Review Intelligence Scraper — Product Review Aggregation & Sentiment

Collect product reviews and review signals from across the web in one run. Discover where a product is reviewed, extract ratings, review text, authors, and dates, and get automatic sentiment analysis, theme extraction, and aggregate stats for competitive intelligence and product research.

How It Works

The actor queries each selected source for your product, parses ratings and review content, and normalizes everything into a single schema with built-in sentiment scoring (positive / neutral / negative), top discussion themes, and average-rating aggregates. Every run finishes with a structured OUTPUT record including per-source status, so you always know exactly what was reached.

Source Reliability Tiers

Review platforms protect themselves aggressively, so sources are organized by how reliably they respond to cloud IPs:

• Reliable: duckduckgo (review-page discovery with excerpts), google (Google AI Overview rating summary — organic links are frequently withheld from cloud IPs)
• Best-effort: trustpilot, g2, capterra, trustradius, appstore (attempted on every run; these typically require a residential or private proxy — attach one via proxyConfiguration — otherwise they are reported as blocked/unreachable instead of failing the run)

Features

• Multi-source reviews: App Store, Google discovery, Trustpilot, G2, Capterra, TrustRadius
• Sentiment analysis: Heuristic positive/neutral/negative scoring per review
• Theme extraction: Top recurring discussion topics across reviews
• Aggregates: Average rating, sentiment split, per-source status
• Verified badges: Captured where the source exposes them
• Proxy support: Apify Proxy included; residential recommended for G2/Trustpilot depth
• Never fails on blocks: Time-budgeted crawl, graceful per-source fallback

Input Parameters

Parameter	Type	Default	Description
productName	string	Salesforce	Product to analyze (required)
sources	array	["duckduckgo", "google", "g2"]	Review sources to query
maxReviews	integer	100	Max reviews to collect (1-1000)
includeSentiment	boolean	true	Score sentiment + extract themes
outputFormat	string	json	Export format
proxyConfiguration	object	Apify Proxy	Proxy for protected sources

Output Example

{
  "product": "Slack",
  "source": "duckduckgo",
  "author": "pcmag.com",
  "rating": 0,
  "ratingScale": 5,
  "title": "Slack Review: The Top Team Communication App",
  "text": "Slack receives high praise for its intuitive design and massive integration library...",
  "date": "",
  "verified": false,
  "sentiment": "positive",
  "sentimentScore": 0.03,
  "reviewUrl": "https://www.pcmag.com/reviews/slack",
  "scrapedAt": "2026-09-06T05:32:32.000Z"
}

The OUTPUT record also includes average_rating, sentiment_split, top_themes, and source_status (e.g. {"duckduckgo": "ok (8 reviews)", "google": "unreachable"}).

Use Cases

• Competitive intelligence: Where are competitors reviewed, and how are they rated?
• Product research: Aggregate sentiment and recurring complaint themes
• Review coverage audit: Which platforms list you (or don't)?
• Market mapping: Discover review sources per product category