App Store Review Pain Miner

Find product pain points and app ideas by mining low-star public Apple App Store reviews.

This Actor searches Apple's public App Store APIs for a keyword/category, fetches recent public reviews for matching apps, filters to low-star reviews, classifies complaints with transparent deterministic rules, and outputs both review-level rows and grouped opportunity summaries.

What you can use it for

• Indie app idea research
• Competitor complaint mining
• App Store niche validation
• Product backlog discovery
• Pricing/UX/support/reliability pain scans

Low-star reviews often reveal paid-pain patterns: subscription resentment, unreliable sync, confusing workflows, missing integrations, poor support, or privacy concerns.

Example input

{
  "query": "habit tracker",
  "country": "us",
  "maxApps": 5,
  "maxReviewsPerApp": 40,
  "maxReviewRating": 3,
  "includeUnknown": false
}

Input fields

Field	Type	Default	Description
query	string	required	App Store keyword/category to search, e.g. habit tracker, budget app, invoice maker.
country	string	us	Two-letter App Store country code used by Apple's public APIs. Invalid values fall back to us.
maxApps	integer	5	Number of search results to inspect. Capped at 25.
maxReviewsPerApp	integer	25	Number of recent reviews per app to inspect. Capped at 100.
maxReviewRating	integer	3	Only include reviews at or below this star rating.
includeUnknown	boolean	false	Include low-star reviews that do not match a known pain category.

Output

Dataset: review-level pain points

Each dataset row is one public App Store review with classification metadata:

{
  "reviewKey": "983826477:costs-too-much",
  "appId": 983826477,
  "appName": "Productive - Habit Tracker",
  "developer": "Mosaic S.r.l.",
  "appRating": 4.6,
  "appUrl": "https://apps.apple.com/us/app/productive-habit-tracker/id983826477",
  "reviewTitle": "Too expensive",
  "reviewText": "Costs too much for a simple habit tracker...",
  "reviewRating": 1,
  "painCategory": "pricing",
  "confidence": 0.67,
  "severity": 1.0,
  "opportunityScore": 0.84,
  "matchedTerms": ["expensive", "too much"],
  "painSummary": "Pricing pain: Costs too much for a simple habit tracker...",
  "opportunity": "Offer a clearer, cheaper, or one-time-purchase alternative with fewer ads and transparent limits."
}

Key-value output: grouped opportunities

The Actor also writes OPPORTUNITIES, a grouped list by app and pain category. This is usually the best starting point for research triage:

{
  "appName": "Strong Workout Tracker Gym Log",
  "painCategory": "reliability",
  "reviewCount": 7,
  "averageConfidence": 0.72,
  "averageSeverity": 0.87,
  "opportunityScore": 0.76,
  "matchedTerms": ["bug", "crashes", "sync"],
  "topPainSummaries": ["Reliability pain: ..."],
  "opportunity": "Differentiate on stability, fast bug fixes, dependable sync, and trustworthy data handling."
}

The SUMMARY key includes run metadata and the top 10 opportunity groups.

Pain categories

Category	Detects	Opportunity angle
pricing	subscriptions, expensive plans, paywalls, ads	clearer pricing, cheaper tiers, one-time purchase, fewer ads
reliability	crashes, bugs, broken sync, lost data	stable sync, data safety, fast bug fixes
ux	confusing UI, clunky flows, too many steps	simpler workflow and opinionated defaults
features	missing features, imports/exports/integrations	focused workflow or integration gap
support	customer support, refunds, no response	responsive support and visible changelogs
privacy	privacy/tracking/permissions/personal data	privacy-first positioning and local-data options
unknown	unmatched low-star reviews	manual review when includeUnknown is enabled

How to read the scores

• confidence: how specifically the review matched the category keywords.
• severity: rating-derived pain score; 1-star reviews score highest.
• opportunityScore: average of confidence and severity, useful for sorting.

Scores are triage signals, not proof of market demand. Always read the source reviewText before making product decisions.

Dogfood results from MVP validation

The Actor found useful opportunity clusters in real App Store scans, including:

• invoice maker: repeated pricing complaints around invoice limits and subscription cost.
• calendar: paid users complaining about ads and missing/unstable calendar features.
• receipt scanner: sync/data-loss/support complaints around tax/reimbursement workflows.
• workout tracker: repeated watch sync and crash complaints.
• family organizer: setup confusion and missing attachment/recurrence workflows.

These are promising enough for public beta, with the caveat that classification is deterministic and should be used for triage.

Notes and limitations

• Uses Apple's public iTunes Search and RSS customer-review endpoints.
• Does not log in to Apple, bypass access controls, or collect private data.
• Classification is deterministic keyword matching, not LLM analysis. This keeps runs cheap and explainable.
• Apple endpoint availability varies by country/app; some apps may have unavailable review feeds.
• Search results can include adjacent apps when a query is broad.

Local development

python -m pytest tests/test_pain_miner.py -q
apify validate-schema
apify run --purge

Deployment

$apify push