Agoda Reviews Scraper — Export Hotel Guest Reviews to JSON, CSV or Excel

Extract hundreds of reviews per hotel in minutes — including the ones Agoda's own page never shows you. Paste a hotel URL and get clean, structured review data: ratings, guest text in both the original language and yours, traveller types, stay dates, room types, photos and owner replies.

• 📈 5–6× more reviews than the hotel page displays — 669 to 763 across test runs, against 121 without the sweep
• 🌍 Four review sources in one dataset — Agoda, Booking.com, Priceline and partner sites, merged and de-duplicated
• 🗣️ Original + translated text on every review, so nothing is lost in translation
• ⚡ No browser, no CAPTCHA solving — a lightweight HTTP scraper, so runs are fast and cheap
• ✅ Honest coverage reporting — every record tells you how many reviews it collected and flags partial runs

---

📌 What This Actor Does

This Apify Actor turns any Agoda hotel page into a structured dataset of guest reviews. You give it one or more Agoda hotel URLs; it returns one record per hotel containing the property's rating summary and every review it could reach, ready to download as JSON, CSV, Excel or XML, or pull straight from the Apify API.

Agoda's website only ever shows a slice of a property's reviews, and it hides them behind pagination, language filters and traveller-type filters. This Actor works through those filters systematically and merges the results, so you get a far more complete picture than manual browsing or a naive scraper produces.

Who it's for

• Hotel and hospitality teams monitoring guest sentiment and comparing against competitors
• Revenue and reputation managers tracking how scores move over time
• Market researchers and analysts studying travel demand, guest expectations and regional differences
• Data scientists and AI teams building training sets or running sentiment and topic analysis
• Agencies and consultants producing reputation audits and competitive benchmarking reports for hotel clients
• Developers who need Agoda review data through a clean API instead of maintaining a scraper

Real-world uses

• Pull every reachable review for your hotel and your five closest competitors, then compare per-aspect scores
• Feed review text into an LLM to summarise recurring complaints about breakfast, noise or cleanliness
• Track a property's rating trend month over month by re-running on a schedule
• Build a multilingual review corpus using the untranslated originalComment field
• Identify which traveller segments rate a property worst and why

---

✨ Key Features

• Deep review extraction. Sweeps Agoda's per-language filter (34 languages on the test property) and its six traveller-type filters, de-duplicating as it goes.
• Multi-source data. Agoda merges reviews from Booking.com, Priceline and partner sites. Each review is labelled with its providerName so you always know the source.
• Both languages, always. reviewComments gives you the translation in your chosen language; originalComment and originalLanguage preserve what the guest actually wrote.
• Rich property summary. Every record includes the aggregate score, per-aspect grades (location, cleanliness, service, room comfort, value, facilities, food), rating distribution, per-traveller-type scores and topic sentiment tags.
• 40 fields per review — rating, title, positives, negatives, merged text, reviewer name and country, traveller type, room type, length of stay, check-in/out dates, helpful votes, photo URLs and the hotel's own reply.
• Photo URLs included, fully qualified and ready to download.
• Transparent coverage. reviewsCollected, reviewsAvailable and a partial flag on every record, so you never mistake a throttled run for a complete one.
• Choose your language. 15 output languages, from English and Spanish to Japanese, Korean, Thai and Vietnamese.
• Fast mode for a cheap sample (~121 reviews in a single request) when you don't need everything.
• Incremental saving. Each hotel is written to the dataset the moment it finishes, so an interrupted run keeps what it already collected.

---

🧠 Why This Actor Is Different

It doesn't stop where Agoda's page stops. A straightforward scraper paginates the review list and hits Agoda's ceiling at around 121 reviews per property — no matter how many pages it requests or how it sorts. This Actor uses Agoda's own language and traveller-type filters as a lever. Measured, on hotel 49487:

Approach	Reviews returned
One plain API request (fastMode)	121
Re-sorting the list, all four sort orders	121 — no gain
Language sweep	611
Language + traveller-type sweep (default)	669 – 763

The range is real: Agoda intermittently answers a request with an empty list, so no two runs return exactly the same total. The Actor retries through that, and tells you via partial when it couldn't.

It keeps the original text. Most tools give you Agoda's machine translation and throw away what the guest actually typed. That is the difference between usable multilingual analysis and a lossy English-only corpus.

It's honest about what it got. Agoda signals rate limiting with an HTTP 200 and an empty review list — indistinguishable from "you've reached the end" unless you look closely. This Actor detects that difference, retries, and marks the record partial: true if anything was still missed. Silent truncation reported as success is the most common failure mode in review scrapers, and it's the one that quietly corrupts your analysis.

It's cheap to run. No headless browser, no CAPTCHA solver — just HTTP requests. That means lower compute usage and faster runs than browser-based alternatives.

It's structured for analysis. One record per hotel with reviews nested inside, so hotel-level and review-level data never get tangled or duplicated across thousands of rows.

---

⚙️ Input Configuration

Minimal example

{
  "startUrls": [
    { "url": "https://www.agoda.com/ibis-budapest-city/hotel/budapest-hu.html" }
  ]
}

Full example

{
  "startUrls": [
    { "url": "https://www.agoda.com/ibis-budapest-city/hotel/budapest-hu.html" },
    { "url": "https://www.agoda.com/park-hyatt-tokyo/hotel/tokyo-jp.html" }
  ],
  "maxReviews": 500,
  "fastMode": false,
  "language": "en-us",
  "includeDemographicSweep": true,
  "requestDelayMs": 500,
  "proxyConfiguration": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["RESIDENTIAL"]
  }
}

Field reference

Field	Type	Default	What it does
startUrls	array	required	Agoda hotel URLs. The /reviews/ version of a URL works too.
maxReviews	integer	0	Stop after this many reviews per hotel. 0 means collect everything reachable.
fastMode	boolean	false	One request per hotel instead of a full sweep. ~121 reviews rather than ~763.
language	string	en-us	Language Agoda translates reviews into. The original text is kept regardless.
includeDemographicSweep	boolean	true	Adds six traveller-type passes. Worth ~150 extra reviews; ignored in fast mode.
requestDelayMs	integer	500	Pause between requests. Raise it if you see throttling.
proxyConfiguration	object	Residential	Proxy settings. See Advanced Tips.

Supported language values: en-us, de-de, fr-fr, es-es, it-it, pt-pt, nl-nl, ru-ru, ja-jp, ko-kr, zh-cn, zh-tw, th-th, id-id, vi-vn.

Tips for best results

• Keep the residential proxy on, especially for multiple hotels or scheduled runs — Agoda rate-limits per IP and a full sweep is 85–115 requests per hotel.
• Use fastMode to test. Validate your URLs and output shape cheaply before committing to a full run.
• Set maxReviews if you only need a sample. It stops the sweep early and cuts your run time proportionally.
• Add hotels in batches, not hundreds at a time. Each property takes several minutes for a complete sweep.
• Copy URLs straight from your browser. Any Agoda hotel page URL works; extra query parameters are harmless.

---

📤 Output Format

You get one dataset record per hotel. Property data sits at the top level; reviews are nested in a reviews array.

{
  "hotelId": 49487,
  "hotelName": "Ibis Budapest City",
  "url": "https://www.agoda.com/ibis-budapest-city/hotel/budapest-hu.html",
  "reviewsPageUrl": "https://www.agoda.com/ibis-budapest-city/reviews/budapest-hu.html",

  "score": 8.6,
  "scoreScale": 10,
  "scoreText": "Excellent",
  "reviewCount": 2022,
  "reviewCommentsCount": 2158,

  "gradesPerAspect": {
    "location": 9,
    "staffPerformance": 8.8,
    "cleanliness": 8.7,
    "roomComfort": 8.6,
    "valueForMoney": 8.5,
    "facilities": 8.2,
    "foodDining": 7.1
  },

  "providers": [
    { "providerId": 332,  "name": "Agoda",               "reviewCount": 167,  "textReviewCount": 70 },
    { "providerId": 3038, "name": "Booking.com",         "reviewCount": 1719, "textReviewCount": 769 },
    { "providerId": -999, "name": "Additional providers","reviewCount": 136,  "textReviewCount": 3 }
  ],

  "ratingDistribution": [
    { "bucket": "exceptional",       "name": "9+ Exceptional",       "count": 259 },
    { "bucket": "excellent",         "name": "8-9 Excellent",        "count": 197 },
    { "bucket": "very_good",         "name": "7-8 Very Good",        "count": 217 },
    { "bucket": "good",              "name": "6-7 Good",             "count": 87 },
    { "bucket": "below_expectation", "name": "<6 Below Expectation", "count": 79 }
  ],

  "travelerTypes": [
    { "id": 2, "name": "Couples",        "score": 8.7, "count": 641 },
    { "id": 3, "name": "Solo travelers", "score": 8.7, "count": 309 }
  ],

  "reviewTags": [
    { "tagName": "Location", "mentioned": 8, "positivePercentage": 100, "negativePercentage": 0 }
  ],

  "reviewsAvailable": 842,
  "reviewsCollected": 763,
  "partial": false,

  "reviews": [
    {
      "hotelId": 49487,
      "hotelName": "Ibis Budapest City",
      "hotelUrl": "https://www.agoda.com/ibis-budapest-city/hotel/budapest-hu.html",

      "hotelReviewId": 1057164279,
      "providerId": 332,
      "providerName": "Agoda",

      "rating": 9.6,
      "ratingText": "Exceptional",

      "reviewTitle": "I am satisfied with the Ibis Budapest City Hotel.",
      "reviewComments": "The cleaning was quick and the staff were friendly…",
      "reviewPositives": null,
      "reviewNegatives": null,
      "text": "The cleaning was quick and the staff were friendly…",

      "originalTitle": "이비스 부다페스트 시티 호텔 만족합니다.",
      "originalComment": "청소도 빠르고 직원들도 친절했어요…",
      "originalLanguage": "ko",
      "translatedTo": "en",

      "reviewDate": "2026-07-28T00:00:00+07:00",
      "checkInDate": "2026-07-20T00:00:00",
      "checkOutDate": "2026-07-26T00:00:00",
      "stayMonth": "July 2026",

      "reviewerName": "cho",
      "reviewerCountry": "South Korea",
      "reviewerCountryCode": "KR",
      "travelerType": "Solo traveler",
      "roomType": "Room with 1 double bed",
      "lengthOfStay": 6,

      "helpfulVotes": 0,
      "unhelpfulVotes": 0,
      "photos": ["https://pix6.agoda.net/generic/…jpeg"],

      "responderName": "Ibis Budapest City",
      "responseText": "Thank you for your wonderful review…",
      "responseDate": "2026-07-29T00:00:00"
    }
  ]
}

Property fields

Field	Description
hotelId, hotelName, url	Agoda's numeric id, the property name, and the URL you supplied
reviewsPageUrl	Direct link to the property's review page on Agoda
score, scoreScale, scoreText	Aggregate rating (e.g. 8.6 out of 10, "Excellent")
reviewCount	Every rating Agoda counts, including score-only ratings with no text
reviewCommentsCount	A second count Agoda reports alongside reviewCount; passed through as-is
gradesPerAspect	Sub-scores: location, cleanliness, service, room comfort, value, facilities, food
providers[]	Each review source with its total and its text-review count
ratingDistribution[]	How many reviews fall in each rating band
travelerTypes[]	Score and review count per travel party, count-weighted across sources
reviewTags[]	Topics guests mention, with positive/negative sentiment percentages
reviewsAvailable	Agoda's estimate of reviews with text — see the FAQ, it's an estimate
reviewsCollected	How many this run actually retrieved
partial	true if rate limiting cut the sweep short — treat the record as incomplete

Review fields

Field	Description
hotelReviewId	Unique review id — use it to de-duplicate across runs
providerId, providerName	Which site the review came from
rating, ratingText	Score out of 10 and its label
reviewTitle, reviewComments	Title and body, translated into your chosen language
reviewPositives, reviewNegatives	Liked/disliked halves, where the source splits them
text	The field to read. Merges whichever fields the source filled in
originalTitle, originalComment, originalLanguage	What the guest actually wrote, untranslated
reviewDate, checkInDate, checkOutDate, stayMonth	When the review was posted and when the stay happened
reviewerName, reviewerCountry, reviewerCountryCode	Who wrote it and where they're from
travelerType, roomType, lengthOfStay	Travel party, room booked, nights stayed
helpfulVotes, unhelpfulVotes	Community votes on the review
photos[]	Fully qualified guest photo URLs
responseText, responderName, responseDate	The property's public reply

Why you should read text, not reviewComments

Different sources fill different fields. Agoda's own reviews put everything in reviewComments. Booking.com reviews split into reviewPositives and reviewNegatives and usually leave reviewComments empty — in one measured run, 607 of 680 reviews came from Booking.com. text merges whichever the source used, so it's populated on virtually every row. The individual fields stay available if you need them separately.

---

▶️ How to Use

1. Open the Actor in the Apify Console and click Try for free.
2. Paste your hotel URLs into the Start URLs field — one per line, straight from your browser.
3. Leave the proxy on Residential. It's pre-filled and strongly recommended.
4. (Optional) Tick Fast mode for a quick, cheap sample first.
5. Click Start. Watch the log — it reports each hotel as it finishes.
6. Download your data from the Storage tab as JSON, CSV, Excel or XML — or fetch it from the API.

Run it on a schedule

Use Apify Schedules to re-run weekly or monthly and track how a property's reviews change. De-duplicate on hotelReviewId to append only new reviews.

Call it from code

curl -X POST "https://api.apify.com/v2/acts/YOUR_USERNAME~agoda-reviews-scraper/runs?token=YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "startUrls": [{ "url": "https://www.agoda.com/ibis-budapest-city/hotel/budapest-hu.html" }],
    "proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"] }
  }'

Official clients are available for JavaScript and Python, and every run's dataset is reachable over the Apify API.

---

📈 Use Cases

1. Competitive reputation benchmarking Scrape your hotel plus your closest competitors and compare gradesPerAspect side by side. You'll see precisely where you lose — cleanliness, breakfast, noise — instead of guessing from a single overall score.

2. Guest sentiment analysis with AI Feed the text field into an LLM or sentiment model to surface recurring complaints and praise at scale. reviewTags gives you Agoda's own topic breakdown as a cross-check.

3. Reputation monitoring over time Schedule monthly runs and track score, ratingDistribution and review volume. Catch a downward trend while you can still act on it.

4. Market and location research Compare guest satisfaction across neighbourhoods, cities or star ratings before an acquisition or a new development. travelerTypes shows which segments a market actually serves.

5. Agency reporting and lead generation Reputation-management and hospitality-marketing agencies use review data to build audit reports that win clients — showing a prospect exactly what guests complain about is a far stronger pitch than a generic deck.

6. Multilingual research datasets Use originalComment and originalLanguage to build genuine multilingual corpora for translation, NLP or academic work.

---

🛠️ Advanced Tips

Proxy: what it does and doesn't buy you

Agoda rate-limits per exit IP, and a full sweep is roughly 85–115 requests per hotel, so residential proxy is pre-filled and is the right default — especially for multi-hotel runs, repeated runs, or anything on a schedule, where a single IP would accumulate limits.

Be aware of what it doesn't fix, though. Agoda's throttling is largely random per request rather than a durable per-IP ban: the same filter can return reviews on one attempt and an empty list seconds later on the same IP. What recovers those reviews is the Actor's retry ladder, not a new IP. In our tests a residential run returned 669 reviews and unproxied runs returned 680 and 763 — well within run-to-run variance, with no clear advantage either way on a single property. So use the proxy for the reasons above, but don't expect it to raise your review count on one hotel.

Tune your run for cost or completeness

Goal	Settings
Cheapest sample	fastMode: true
Balanced	maxReviews: 300, includeDemographicSweep: false
Maximum coverage	Defaults, residential proxy, requestDelayMs: 800

Scaling to many hotels

• Split large lists across several runs rather than one very long one — easier to monitor and retry.
• Raise requestDelayMs to 800–1500 when running many hotels, to trip the rate limiter less often.
• Check partial on each record and re-run just those hotels rather than the whole batch.
• Failed URLs are written to the FAILED_TARGETS record in the key-value store with the reason.

Building an incremental pipeline

hotelReviewId is stable across runs. Store the ids you've already seen and insert only new ones — that turns a weekly run into a cheap append rather than a full re-import.

---

❓ FAQ & Troubleshooting

Why did I get fewer reviews than the number Agoda advertises? Agoda's headline count includes score-only ratings — guests who left a star rating but wrote nothing. Those are never served to anyone who isn't logged in, so no scraper can retrieve them. On our test property, 2,022 advertised ratings corresponded to several hundred reviews with actual text. reviewsCollected tells you what you got.

What does partial: true mean? Rate limiting or an internal safety cap cut that hotel's sweep short, so some reachable reviews are probably missing. Re-run that hotel with a residential proxy and a higher requestDelayMs. We flag this deliberately — reporting a truncated run as complete would quietly corrupt your analysis.

Is reviewsAvailable an exact total? No, it's an estimate. Agoda calculates it per request, and the same property has reported different figures depending on how the request was paginated. Use it as a rough coverage check; use partial as the reliable completeness signal.

Do I need a proxy? Keep it on — it's pre-configured, and it matters as soon as you scrape more than one hotel or run on a schedule, because Agoda rate-limits per IP. On a single property we did not measure a clear gain from it; see Proxy under Advanced Tips for the detail.

The run failed with "Agoda refused every attempt." Your IP is rate-limited. Enable or widen the residential proxy configuration and retry. If it persists, raise requestDelayMs.

Can I get reviews in my own language? Yes — set language to any of the 15 supported locales. Agoda translates the reviews, and originalComment still holds what the guest wrote.

Some reviews have text: null. Is that a bug? No. A few reviews genuinely carry only a rating. Some sources send a placeholder sentence ("This guest did not leave comments about this hotel"); we strip that rather than store it as if a guest wrote it, keeping the rating and dates, which are real.

Can I scrape by hotel ID instead of URL? No — URLs only. Copy them from your browser.

Can I filter to only recent reviews? Not through input. Collect the reviews and filter on reviewDate afterwards.

How long does one hotel take? A full sweep is a few minutes per property. fastMode is seconds. Multiply by your URL count.

How do I get a CSV or Excel file? Open the run's Storage tab and pick your format, or append ?format=csv to the dataset API URL.

Is web scraping Agoda legal? This Actor collects only publicly visible review data — no logins, no personal accounts. You are responsible for how you use the data, including complying with the GDPR and applicable local law. If you plan to republish reviews or store personal data, take your own legal advice.

---

📞 Support

Questions, bug reports and feature requests are welcome.

• Email: me@ahmedhrid.com
• Issues: Open an issue on the Actor's page in the Apify Store

When reporting a problem, please include the run ID and the hotel URL that failed. If you can, re-run it with the environment variable APIFY_LOG_LEVEL=DEBUG set — that records every request, retry and proxy rotation, and makes most issues diagnosable straight away.

---

Keywords: Agoda scraper, Agoda reviews scraper, hotel reviews scraper, Agoda API, web scraping, Apify Actor, hotel review data extraction, guest review analysis, hospitality data, travel data scraping, review monitoring, reputation management, Booking.com reviews, hotel sentiment analysis.