Hotel Price Scraping: Rate Parity & Competitor Monitoring with Booking.com Data
Hotel prices change by the hour. A room that costs one amount on Monday can cost another on Friday, and a competitor two streets away may undercut you without any announcement. Hotel price scraping is the practical way to see those moves: you collect the public prices shown on booking pages, store them with a timestamp, and compare them over time.
To make this concrete, we collected a free sample of 100 Barcelona listings from Booking.com. You can download it on the Booking.com Barcelona dataset page. This article uses that file to show how price relates to star rating, distance to the center and free cancellation, and then explains how teams turn the same approach into rate parity monitoring and competitor tracking.
Why hotel price scraping matters
Revenue managers, channel managers and travel analysts all face the same problem: the market price is public, but it is spread over thousands of pages that change constantly. Checking them by hand does not scale beyond a handful of competitors.
Automated collection solves three recurring needs:
- Competitive set tracking. See what comparable hotels charge for the same night and the same occupancy.
- Rate parity checks. Confirm that the price on a third-party channel matches the price on your own site or other channels.
- Market research. Understand how location, stars and cancellation terms translate into price in a given city.
If you want the broader business case for this kind of work, our post on web scraping as a competitive advantage covers it. For another travel source, see TripAdvisor travel insights.
What our Booking.com sample shows
The sample covers one night (check-in 2026-11-21, one room, two adults), priced in EUR as displayed on the result page. It has 100 properties and 12 columns. Here is what we computed from it.
- Average total price: €230.58. Median: €193.09. The range runs from €59.57 to €656.40. The average sits above the median because a few luxury properties pull it up.
- Stars drive price. The correlation between price and star rating (82 properties with a star rating) is 0.55. The 22 five-star hotels average €390.24, the 35 four-star hotels €226.86 and the 16 three-star hotels €218.03. The jump from four to five stars is much bigger than from three to four.
- Unrated properties are cheapest. 18 listings have no official star rating and average €94.59.
- Distance matters, but less than stars. Price and distance to the center have a correlation of -0.21. The 47 properties within 1 km of the center average €257.42; the 28 between 2 and 4 km average €201.26; the 10 beyond 4 km average €187.68. The median distance in the sample is 1.15 km.
- District spread. Eixample (32 properties) averages €281.24, Ciutat Vella (28) €229.71 and Sant Martí (8) €158.92.
- Free cancellation is the minority. 20 of 100 listings (20%) show free cancellation. Those average €261.91 against €222.74 for the rest. By class, 31.4% of four-star hotels offer it, but only 9.1% of five-star hotels.
- Guests rate the market high. The average review score is 8.44 out of 10.
An excerpt of the data:
| Hotel | Stars | District | Distance to center (km) | Review score | Total price (EUR) |
|---|---|---|---|---|---|
| Eurostars Grand Marina Hotel GL | 5 | Ciutat Vella | 1.9 | 8.5 | 169.57 |
| Barceló Sants | 4 | Sants-Montjuïc | 2.7 | 8.7 | 164.24 |
| Grand Hyatt Barcelona | 5 | Les Corts | 3.9 | 8.3 | 249.40 |
| Via Sants Hotel Barcelona, Tapestry Collection By Hilton | 4 | Eixample | 2.3 | 8.6 | 152.75 |
| Barcelona Princess | 4 | Sant Martí | 4.8 | 8.8 | 133.23 |
One caution: this is a sample for one night and one search, not a full census of Barcelona. It combines several result orderings of the same search, and the share of free cancellation reflects what the result page displayed at the time. Treat the figures as an illustration of the method, not as a market benchmark.
Rate parity monitoring with Booking.com data
Rate parity means that a hotel offers the same price for the same room and conditions across channels. When a room is cheaper on one channel than on the hotel's own site, direct bookings suffer and contract terms may be at risk.
Rate parity monitoring with this kind of data works like this:
- Pick the stay dates and occupancies that matter, for example the next 30 days, one night, two adults.
- Collect your own price and the prices on the channels you want to compare, with the same filters.
- Match records by property and room or rate type.
- Flag any gap above a threshold, for example more than 2%.
- Keep the history so you can see whether a gap is a glitch or a pattern.
The fields that make this possible are hotel_name, hotel_url, checkin_date, price_total, currency and free_cancellation. Without the cancellation flag you can end up comparing a non-refundable price with a flexible one, which looks like a parity breach but is not.
Hotel rate shopping: building a competitive set
Hotel rate shopping is the competitor side of the same coin. Instead of checking one property across channels, you check many properties on one channel.
A useful competitive set is narrower than "all hotels in the city". Filter on the dimensions the sample shows matter:
- Stars. Compare four-star hotels with four-star hotels. In our sample the gap between classes is large enough to distort any average that mixes them.
- Distance to center. A hotel 0.5 km from the center and one 4 km away serve different guests. Use bands such as under 1 km, 1 to 2 km and 2 to 4 km.
- District. Barcelona districts behave differently. Eixample and Ciutat Vella tend to sit higher than outer districts in this sample.
- Cancellation terms. Compare like with like: flexible with flexible.
Once the set is defined, track the median price of the set per check-in date, and compare your own price with that median. A median is more stable than an average when a few luxury properties are in the group.
Which fields matter and why
| Field | Why it matters |
|---|---|
price_total and currency |
The core signal; always keep the currency next to the number |
checkin_date |
Prices are date-specific, so a price without a date is useless |
stars |
Strongest single predictor of price in the sample |
distance_to_center_km |
Lets you build distance bands |
district |
Separates micro-markets inside one city |
review_score and review_count |
Quality context; a high price with a low score is a different story |
free_cancellation |
Separates flexible and restrictive rates |
hotel_url |
Stable key for matching records across runs |
scraped_at_utc |
Needed to build a price history |
Use cases
- Revenue management. Compare your rate with a competitive set every day and adjust.
- Channel managers and agencies. Run parity reports for several hotels and show the gaps.
- Travel startups. Seed a price-comparison or deal-alert product with structured data.
- Tourism boards and analysts. Track how prices move around events, holidays and conferences.
- Investors. Compare price levels and quality scores across districts or cities before a purchase.
Common challenges in hotel price scraping
Travel sites are among the harder targets, and it is worth knowing why before you plan a project.
Anti-bot measures. Booking pages respond differently to plain requests, residential IPs and full browser sessions. A setup that works on Monday can start returning blocked pages on Thursday. We run our collection on managed infrastructure that escalates from simple requests to rendered sessions only when needed.
Pagination that does not behave. In our own run, the offset parameter did not return later result pages reliably; several requests came back with results we already had. We worked around it by collecting different orderings of the same search and removing duplicates. This works for a sample, but a full-market project needs a more careful strategy, for example splitting the city by district or price band.
Dates, occupancy and currency. A price is only meaningful with its check-in date, length of stay, number of guests and currency. Change one of them and the result is a different dataset.
Availability. Some properties are sold out for the chosen date and show no price. Decide upfront whether to keep them as empty rows or drop them.
Layout changes. Page markup and embedded data structures change without notice. Parsers need tests and monitoring so that a silent change does not produce a week of empty or wrong data.
Data quality. Normalize prices into plain numbers, convert distances such as "700 m" into kilometers, and deduplicate by property. In our sample, 18 of the 1,200 cells (1.5%) were empty, mostly star ratings for properties without an official class.
How often should you refresh the data?
It depends on the job:
- Rate parity monitoring: daily, or several times a day for key dates.
- Competitor rate shopping: daily for the next 30 to 60 days of check-ins.
- Market research and reporting: weekly or monthly snapshots are enough.
- Event periods: increase the frequency around trade fairs, holidays and sports events, when prices move fastest.
Always store each run with its timestamp. A single snapshot tells you the price; a series tells you the pattern.
FAQ
Is it legal to scrape hotel prices? Public prices shown to any visitor are commonly collected for market research, but rules differ by country and by site terms. We collect public data only, no login-protected pages and no personal data. For a commercial project, check your own legal requirements.
How many hotels does a typical Booking.com search return? Our Barcelona search reported 749 results in total on the result pages we loaded. The free sample contains 100 of them.
Can I compare prices for several check-in dates?
Yes. Each check-in date is a separate collection run, and the checkin_date column keeps them apart. Many teams track the next 30 to 60 days.
Why do averages and medians differ so much? A few five-star properties are priced far above the rest. In our sample the average is €230.58 and the median €193.09, so use the median for a competitive set.
Get the data
Download the free Barcelona sample and test it against your own revenue questions. If you need other cities, more dates or a daily refresh, DataHarbor delivers datasets from $49; see pricing.
Want data from a different hotel market or channel? Request a free sample of your own target and we will show you what the data looks like before you commit.
Author: DataHarbor Team