App Store Review Scraper — with version tracking

Scrape App Store reviews and see which app version each review was written against. Get per-release rating breakdowns and automatic regression detection, so you can answer the question every mobile team actually asks: did the release we just shipped make things worse?

Why this one is different

Every App Store review scraper hands you a flat list of reviews. That tells you people are unhappy. It does not tell you when it started or what caused it.

Apple publishes the app version on every review. Almost nobody uses it. This actor joins reviews to releases and gives you:

• Rating per version — the full 1-to-5 distribution for every release
• Negative share — the percentage of 1 and 2 star reviews per version, which is the number that moves first
• Regression detection — an automatic comparison of the newest release against the one before it

A drop from 4.6 to 4.1 across a release, with 1-star reviews tripling, is a shipped bug. In a flat review list that is invisible.

What you get

Review rows — rating, headline, body, author, date, the app version it was written against, and whether that version is currently live.

Version rows — one per release: review count, average rating, full star distribution, negative share, and the date range those reviews cover.

Delta rows — newest release versus previous: rating change, negative-share change, sample sizes, and a regression flag.

The regression flag is deliberately conservative. It only fires when the rating dropped meaningfully and the share of 1-2 star reviews rose. A single angry reviewer will not trigger it.

Input

{
  "appIds": ["324684580"],
  "countries": ["us", "gb"],
  "maxPages": 3,
  "includeVersionSummary": true
}

App IDs — the digits after /id in an App Store URL. Paste the whole URL if easier; the ID gets extracted.

Countries — each country store holds its own reviews. A bug affecting one locale shows up in one store first.

Pages — 50 reviews each. Apple caps its public feed at 10 pages (500 reviews) per country, and the actor clamps to that rather than silently looping.

Rating filters — set maxRating: 2 to pull only complaints, which is the fastest way to find what broke.

Use cases

• Release monitoring — schedule daily and get alerted when a version regresses
• Competitive teardown — track a rival's ratings release by release
• Support triage — filter to 1-2 star reviews on the current version to see live problems
• Pre-launch research — read what users hate about the incumbent before you build

Reliability

One app or country failing never fails the run. If you track twenty apps and one is unavailable in one store, you get the other nineteen plus a record of what failed and why. You are only charged for rows actually returned.

Wrong app IDs produce a clear message telling you where to find the right one, rather than an empty dataset.

Data comes from Apple's public review feed — no login, no browser automation. Note that Apple's feed exposes recent reviews only; it is not a complete historical archive, and this actor does not pretend otherwise.

Notes

Unofficial and not affiliated with Apple. Review text remains the property of its authors — use it in line with applicable laws and terms.