Agoda Reviews Scraper — Hotel Guest Reviews & Ratings

An agoda reviews scraper that returns every guest review a property has, as structured rows: agoda ratings on both the 0–10 and 5-star scales, the rating word, title, review text, separate positives and negatives, stay dates and length, room type, reviewer name and country, and the hotel's public reply. It turns agoda hotel reviews into hotel review data you can query — a hotel review scraper built for bulk work rather than one page at a time.

Give it a hotel name, an Agoda URL, or a numeric hotel id — a name is resolved through Agoda's own search, so you never have to hunt for an id. No login, no agoda api key, no browser. This is a hotel review scraper that bills per review row, and two filters cut what you pay for before you are billed.

Last verified working: 2026-08-26.

What does the Agoda reviews scraper do?

You give it properties. It gives you their agoda hotel reviews as rows you can sort, filter and export.

• review rows — rating (0–10, Agoda's scale), rating_5 (the same score out of five), rating_text ("Exceptional"), title, text, positives, negatives, language, review_date, check_in, check_out, stay_month, stay_length, room_type, reviewer_name, reviewer_country, group_type, helpful_votes, unhelpful_votes, responder, response, response_date, images, source and provider_id.
• hotel rows — free context: the property id, name, URL and how many reviews were taken.
• status rows — free: per property, what was harvested, how many were filtered, and why anything failed.

Reviews come back in the language they were written in, with the language tagged. One Lisbon property returned 18 languages and 57 reviewer countries in a single run.

Agoda reviews by hotel, by name, or by id

Most tools make you find Agoda's internal id first. This one does not:

• hotelNames — type the property as you would into Agoda's search box: Park Hyatt Tokyo. Resolved through Agoda's own autocomplete to the exact property.
• startUrls — Agoda property pages. The numeric id is read from the page for you.
• hotelIds — the numeric id directly, e.g. 63820, which is what an earlier run's hotel_id gives you. This is the agoda reviews by hotel id path.

A name, its URL and its id all count as the same property and are harvested once, so mixing inputs never double-bills.

Agoda negative reviews without paying for the happy ones

Two filters run before billing, and together they are the reason a complaints feed is affordable:

• maxRating keeps only reviews at or below a score on Agoda's 0–10 scale. 5 is "a poor stay".
• requireText drops the score-only reviews that have no words — a large share of Agoda's corpus.

In a verified run, a property with 324 visible reviews delivered 8 rows under maxRating: 5 with requireText on. You pay for eight. That is what makes agoda negative reviews monitoring across a portfolio practical rather than theoretical.

Hotel review data for revenue, ops and analysts

• Hotel reputation monitoring. Schedule a daily run sorted most_recent and every new review arrives with its score and text. Nothing to poll.
• Agoda competitor reviews. Point it at the properties you compete with and pull their whole review corpus — what guests praise, what they complain about, which hotels bother to reply.
• Hotel sentiment analysis. positives and negatives arrive as separate fields, which is unusual and far easier to model than free text alone.
• Operations. Reviews carry check_in, check_out, stay_length and room_type, so a score ties back to the actual stay — you can find the room type that generates complaints.
• OTA review data benchmarking: source and provider_id say which channel a review came from, because Agoda aggregates partner reviews alongside its own.

Input

Field	What it does
hotelNames	Property names as typed into Agoda's search box, e.g. Park Hyatt Tokyo. Resolved through Agoda's own autocomplete.
startUrls	Agoda property pages. The numeric hotel id is read from the page for you.
hotelIds	Agoda's numeric ids, e.g. 63820 — the fastest input.
maxReviewsPerHotel	0 = every review the property has (Agoda pages ~200 at a time; large hotels run into the thousands). N = the first N. The main cost control.
reviewsSort	most_recent, highest_rated, lowest_rated, most_relevant.
maxRating	0 = keep everything. 1–10 = keep only reviews at or below that score. Filtered reviews are never charged.
requireText	Keep only reviews that actually have a written comment.
locale	Agoda locale for the request (en-us, es-es, de-de…).
includeHotelRow	Emit the free per-property context row.
sessions, perIp	Parallel proxy sessions and the pace of each. Leave alone unless a run is throttled.
proxyConfiguration	Apify Proxy. Required — Agoda throttles datacentre traffic that is not rotated.

Example: a scheduled complaints feed by hotel name

{ "hotelNames": ["Park Hyatt Tokyo"], "maxReviewsPerHotel": 0, "reviewsSort": "lowest_rated", "maxRating": 5, "requireText": true }

Example: the full corpus for a competitor set

{ "startUrls": ["https://www.agoda.com/the-central-house-lisbon-baixa/hotel/lisbon-pt.html"], "maxReviewsPerHotel": 0 }

Example: newest reviews only, by id, in Spanish

{ "hotelIds": ["63820"], "maxReviewsPerHotel": 200, "reviewsSort": "most_recent", "locale": "es-es" }

Output

{
  "type": "review", "hotel_id": "63820",
  "hotel_name": "Tivoli Avenida Liberdade", "url": null, "locale": "en-us",
  "review_id": "1059514024", "rating": 10.0, "rating_5": 5.0, "rating_text": "Exceptional",
  "title": "Nice Hotel", "text": "Beautiful hotel in the center!", "positives": null, "negatives": null,
  "original_title": null, "original_text": null, "language": "en", "translated_to": "en",
  "review_date": "2025-12-15T10:54:00+07:00", "review_date_text": "December 15, 2025",
  "check_in": "2025-12-08T00:00:00+07:00", "check_out": "2025-12-14T00:00:00+07:00",
  "stay_month": "December 2025", "stay_length": null, "room_type": null, "group_type": null,
  "reviewer_name": "VARDAN", "reviewer_country": "Mozambique", "reviewer_type": null,
  "helpful_votes": 0, "unhelpful_votes": 0, "images": null,
  "source": "Agoda", "provider_id": "332",
  "responder": "Tivoli Avenida Liberdade", "response": null, "response_date": null
}

The free hotel row gives you context:

{ "type": "hotel", "hotel_id": "63820", "hotel_name": "Tivoli Avenida Liberdade", "url": null, "reviews_fetched": 108, "locale": "en-us" }

status rows carry target, hotel_id, hotel_name, status, reviews, filtered, pages, duplicates, error and fetched_at, so every run says exactly what it did.

How much does it cost?

Pay per event, one event per delivered review row. Free: hotel rows, status rows, every review removed by maxRating or requireText, properties with no reviews, and failed properties. A property is harvested once per run whether you give its name, its URL, its id, or all three.

A mid-size hotel is a few hundred reviews; a busy hostel can be over a thousand — 1,457 from a single property was verified in testing. maxReviewsPerHotel and the two filters make that a deliberate choice. Verified runs behind this listing cost $0.0002–$0.0026 each in platform compute. Exact per-row pricing is on this page's pricing tab.

Agoda reviews scraper in Python, JavaScript, curl, n8n, Make or an AI agent

In Python:

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("kestrel/agoda-reviews-scraper").call(run_input={
    "hotelNames": ["Park Hyatt Tokyo"], "maxReviewsPerHotel": 0,
    "reviewsSort": "lowest_rated", "maxRating": 5, "requireText": True,
})
for row in client.dataset(run["defaultDatasetId"]).iterate_items():
    if row["type"] == "review":
        print(row["rating"], row["reviewer_country"], row["text"][:80])

In JavaScript (Node):

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: '<YOUR_APIFY_TOKEN>' });
const run = await client.actor('kestrel/agoda-reviews-scraper').call({
    hotelIds: ['63820'], maxReviewsPerHotel: 200, reviewsSort: 'most_recent',
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items.filter((i) => i.type === 'review').length, 'reviews');

With curl, synchronously:

curl -X POST "https://api.apify.com/v2/acts/kestrel~agoda-reviews-scraper/run-sync-get-dataset-items?token=<YOUR_APIFY_TOKEN>" \
  -H 'Content-Type: application/json' \
  -d '{"hotelIds":["63820"],"maxReviewsPerHotel":100}'

For n8n, Make and Zapier, add the actor as a step through Apify's integration and map rating, text and review_date into an alert or a sheet — "new review at or below 5/10" is two nodes. For AI agents, the actor is callable over Apify's MCP server. To get agoda reviews csv, run it and use the dataset's Export button, or append ?format=csv to the dataset items API.

Is it legal to scrape Agoda reviews?

This reads publicly visible property pages through the same JSON endpoint agoda.com's own review widget calls, with no login and no account. Reviews are published deliberately for anyone to read, and hiQ v. LinkedIn is the usual reference point for public data in the US.

Two caveats. Reviews contain personal data — a display name, a country and someone's words — so under GDPR you are a controller once you store them: have a lawful basis, keep them only as long as you need, and do not republish reviewer identities. This actor exposes no email, phone or full name, and has no input that would produce one. Separately, Agoda's Terms of Service discourage automated access, which makes this a terms question rather than a criminal one, with the risk on whoever operates the run. None of this is legal advice.

Limits and honest notes

• Agoda rates 0–10, not 1–5. rating is Agoda's own number; rating_5 is the same score halved so you can compare against five-star sources. Do not mix them.
• Many reviews are a score with no words. That is Agoda, not a parsing gap — requireText drops them if you only want written feedback.
• positives and negatives are only present when the guest used those fields. Older reviews often put everything in text.
• A throttled request answers 200 with no reviews at all, rather than an error. The run detects that and retries on a fresh IP, so a busy property is never reported as having none.
• stay_length, room_type and group_type are frequently null — Agoda only shows them for some reviews.
• HTTP/2 is required by Agoda's endpoint; over HTTP/1.1 it silently returns nothing. The actor handles this, but it is why a naive script fails.

FAQ

Does it need an Agoda API key or login?

No. There is no public agoda api for reviews to hold a key for. This calls the same endpoint agoda.com's own review widget uses, with no session and no account.

Can I download Agoda reviews as CSV or Excel?

Yes. Every run writes an Apify dataset, exportable to CSV, Excel, JSON or XML from the Console or over the API with ?format=csv. That is the quick path to download agoda reviews for one property or a whole competitor set.

How do I get only Agoda negative reviews?

Set maxRating to your ceiling — 5 on Agoda's 0–10 scale is a poor stay — and reviewsSort to lowest_rated. Add requireText: true so you only pay for reviews with actual words. Everything above the ceiling is dropped before billing.

Do I need the hotel id, or can I use the name?

Either. hotelNames resolves a plain name through Agoda's own autocomplete: Park Hyatt Tokyo became id 910 in testing, and Tivoli Avenida Liberdade became 63820. Use startUrls if you already have the page, or hotelIds if a previous run gave you the number.

How many reviews can one property return?

All of them. Agoda pages about 200 at a time and the run follows the pages, de-duplicating as it goes — 1,457 unique reviews from a single property was verified. maxReviewsPerHotel bounds it when you want a cheaper run.

Are reviews translated, and which languages come back?

Reviews arrive in the language they were written in, tagged in language, and translated_to records what Agoda showed them as. original_text holds the untranslated body when Agoda translated one. Multi language hotel reviews are the normal case here, not an edge case — one property returned 18 languages.

Does it include the hotel's replies?

Yes — responder, response and response_date when the property answered publicly. responder is also how the run recovers the hotel's name, because Agoda ships its property pages with an empty <title>.

Can I use it as a general hotel review aggregator?

For Agoda's corpus, yes — and note that Agoda itself aggregates some partner reviews, which is why each row carries source and provider_id. For a true hotel review aggregator across several OTAs you would run this alongside scrapers for the other sites and join on the property.

Can I scrape Agoda reviews without an API?

That is what this is — agoda reviews without api access, no key to apply for, no approval process. You call an Apify actor and read rows back.

What does bulk work cost?

Bulk hotel reviews are billed per delivered row, so cost scales with data, not time. The two filters and the per-hotel cap cut it directly, because filtered rows are never charged.

Review monitoring across a portfolio

Watching one property is a browser tab. Watching thirty is a data problem, and it is the job this was built for.

Schedule one run with every property you care about in hotelNames or hotelIds, sorted most_recent. Each run writes rows stamped with fetched_at, so the dataset becomes a time series without any extra work: reviews per property per week, average score by month, response rate over time. Review monitoring stops being a person checking pages and becomes a table you can chart.

Two settings make the scheduled version cheap. maxReviewsPerHotel bounds each property, and maxRating means a daily complaints run pays only for the reviews that need a human. A thirty-property portfolio checked daily for reviews at or below 5/10 is a handful of rows most days.

Guest feedback data that keeps its structure

The reason this corpus is more useful than a page of text comes down to three fields most tools flatten away.

• Separate positives and negatives. When a guest fills Agoda's two boxes, they arrive as positives and negatives rather than one blob. Sentiment work that would otherwise need a model is a column read. This is the single biggest difference between guest feedback data here and elsewhere.
• Stay context on every row. check_in, check_out, stay_length and room_type mean a complaint is attached to a date and a room, so you can find the week or the room type that generated it.
• The original language, preserved. language says what the guest wrote in, and original_text keeps the untranslated body. You can route by language before spending model tokens, or check a translation you do not trust.

What this does not do

It does not read private messages, it does not touch booking data, and it does not need or accept a login. Public reviews are the boundary, and there is no input that crosses it. It also does not return room prices — that is a different endpoint and a different product.

Choosing between sort orders

most_recent is the one to schedule: it answers "what changed". lowest_rated is the one to run once when you inherit a property and need to know its history of problems — combined with maxRating it is the cheapest possible audit. highest_rated is useful for marketing copy, and most_relevant mirrors what a shopper actually sees on the page, which matters if you are trying to understand why a property converts the way it does.

Related scrapers

Agoda reviews are one channel's opinion. These read the other channels and the prices behind them, with the same row shape and billing model:

• Google Hotels Prices Scraper — hotel prices and every booking site’s rate for a stay.
• Google Flights Scraper — airfare for a route and date, with a cheapest-day sweep.
• Airbnb Scraper — Airbnb prices, availability calendars and occupancy.
• Airbnb Reviews Scraper — every guest review of an Airbnb listing, with translations.

All of them bill per delivered row, never charge for rows a filter or a spending limit removed, and write an Apify dataset you can export to CSV, Excel or JSON.