Apple App Store Scraper

Scrape Apple App Store category charts, full app details, reviews, keyword search-rank, and similar apps — built for app-market research and idea validation: competitor density, positioning, review sentiment, and keyword landscape before committing engineering time to a new app.

Wraps the maintained app-store-scraper npm package (v0.18.0, ISC — same author, facundoolano, as google-play-scraper, which powers the sibling google-play-store-scraper actor) — public, unauthenticated iTunes Search API / App Store RSS endpoints, no login, no anti-bot wall.

What you get per run

Five dataset views, all in one dataset:

View	Rows	Use case
charts	One per listed app	Market sizing, competitor density, category trends
details	One per app	Deep competitor analysis, monetization, rating distribution
reviews	One per review	Sentiment, feature requests, crash reports, version-specific feedback
search	One per ranked result per term	Keyword-rank tracking — the ASO data AppTweak sells
similar	One per related app	Competitor/adjacent-app mapping ("customers also bought")

Modes (single actor, one input)

Mode	What it does
charts	Category/collection listings only
details	Full app details for explicit appIds
reviews	Reviews for explicit appIds
search	Keyword search-rank for searchTerms
similar	Related apps for explicit appIds
all (default)	Charts → auto-feeds discovered apps to details → reviews for everything

search and similar are opt-in — they take explicit keyword/appId lists rather than falling out of chart discovery, so they aren't part of all.

The all mode workflow

{
  "mode": "all",
  "categories": ["BUSINESS", "PRODUCTIVITY"],
  "collections": ["TOP_FREE_IOS", "TOP_GROSSING_IOS"],
  "countries": ["us"],
  "numPerChart": 50,
  "fetchDetailsForListedApps": true,
  "maxReviewsPerApp": 50
}

1. Scrapes 2 categories × 2 collections × 1 country × 50 apps = 200 chart rows
2. Deduplicates the 200 apps → unique bundle IDs
3. Fetches full details for every unique app
4. Fetches up to 50 reviews per app
5. Single run, single dataset, complete market picture

Input

Quick start (zero config)

Hit Run with defaults → gets top 50 free/paid/grossing apps in Business, Productivity, Health & Fitness (US), fetches full details + 50 reviews each.

Full input reference

{
  "mode": "all",
  "categories": ["BUSINESS", "PRODUCTIVITY", "HEALTH_AND_FITNESS"],
  "collections": ["TOP_FREE_IOS", "TOP_PAID_IOS", "TOP_GROSSING_IOS"],
  "countries": ["us"],
  "numPerChart": 50,
  "appIds": [],
  "fetchDetailsForListedApps": true,
  "fetchRatingsHistogram": true,
  "maxReviewsPerApp": 50,
  "searchTerms": [],
  "searchNumPerTerm": 20,
  "maxItems": 5000,
  "maxConcurrency": 3,
  "retryAttempts": 3,
  "retryDelayMs": 1500
}

Field	Type	Default	Description
mode	enum	"all"	charts | details | reviews | search | similar | all
categories	string[]	["BUSINESS","PRODUCTIVITY","HEALTH_AND_FITNESS"]	70 App Store genres. Use ["ALL"] for all 70.
collections	string[]	["TOP_FREE_IOS","TOP_PAID_IOS","TOP_GROSSING_IOS"]	iPhone/iPad/Mac × top/new charts (13 total)
countries	string[]	["us"]	2-letter codes (ISO 3166-1 alpha-2)
numPerChart	int	50	Apps per category/collection/country (max 200, App Store RSS limit)
appIds	string[]	[]	Bundle IDs, numeric App Store IDs, or full App Store URLs — for details/reviews/similar
fetchDetailsForListedApps	bool	true	In all mode, auto-fetch details for every app found in charts
fetchRatingsHistogram	bool	true	Fetch the 1-5 star breakdown per app (one extra request per app)
maxReviewsPerApp	int	50	0 = skip reviews. The App Store review feed hard-caps around 500 (10 pages) per app regardless.
searchTerms	string[]	[]	Keywords for search mode
searchNumPerTerm	int	20	Ranked results to capture per term
maxItems	int	5000	Hard cap on total dataset rows
maxConcurrency	int	3	Parallel requests. Apple rate-limits harder than Google Play.
retryAttempts	int	3	Retries for transient errors
retryDelayMs	int	1500	Base delay for exponential backoff

Output

Chart row

{
  "type": "chart",
  "category": "BUSINESS",
  "collection": "TOP_GROSSING_IOS",
  "country": "us",
  "rank": 1,
  "id": "288429040",
  "appId": "com.linkedin.LinkedIn",
  "title": "LinkedIn: Community & Network",
  "developer": "LinkedIn Corporation",
  "price": 0,
  "free": true,
  "currency": "USD",
  "genre": "Business",
  "genreId": "6000",
  "icon": "https://is1-ssl.mzstatic.com/...",
  "url": "https://apps.apple.com/us/app/linkedin-community-network/id288429040",
  "scrapedAt": "2026-09-13T22:14:49.555Z"
}

Detail row

{
  "type": "detail",
  "id": 294934058,
  "appId": "com.hotschedules.hotschedules",
  "title": "HotSchedules",
  "primaryGenre": "Business",
  "score": 4.60016,
  "reviews": 67269,
  "ratings": 67,
  "histogram": { "1": 3364, "2": 973, "3": 2315, "4": 5892, "5": 54725 },
  "price": 2.99,
  "free": false,
  "currency": "USD",
  "requiredOsVersion": "17.1",
  "released": "2008-10-31T05:30:52Z",
  "updated": "2026-09-12T06:49:02Z",
  "version": "6.0.1",
  "description": "...",
  "screenshots": ["https://...", "..."],
  "supportedDevices": ["iPhone16-iPhone16", "..."],
  "developer": "HotSchedules",
  "developerWebsite": "http://www.hotschedules.com",
  "url": "https://apps.apple.com/us/app/hotschedules/id294934058",
  "country": "us",
  "scrapedAt": "2026-09-13T22:14:55.796Z"
}

Note App Store addresses apps by both a numeric id (trackId) and a string appId (bundle ID) — unlike Play's single package-name scheme. Both are always present on detail/chart rows; appIds input accepts either form plus full URLs.

Key fields for idea validation:

• score + ratings/histogram — rating distribution, not just average (histogram is a separate request per app — see fetchRatingsHistogram)
• reviews / currentVersionReviews — engagement and version-specific reception
• screenshots / description / releaseNotes — positioning and update cadence
• contentRating, languages, supportedDevices — market reach signals
• developerWebsite — contact for partnerships

Review row

{
  "type": "review",
  "reviewId": "14540170480",
  "appId": "com.linkedin.LinkedIn",
  "userName": "TweetyRuns7",
  "version": "9.1.570",
  "score": 1,
  "title": "Can't login",
  "text": "I am not sure what's going on...",
  "url": "https://itunes.apple.com/us/review?id=288429040&type=Purple%20Software",
  "updated": "2026-09-12T01:34:22-07:00",
  "country": "us",
  "scrapedAt": "2026-09-13T22:20:52.940Z"
}

The App Store review RSS feed caps at ~500 reviews (10 pages) per app — this is Apple's limit, not ours.

Search row (keyword-rank)

{
  "type": "search",
  "term": "budgeting app",
  "rank": 1,
  "id": 1621020173,
  "appId": "com.fleurbudget",
  "title": "Fleur - Budget Planner App",
  "developer": "Akash Jain",
  "primaryGenre": "Finance",
  "score": 4.80138,
  "url": "https://apps.apple.com/us/app/fleur-budget-planner-app/id1621020173",
  "country": "us",
  "scrapedAt": "2026-09-13T22:30:07.974Z"
}

This is the keyword-rank data AppTweak/ASO tools sell — automated App Store search, capturing which apps rank where for each term.

Similar row

{
  "type": "similar",
  "sourceAppId": "310633997",
  "relatedId": 454638411,
  "relatedAppId": "com.facebook.Messenger",
  "relatedTitle": "Messenger",
  "relatedDeveloper": "Meta Platforms, Inc.",
  "relatedScore": 4.6613,
  "relatedPrimaryGenre": "Social Networking",
  "relatedUrl": "https://apps.apple.com/us/app/messenger/id454638411",
  "country": "us",
  "scrapedAt": "2026-09-13T22:30:25.503Z"
}

Apple's own "customers also bought" data per app listing — competitor/adjacent-app mapping.

What this doesn't do

This actor covers everything that's actually scrapable — public store metadata, charts, reviews, keyword search-rank, and similar-apps. Two things 42matters/AppTweak/Sensor Tower sell that this deliberately does not attempt:

• Install/revenue estimates. These come from proprietary SDK panels (analytics SDKs embedded in thousands of apps, giving device-level usage samples extrapolated statistically) and calibration deals with Google Play's disclosed install bands. Not scrapable — no public data source exists. If this is ever genuinely needed, buy API access from one of those vendors rather than trying to rebuild the model from rank/rating signals; that's a data-science project of its own with a poor accuracy ceiling.
• iOS SDK/tech-stack detection. App Store binaries are FairPlay-DRM-encrypted per purchase — getting a clean binary to inspect for embedded SDKs requires circumventing that DRM, a DMCA anti-circumvention issue, not a difficulty issue. (Android SDK detection via public APK teardown is technically feasible but was explicitly scoped out of this build — separate concern, ask if it's needed later.)

Reliability

• Concurrency control — p-limit semaphore, default 3 (lower than the Play actor's default 5 — Apple's iTunes endpoints rate-limit harder, returning 403 rather than 429)
• Retry with backoff — exponential backoff + jitter on 403/429/5xx and network errors (ETIMEDOUT/ECONNRESET/ECONNREFUSED/ENETUNREACH/ENOTFOUND/EAI_AGAIN)
• Graceful degradation — one failed appId/category/term doesn't kill the run; a missing ratings histogram doesn't lose the rest of a detail row
• Fail loud — a run that pushes 0 rows fails with reason — not a green run with an empty dataset
• Deduplication — apps deduplicated across chart discoveries and explicit appIds (by resolved bundle ID or numeric ID)
• Global maxItems cap — stops cleanly at limit

Categories & Collections (from the package, not hardcoded)

import appStore from 'app-store-scraper';
// category: 70 genres (BOOKS, BUSINESS, GAMES, ... WEATHER)
// collection: 13 charts (TOP_FREE_IOS, TOP_PAID_IPAD, TOP_GROSSING_MAC, NEW_IOS, ...)

ALL_CATEGORIES/ALL_COLLECTIONS in src/utils.ts read these directly from the library at runtime — use ["ALL"] in categories to scrape all 70.

Local development

cd apple-app-store-scraper
npm install
npm test                # runs unit tests (no network)
npm start                # runs locally, reads storage/key_value_stores/default/INPUT.json

Local storage directory is controlled by CRAWLEE_STORAGE_DIR (not APIFY_LOCAL_STORAGE_DIR), and its default-store input file must live at <dir>/key_value_stores/default/INPUT.json — running via the apify CLI (apify run) sets this up automatically; running node directly does not.

Test input (save as storage/key_value_stores/default/INPUT.json)

{
  "mode": "all",
  "categories": ["BUSINESS"],
  "collections": ["TOP_FREE_IOS"],
  "countries": ["us"],
  "numPerChart": 5,
  "fetchDetailsForListedApps": true,
  "fetchRatingsHistogram": false,
  "maxReviewsPerApp": 5,
  "maxItems": 100,
  "maxConcurrency": 3
}

Deployment

$apify push

Is scraping the App Store legal?

Scraping publicly available app listings, ratings, and reviews is generally legal. This actor collects only public, non-personal data — app metadata, aggregate ratings, and public reviews — via the same endpoints the App Store app itself uses. No user accounts, no private data, no DRM circumvention, no authentication bypass.

You are responsible for how you use the output, including under GDPR/CCPA where relevant. Apple's ToS prohibits automated access; use at your own risk. When in doubt, take legal advice.

Known dependency vulnerabilities

app-store-scraper depends on the deprecated request HTTP library, which transitively pulls in form-data/qs/tough-cookie/uuid versions with known CVEs (npm audit flags 2 critical, 7 moderate). All of them are exploitable only via attacker-controlled multipart form bodies or cookie jars — this actor only issues GET requests to fixed, hardcoded Apple endpoints and never sends multipart data or persists cookies, so none of these are reachable in practice. No non-breaking fix exists upstream (the fix downgrades to app-store-scraper@0.4.0, a materially older and less capable version); documenting the ceiling here rather than downgrading.

Changelog

• 1.0 — Initial release: charts, details, reviews, keyword search-rank, similar-apps, 70 categories, 13 collections, retry/backoff, fail-loud, deduplication, dataset views.

License

Apache-2.0

---

Pricing & cost estimator

Pay-per-event: $1.00 per 1,000 app rows. You are charged only for rows that contain data; empty or failed targets are free. A 100-item run typically finishes in well under a few minutes and costs a fraction of a cent in platform usage. Use maxItems to control spend.

Example output (real run)

{
  "type": "chart",
  "category": "BUSINESS",
  "collection": "TOP_FREE_IOS",
  "country": "us",
  "rank": 1,
  "id": "1113153706",
  "appId": "com.microsoft.skype.teams",
  "title": "Microsoft Teams",
  "developer": "Microsoft Corporation",
  "developerUrl": "https://apps.apple.com/us/developer/microsoft-corporation/id298856275?uo=2",
  "price": 0,
  "free": true,
  "currency": "USD",
  "genre": "Business",
  "genreId": "6000",
  "icon": "https://is1-ssl.mzstatic.com/image/thumb/Purple221/v4/5b/d9/40/5bd940d9-d71d-17b0-7515-8fff4d9e7aa4/AppIcon-0-…",
  "url": "https://apps.apple.com/us/app/microsoft-teams/id1113153706?uo=2",
  "scrapedAt": "2026-10-01T19:39:02.027Z"
}

Limitations

Chart positions and ratings reflect the moment of the run and the chosen country. Review texts are not collected unless maxReviewsPerApp is set.

No login, no cookies

This actor only reads public Apple App Store listings and charts. It never asks for a cookie, password or account and does not access private data. Use the data in line with the source site's terms and applicable law (GDPR/CCPA for personal data).

Need a custom version?

Need a field this actor doesn't return, a different site, a schedule, or delivery to Google Sheets / your CRM? We build and maintain custom scrapers, public data only. Open an Issue on this actor titled "Custom request", or email kashif@dialnorth.com.