Review Intelligence Analyzer

Turn raw customer reviews into an action-ready product and marketing report—without reading thousands of comments manually.

The Actor accepts review objects pasted directly into the input, loaded from any authorized Apify dataset, or collected in one click from Google Maps and YouTube through established source Actors. It automatically detects common fields from Google Maps, Amazon, app-store, social-comment, and generic review scraper outputs.

One-click sources

Choose Google Maps reviews or YouTube comments, add URLs, and run. The Actor calls the selected source Actor under the current user's Apify account, waits for its dataset, and analyzes the returned reviews or comments automatically.

• Google Maps uses compass/google-maps-reviews-scraper.
• YouTube uses streamers/youtube-comments-scraper.
• Source Actor pricing is separate and is shown by Apify before the run.
• Personal reviewer data is disabled for Google Maps source calls because the analysis only needs review text and ratings.

What you get

• sentiment totals and per-review sentiment
• recurring themes with mention share and net sentiment
• prioritized pain points with evidence quotes and source URLs
• feature-request clusters
• tracked competitor mentions
• positive themes that can become marketing angles
• ranked product or operational actions
• JSON dataset plus Markdown and standalone HTML reports

Basic analysis is deterministic, multilingual, fast, and needs no external AI key. English and Simplified Chinese are supported in the first release.

Input options

Use either or both:

1. reviews: an array of review objects.
2. datasetId: the ID of an existing Apify dataset.

Common text fields such as text, reviewText, content, comment, and body are detected automatically. Use textField, ratingField, or urlField when your input uses custom nested fields.

Example

{
  "reviews": [
    {
      "text": "Fast delivery and excellent quality. I would buy again.",
      "rating": 5,
      "url": "https://example.com/review/1"
    },
    {
      "text": "The app crashes often. Please add offline mode.",
      "rating": 2,
      "url": "https://example.com/review/2"
    }
  ],
  "competitorNames": ["Competitor A"],
  "outputLanguage": "en"
}

Outputs

• Default dataset: one summary row followed by normalized, analyzed review rows.
• OUTPUT: complete report as JSON.
• REPORT.md: portable Markdown report.
• REPORT.html: browser-ready executive report.

Best workflow

1. Choose a one-click source, paste URLs, and run; or provide an existing dataset/direct review list.
2. Add competitor names if useful.
3. Run the Actor and download the report.

Privacy

The Actor processes only the input supplied to the run. It does not send review content to an external AI provider and does not require third-party credentials.

Limits

The first version uses explainable keyword and rating signals. Sarcasm, implicit context, and highly specialized terminology may need manual review. Evidence quotes and source links are included so every important conclusion can be checked quickly.