It returns reviews for apps in your own account, with a territory field per review, and lets you post and edit responses. App Store Connect also shows ratings and reviews per territory in the web interface, with filters for rating and version.
These routes are the better choice whenever you own the app. They are complete, they are free, and they include the fields the public page hides. Their one shared limit is scope: both only cover apps in your own developer account. For competitor research, market sizing or sales prospecting, they give you nothing.
Where the public data is incomplete, in one list
- Star-only ratings are invisible on both stores. The written reviews you can collect are a self-selected subset.
- Google Play exposes language but not country. The App Store exposes country but not language.
- Neither store exposes the device or OS version to a logged-out visitor.
- Both stores show the latest state of a review and a reply, not the edit history.
- The App Store web page shows only a few reviews per storefront, and the RSS feed caps at recent reviews.
- Google Play shows reviews filtered to a language, so a complete pull means iterating over every language the app has reviews in.
- Both stores remove reviews that violate their policies, and removed reviews disappear from the public page without any trace.
- Rating resets on the App Store are invisible from the outside.
None of this makes public review data useless. It makes it a sample with known biases. If you compare apps, compare them under the same conditions: same language slice on Google Play, same storefronts on the App Store, same date window on both. If you compare stores, remember you are comparing a language slice with a country slice, and say so in whatever you produce.
Picking the store for the question
Ask which dimension your question is about. "How do Spanish speakers feel about this app" is a Google Play question, because Google filters by language and Spanish speakers in Mexico, Spain and the United States all land in one slice. "How does this app perform in Japan" is an App Store question, because the Japan storefront is a clean boundary and Apple's rating summary is per storefront. "Does this company reply to complaints" works on either store but means something slightly different on each, since the Play Console gives developers a stronger nudge to reply and the reply rate on Google Play is often not comparable with the one on the App Store for the same app.
For apps you publish yourself, start with App Store Connect and the Play Console exports. For apps you do not own, where you need the public Google Play reviews for many apps, in several languages, with developer replies included and paid per result rather than per month, the Google Play reviews scraper on this site collects exactly the fields described above and nothing it cannot see.
Want the signal instead of the raw filings? Get a free report preview. Prefer the tool to the write-up? Browse all data feeds or connect the free MCP server.