Google Hotels Scraper & Rate Monitor: Prices by Booking Site

Get hotel prices from Google Hotels for the exact dates you choose, broken down by booking site: Booking.com, Expedia, Agoda, Trip.com, Hotels.com, Priceline, the hotel's own website and the rest. Each row also has the lowest price, rating, review count and star class. In monitor mode a daily run returns only the prices that changed since the last run, with old price, new price and delta per site.

It's built for hotel revenue managers doing daily rate shopping, small hotel groups and consultants watching a competitor set, and travel-deal sites.

• Real dates, not Google's default. Pick fixed stays (2026-12-04/2026-12-06) or rolling dates ("the next 14 nights from today", "every Friday for 8 weeks"). Every result is checked against the dates Google actually priced. A mismatch is never charged.
• Any guests and currency: adults, currency (USD, EUR, GBP…), country market and language.
• Hotels by name, URL or destination. Use "Hotel Botanico Lisbon", a google.com/travel/hotels/... link, or "hotels in Paris" (top 1–20 hotels, pinned across runs).
• Parity check: the hotel's own-site price against the cheapest OTA (parityGapPct).
• No login, no API key, no browser. Plain HTTP through Apify datacenter proxy, about 1–3 seconds per hotel and date.

Sample output

One hotelDate row from a real cloud run (Oct 2026). The price list is shortened to 3 of 20 sites, and links are cut:

{"type": "hotelDate", "hotelName": "Hotel Botânico", "checkIn": "2026-10-23", "checkOut": "2026-10-25", "nights": 2, "adults": 2,
 "currency": "EUR", "stars": 3, "rating": 4.4, "reviewCount": 873,
 "lowestPrice": 103.94, "lowestPriceSite": "Vio.com", "lowestTotalPrice": 207.88,
 "officialSitePrice": 120.5, "cheapestOta": "Vio.com", "cheapestOtaPrice": 103.94, "parityGapPct": 15.93, "sitesCount": 20,
 "prices": [
   {"site": "Vio.com", "pricePerNight": 103.94, "totalPrice": 207.88, "isOfficialSite": false, "isSponsored": false, "link": "https://deals.vio.com/?sig=…"},
   {"site": "Trip.com", "pricePerNight": 103.98, "totalPrice": 207.97, "isOfficialSite": false, "isSponsored": false, "link": "https://us.trip.com/hotels/redirect?hotelid=3483974…"},
   {"site": "Hotel Botânico", "pricePerNight": 120.5, "totalPrice": 241, "isOfficialSite": true, "isSponsored": false, "link": "https://be.heytravel.net/…"}
 ],
 "googleHotelsUrl": "https://www.google.com/travel/hotels/entity/ChgIj5u_rIHe5tKxARoLL2cvMXRqN2c5eXcQAQ/prices?…"}

A change row from the next scheduled run (real, from the same test):

{"type": "change", "changeType": "priceChange", "hotelName": "The Central House Lisbon Baixa", "checkIn": "2026-10-30", "nights": 2,
 "site": "Kiwi.com", "oldPrice": 136.25, "newPrice": 133.27, "delta": -2.98, "changePct": -2.19, "direction": "down", "currency": "EUR",
 "previousCheckAt": "2026-10-09T16:09:16Z", "detectedAt": "2026-10-09T16:10:41Z"}

How it works

1. Hotels. Names are searched on Google Hotels, and the first hotel listed is used. The matched name is in every row (hotelName, matchedFrom), and the match is remembered for later runs. URLs and entity keys are used directly. For a destination, the top hotels on the first run are saved and checked again on every run, because Google reorders destination results even minutes apart. Turn on refreshDestinationHotels to pick them again.
2. Dates. Google ignores simple checkin= URL parameters. The Actor builds Google's own encoded travel parameter for your dates, guests and currency. It then verifies the stay and currency Google echoes back before saving or charging anything.
3. Prices. For each hotel and stay it reads the hotel's Google Hotels Prices page, which shows every booking site with price per night, total and booking link.
4. Changes. The last prices for each hotel, stay, adults, currency and country are kept in a named key-value store (stateStoreName). The next run compares against them:
   • priceChange: one booking site's nightly price moved (old, new, delta, %).
   • lowestPriceChange: the lowest price moved, or another site became the cheapest.
   • soldOut / backAvailable: the stay has no prices any more, or has prices again.
   • newSite / siteRemoved (off by default): a booking site appeared in the list or left it.

Rolling dates fit this design. A stay is remembered by its calendar date, so "next 14 nights" compares each night with the price seen for that night on earlier runs. A night that enters the window is a first check, and nights in the past are forgotten.

Input example

{
  "hotels": ["Hotel Botanico Lisbon", "https://www.google.com/travel/hotels/entity/ChkIg-b2ismUj7M1Gg0vZy8xMWg3MThreGg1EAE"],
  "destinations": ["hotels in Lisbon"],
  "maxHotelsPerDestination": 5,
  "daysAhead": 1,
  "rollingDates": 14,
  "nights": 1,
  "adults": 2,
  "currency": "EUR",
  "country": "gb",
  "mode": "changes",
  "minChangePct": 2,
  "stateStoreName": "lisbon-compset"
}

Option	What it does
hotels	Names with city, Google Hotels URLs or entity keys
destinations, maxHotelsPerDestination	"hotels in …" searches. The top 1–20 hotels are pinned across runs (refreshDestinationHotels re-picks them)
checkInDates	Fixed stays: YYYY-MM-DD (uses nights) or YYYY-MM-DD/YYYY-MM-DD
daysAhead, rollingDates, rollingStepDays	Rolling check-ins: start N days from today, how many, days between them (7 = same weekday). Without any dates: one stay 7 days ahead
nights, adults, currency, country, language	The search itself
mode	changes (default): full rows the first time a hotel and date is checked, then only change rows. allPrices: a full row on every run, plus change rows
changeTypes, minChangePct, minChangeAmount	Which changes you want and how small is too small
sitesFilter	Track only some sites, e.g. ["booking", "expedia", "official"]
includeSponsored, includeLinks, includeVacationRentals	Extra filters

Output

• hotelDate rows, one per hotel and stay. You get them on the first check, and on every run in allPrices mode. Each row has the hotel name, key, address, coordinates, stars, rating and reviews; the stay (check-in, check-out, nights, adults, currency); available; the lowest price and its site; total for the stay; own-site price; cheapest OTA and parityGapPct; sitesCount; prices[] (site, price per night, total, display strings, own site / sponsored flags, booking link); changesSinceLastCheck; and a Google Hotels link that opens the same search. When a stay has no prices, noPricesSuggestedStay shows the stay Google suggests instead, for example a minimum stay of 3 nights.
• change rows: changeType, site, oldPrice, newPrice, delta, changePct, direction, oldCheapestSite/newCheapestSite, plus the hotel and stay context.
• One summary row per run (free): hotels, dates, checks done or failed, sold-out stays, change counts by type, and the 5 cheapest stays.
• error rows (free, never charged): bad_input, not_found, search_failed, blocked, parse_error, dates_not_applied.

The dataset has ready-made views: Prices by hotel and date, Price changes, Run summary and Errors.

Honest limits

• Prices are what Google Hotels shows for the chosen country. That is usually the cheapest standard room per night. Whether taxes and fees are included depends on the market: US results usually exclude them, while UK and EU results usually include them. Member, mobile-only and loyalty rates aren't visible.
• The list of booking sites varies between runs. Small meta-search sites appear and disappear, so newSite/siteRemoved are off by default. Big OTAs and the hotel's own site are stable.
• Name matching takes the first hotel Google lists. Check hotelName on the first run, or paste the hotel's URL to be exact.
• Destination searches read Google's first results page, so at most 20 hotels per destination. To track more, add hotels by name or URL.
• Sold-out stays are charged as a check. "No prices for these dates" is real information, and the row says so. Failed or mismatched pages are never charged.
• Google can change its page format. The Actor runs a daily self-test, and pages it cannot read become free error rows instead of wrong data.
• About 330 days ahead at most, up to 30 nights, 1–8 adults, and no children yet.

In our Apify cloud tests (Oct 2026): 116 of 116 hotel-date checks succeeded across Lisbon, Paris, London and New York in USD, EUR and GBP, with no captcha or blocked page. About 5% of requests hit a slow proxy connection and were retried automatically. A 60-check run took 93 seconds.

Pricing (pay per event)

Event	Price
Actor start	$0.005 per run
Hotel and date checked (prices read and compared)	$0.003
Price change row	$0.0005

hotelDate rows (first checks and allPrices mode), summary and error rows cost nothing extra.

Examples:

• Competitor set of 10 hotels, next 14 nights, checked daily: 140 checks plus about 30 changes a day = $0.005 + $0.42 + $0.015 = $0.44 per day (about $13 per month).
• 5 hotels on 4 key weekends, daily: 20 checks = about $0.07 per day.
• One-off price grid for 1 hotel × 30 nights: $0.005 + $0.09 = $0.095.

Daily alerts to Slack or email

1. Fill in the input and click Save as a new task. One task per competitor set or client works well.
2. In Schedules, create a schedule, for example every day at 07:00 in your time zone, and add the task.
3. In the task's Integrations tab, add the Slack or Gmail integration for a message when a run finishes. Or add a webhook on "Run succeeded" for Zapier, Make, n8n or your own endpoint, which can read https://api.apify.com/v2/datasets/{defaultDatasetId}/items?view=changes.

In monitor mode every scheduled run's dataset is your alert list: an empty changes view means no price moved. Use minChangePct to hear only about meaningful moves, and turn on Apify's run-failure notifications.

Related actors

This Actor is part of a small competitor-intelligence suite from the same developer. They all work the same way: pay per event, failed items never charged, and monitors that return only what changed since the last run.

• Google Ads Transparency Scraper & New Ads Monitor: what your competitors advertise on Google Search, Display and YouTube.
• LinkedIn Ad Library Scraper & New Ads Monitor: competitors' LinkedIn ads without login, including EU impressions and targeting.
• Bing Ads Library Scraper - Microsoft Ads Monitor (EU): Bing ads from Microsoft's official Ad Library (EU/EEA).
• Google Trends Scraper & API: interest over time, by region/city, top queries and Trending now.
• ATS Jobs Scraper & Hiring Monitor: new and closed jobs from Greenhouse, Lever, Ashby, Workday and 6 more job boards.
• App Store & Google Play Scraper: ratings, versions, chart and keyword ranks of iOS and Android apps.
• Shopify Store Monitor: competitor Shopify stores' new products, price changes and stock.

FAQ

Do I need a proxy? No. The default Apify datacenter proxy worked for every check in our tests. A residential fallback is available but off by default.

Can I get prices for one hotel over many dates? Yes. Use rollingDates (up to 60 check-ins per run) or list the stays in checkInDates.

Why did my destination always return the same hotels? They are pinned on purpose, so the monitor compares like with like. Set refreshDestinationHotels to re-pick them, or use a new stateStoreName.

Is it legal? It reads publicly available prices from Google Hotels. You are responsible for how you use the data and for complying with applicable law and Google's terms.