App Store Reviews Scraper

Export user reviews from the Apple App Store and Google Play as clean, normalized JSON — in a single run, for any list of apps. Every review comes back with the same flat shape (rating, title, text, author, date, app version, helpful votes and a sentiment hint), so results drop straight into a spreadsheet, database or BI tool.

Great for release monitoring (catch regressions after an update), competitor tracking, voice-of-customer research and support triage. Run it on a schedule to turn review streams into a living dataset.

• App Store reviews come from Apple's public RSS customer-reviews JSON feed — no API key, capped at ~500 reviews per app (10 pages), always most-recent.
• Google Play reviews are fetched with google-play-scraper and paginate beyond that.
• No LLM key, no proxy, no login required.

Input

Field	Type	Default	Description
appStoreAppIds	array of strings	[]	Numeric Apple iTunes app IDs, e.g. "310633997". Aliased as iosAppIds.
googlePlayAppIds	array of strings	[]	Google Play package names, e.g. "com.whatsapp". Aliased as androidAppIds.
country	string	"us"	Two-letter country / storefront code (aliased as countryCode).
lang	string	"en"	Two-letter language code for Google Play (aliased as language).
maxReviewsPerApp	integer	200	Max reviews per app (aliased as maxReviews). App Store RSS is limited to ~500.
sort	enum	"newest"	newest, rating, or helpful (helpful = Google Play only; App Store is always most-recent).

Provide at least one of appStoreAppIds or googlePlayAppIds. Original field names from compatible actors are accepted, so existing input configs keep working.

Example input

{
    "appStoreAppIds": ["310633997"],
    "googlePlayAppIds": ["com.whatsapp"],
    "country": "us",
    "lang": "en",
    "maxReviewsPerApp": 200,
    "sort": "newest"
}

Output

Each dataset item is one review (recordType: "review"):

{
    "recordType": "review",
    "store": "app_store",
    "appId": "310633997",
    "appName": "WhatsApp Messenger",
    "reviewId": "1234567890",
    "userName": "Jane Doe",
    "rating": 5,
    "title": "Works great",
    "text": "Love the new update!",
    "date": "2026-07-10T12:34:56.000Z",
    "thumbsUp": 3,
    "version": "2.26.1",
    "url": "https://apps.apple.com/us/review?id=...",
    "sentiment": "positive"
}

Field	Description
recordType	Always "review".
store	"app_store" or "google_play".
appId	The app id you supplied.
appName	Resolved human-readable app title.
reviewId	Stable review id (or null).
userName	Review author display name (or null).
rating	Star rating 1–5 (or null).
title	Review title (or null; App Store only).
text	Review body (or null).
date	ISO 8601 timestamp (or null).
thumbsUp	Helpful / vote count (0 when unavailable).
version	App version reviewed (or null).
url	Link to the review (or null).
sentiment	Heuristic: positive (rating ≥ 4), negative (≤ 2), else neutral.

Pricing (pay-per-event)

This actor uses the pay-per-event model with a single event, review-scraped, charged once per review delivered to the dataset. You only pay for reviews you actually receive — apps that return no reviews or fail to load cost nothing.

Notes & limitations

• Apple's RSS feed returns at most ~500 reviews per app (10 pages) and always sorts by most recent; the sort option only affects Google Play.
• thumbsUp for App Store reviews reflects the RSS vote count and may be 0 when unavailable.
• A single app failing (network or an invalid id) is logged and the run continues with the next app.

License

MIT © Assia Fadli