Tripadvisor Reviews Scraper

Collect Tripadvisor reviews for attractions, hotels and restaurants as clean, flat rows: rating, title, full text, review date, trip type, author, sub-ratings (rooms, service, food, ...), photos and the owner's reply. Point it at place URLs or IDs, or just type place names. It is built for reputation monitoring, competitor benchmarking, sentiment and NLP datasets, and travel research.

The actor reads Tripadvisor's mobile-app data feed, so it needs no login and no browser. Every place is read page by page (20 reviews per page) until your limit or the last review.

Output fields

Field	Description
review_id, review_url	Tripadvisor review ID and its canonical link
place_id, place_type, place_name, place_url	The attraction, hotel or restaurant the review belongs to
rating	1 to 5
title, text	Review headline and full text (plain text, line breaks kept)
published_text, published_date, published_date_is_approximate	Date as Tripadvisor shows it ("3 days ago", "Dec 3, 2022") and an ISO date. Recent reviews only carry relative text, so their ISO date is approximate (true)
visit_date_text, visit_month	When the reviewer travelled ("Sept 2026" and 2026-09)
trip_type	Business, Couples, Family, Friends or Solo
author_name, author_location, author_contributions, author_profile_url	Reviewer details as shown on the review
helpful_votes	Number of "Helpful" votes
sub_ratings	Hotels and restaurants: for example {"Rooms": 5, "Service": 4}
owner_response	Management reply: text, author, author_position, published_text, published_date, profile_url
tip, label	Reviewer tip and badges such as "HOTEL'S FAVORITE", when present
is_translated	true when the text was machine-translated to English by Tripadvisor
photo_count, photos	Reviewer photo links
scraped_at	UTC timestamp

Use cases

• Reputation monitoring: schedule a daily run with reviewsSince set to yesterday and send new reviews and owner replies to Slack, a sheet or a webhook.
• Competitor benchmarking: compare ratings, complaints and response rates across hotels or restaurants in one city.
• Sentiment and NLP datasets: large review corpora with star labels, trip type and language filters.
• Hospitality and tourism research: seasonality (publishedMonths), trip-type mix and sub-rating trends.

Input

Field	Description
placeUrls	Tripadvisor Attraction_Review, Hotel_Review or Restaurant_Review URLs from any Tripadvisor domain, or IDs such as hotel:188730, restaurant:719141, attraction:188151
searchQueries	Place names to look up, e.g. Hotel Plaza Athenee Paris. placeTypes and maxPlacesPerQuery control what is scraped
maxReviewsPerPlace	Reviews to collect per place. 0 = all matching reviews
language	all (default) or one language. With all or en, non-English reviews arrive translated to English (is_translated); a specific language returns the original text
ratings, tripTypes, reviewsSince, publishedMonths	Tripadvisor's own review filters (stars, trip type, written on or after a date, written in these calendar months)
concurrency	Places scraped in parallel (reviews inside one place are read in order)
proxyConfiguration	Datacenter proxy by default. After repeated blocks the run switches to Apify RESIDENTIAL automatically

{
    "placeUrls": [
        "https://www.tripadvisor.com/Hotel_Review-g187147-d188730-Reviews-Hotel_Plaza_Athenee-Paris_Ile_de_France.html"
    ],
    "searchQueries": ["Le Jules Verne Paris"],
    "placeTypes": ["restaurant"],
    "maxReviewsPerPlace": 200,
    "language": "all",
    "ratings": ["1", "2"],
    "reviewsSince": "2026-01-01"
}

Scheduling and webhooks

Run it on a schedule with reviewsSince set to your last run date, then use an Apify webhook or integration to push the new rows to your tools. Rows are pushed as they are collected, and an interrupted run resumes from where it stopped.

Using it from AI agents and pipelines

The dataset is flat JSON/CSV, so it works with the Apify API, the Python and JavaScript clients, and any MCP-compatible agent that can call Apify Actors.

FAQ

Do I need a Tripadvisor account? No.

Why is the date sometimes approximate? Tripadvisor's app shows "3 days ago" for recent reviews and a full date for older ones. The ISO date for relative text is calculated from the scrape time, so it can be a day off (for weeks and months it is rougher). published_date_is_approximate tells you which.

Can I get every review of a very popular place? Yes, set maxReviewsPerPlace to 0. Reviews are read in order at roughly 10 reviews per second per place, so a place with 100,000+ reviews takes hours. Split it with the star-rating or date filters and run the slices in parallel as separate runs.

Are all languages covered? Yes. Choose all or a specific language.

What happens when Tripadvisor blocks an IP? The request is retried on a fresh IP, and after repeated blocks the run switches to RESIDENTIAL proxies.