App Store & Google Play Reviews Scraper — New Reviews Only
Pricing
from $2.00 / 1,000 new review delivereds
Scrape iOS App Store and Google Play reviews and get only NEW reviews since your last run — stateful delta monitor with app-version regression flags. Both stores in one scheduled actor. No reviewer PII. Keyless.
One entry per app. Accepts Google Play package ids (com.whatsapp), Apple numeric ids (310633997 or id310633997), or full store URLs (play.google.com/store/apps/details?id=…, apps.apple.com/…/id…). Auto-detected: pure digits (after stripping an id prefix) → Apple; contains a dot → Play; URLs are parsed. Anything else (including entries over 512 chars) becomes an uncharged error row. Each app is checked in every countries code (app×country pairs). Cap: 30 pairs per run; overflow is reported in one aggregated error row.
ISO 2-letter lowercase country codes from the Apple storefront allowlist (also enforced for Play). Each app is polled in each country. Unknown codes are reported once in a single aggregated uncharged error row (never crash the run).
[ "us"]Names the persistent snapshot this schedule diffs against. Run one schedule per monitorId — a fail-closed KV lease blocks concurrent runs sharing the same id (double-billing guard); if the lease cannot be verified, the run skips uncharged. Empty → default. Truncated to 120 chars, then sanitized to [a-zA-Z0-9._-].
Newest-first cap of new reviews delivered per app×country pair on a delta run (and of ids seeded on baseline). Lossy: overflow beyond this cap is dropped permanently and never re-emitted. Baseline seeds up to this many ids into the snapshot.
When true, emit one uncharged version_summary row per app×country pair that delivered ≥1 new review. Groups by appVersion (Play) or unknown (Apple — no per-review version on the endpoint). Flags regressionSuspect when any version group has count ≥ 5 and avgRating ≤ overallAvg − 1.0.