Tripadvisor Reviews Scraper

Extract publicly visible Tripadvisor reviews from hotel, restaurant, attraction, and vacation-rental place URLs. Each successful row includes the place, rating, title, review text, reviewer, publication date, review URL, and transport path.

Why this actor

• Direct HTTP is attempted first to keep ordinary runs inexpensive.
• Optional Apify Proxy fallback can retry only a direct 403 or rate limit. Residential traffic is not silently used.
• Review pages and places are hard-bounded, and a deadline protects long runs.
• A review is pushed before Actor.charge({ eventName: "review" }); blocked, empty, invalid, sample, and diagnostic rows are never charged.
• Duplicate reviews are removed by review ID, URL, or a stable author/date/text key.

Input

placeUrls accepts Tripadvisor review URLs. maxPlaces is 1-20, maxReviewsPerPlace is 1-100, maxPages is 1-10, and deadlineSecs is 5-120. An empty input returns a labeled sample row and does no network work.

enableProxyFallback is off by default. When enabled, the actor tries direct traffic first and uses the supplied Apify Proxy configuration only after a direct block/rate limit. Set a residential group for Tripadvisor pages.

Output

Successful rows have ok: true, reviewId, placeName, placeUrl, rating, title, reviewText, reviewer, publishedAt, reviewUrl, and transport. Diagnostics have ok: false and an errorCode; they are never billed.

Pricing and margin

The proposed event is review at $0.00025 per successful review, below the closest current comparable at $0.0003 per review. After Apify's 20% platform share, creator proceeds are $0.0002 per review. This is economically defensible only if a 256 MB direct cloud run settles below $0.00012 per delivered review; it is not yet proven because Tripadvisor blocked the local direct smoke and no proxy-backed cloud run has been performed. Residential fallback is intentionally optional because proxy-heavy runs can erase the margin at this price.

Limitations

Tripadvisor can return a bot page or require a residential IP. The actor reports BLOCKED rather than billing empty output. It does not log in, bypass CAPTCHA, infer private data, or scrape beyond the configured bounds.

Example

{
  "placeUrls": ["https://www.tripadvisor.com/Hotel_Review-g60763-d671150-Reviews-The_Empire_Hotel-New_York_City_New_York.html"],
  "maxReviewsPerPlace": 20,
  "maxPages": 2,
  "enableProxyFallback": true,
  "proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"], "countryCode": "US" }
}