Review Pain Miner API

Turn product reviews into ranked customer pain themes, severity scores, trend signals, evidence, and recommended actions. Use inline JSON or enrich an existing Apify dataset.

✅ What you get

• Ranked complaint and product-pain themes
• Severity and evidence-volume confidence
• Earlier-vs-later trend direction when dates are available
• Up to five auditable review excerpts per theme
• Deduplicated counts, rating summary, and recommended next actions

❌ What this isn't

This is not a generic sentiment wrapper and it does not claim causality. Version 1 uses a transparent multilingual taxonomy so the same input produces the same result without LLM cost or API keys.

🔎 Use cases

• Find recurring product complaints before roadmap planning
• Detect pain that increased after a release
• Compare complaint themes across competitors
• Turn App Store, Trustpilot, Google, or marketplace review datasets into action-ready output

👥 Who it is for

Product managers, voice-of-customer teams, app agencies, reputation teams, and researchers who already have review data but need prioritized findings.

⚙️ Input

Pass reviews inline or supply an Apify datasetId. Common source fields are detected automatically. Use fieldMapping when your columns have custom names.

📥 Example input

{
  "productName": "Example App",
  "reviews": [
    { "id": "r1", "text": "The app keeps crashing during upload", "rating": 1, "date": "2026-07-01", "version": "4.2" },
    { "id": "r2", "text": "Upload crashed again after the update", "rating": 1, "date": "2026-07-05", "version": "4.2" }
  ],
  "minimumThemeMentions": 2
}

📤 Output

Each THEME row includes the theme, mention count/share, severity, trend, confidence, recommended action, and evidence objects. A final SUMMARY row and the OUTPUT key-value record contain run-level findings.

💵 Pricing

Pay-per-result pricing starts at $5 per 1,000 output rows. A typical analysis emits only a handful of theme rows plus one summary row. Apify platform usage is included.

🔁 Example tasks

• Mine one-star mobile app complaints
• Detect pain themes rising after a release
• Compare competitor review complaints

⚠️ Limitations

• Taxonomy detection can miss novel or highly domain-specific wording.
• Trend comparison requires at least six dated reviews.
• Evidence confidence measures volume, not causal certainty.
• Review content may contain personal data; only process data you are allowed to use.

🧩 Integration

Run from Apify Console, REST API, schedules, webhooks, or apify-client. Dataset output is JSON/CSV/Excel compatible.

❓ FAQ

Does it need an OpenAI key? No. The MVP is deterministic and has no third-party AI cost.

Can it analyze an existing Actor dataset? Yes. Pass its dataset ID.

Does it support custom schemas? Yes. Map custom columns with fieldMapping.

🛠️ Support

For reproducible issues, include a sanitized sample row, the field mapping, and the run ID. Do not include API keys or sensitive personal data.