This project analyzes hotel pricing in Amsterdam using data collected from the Booking.com public API (via RapidAPI). The goal is to understand what drives hotel prices and identify properties that appear overpriced or underpriced compared to the market.
The analysis combines price, star rating, review score, review count, and location to build a holistic mispricing detection framework. The final output includes visual analysis, mispricing examples, and pricing recommendations.
Hotels on Booking.com aim to balance competitive pricing with strong revenue performance. However, mispriced hotels — either too high or too low — reduce visibility, harm conversion, and lower guest satisfaction.
This project helps answer:
- Which factors influence hotel pricing in Amsterdam?
- Which hotels appear overpriced or underpriced compared to their peers?
- How can pricing be improved to increase competitiveness and fairness?
Data was collected using the Booking.com Hotel Availability API (via RapidAPI).
Endpoints used:
/searchDestination — Fetch destination ID for Amsterdam
/searchHotels — Retrieve listings, prices, ratings, and location details
Data extracted:
hotel_id, name
priceBreakdown.grossPrice.value
accuratePropertyClass (star rating)
reviewScore, reviewCount
latitude, longitude
Dataset size: 220 hotels (1-night stay, standard room for 2 people)
The analysis follows a structured pipeline:
1️. Price distribution by star category
Examines typical price ranges and identifies outliers.
2️. Review score vs star rating
Evaluates guest satisfaction versus expected hotel class.
3️. Price vs review score
Checks whether higher ratings correlate with higher prices.
4️. Price vs review count
Shows how trust/popularity influence pricing stability.
5️. Location-based pricing
Maps hotels across Amsterdam to identify premium and budget zones.
6️. Mispricing detection
Hotels flagged as mispriced if they show:
Price far outside category norms
Weak review score or review count
Location that does not justify pricing
Or a combination of these signals
7️. Pricing recommendations
Clear, actionable adjustments for both overpriced and underpriced hotels.
1. Amsterdam is a mid–upper range hotel market
Most hotels fall into the $200–$400 price band.
2. 3★ and 4★ hotels form the biggest group
They dominate the supply and are spread widely across the entire city.
3. Star rating alone does not predict price
0★, 2★, 3★, and 4★ hotels all overlap in the $200–$500 mid-range.
4. Review score better reflects real value
Some 0★ hotels score as highly as 5★
1★ sometimes outrank 2★
-> Review score is essential for validating price.
5. Review count signals trust & popularity
High-trust hotels cluster predictably in the mid-market pricing band.
6. Location is the strongest price driver
Premium clusters: Old Centre, Museum Quarter, De Pijp
Mid-range hotels: widely spread across the city
Budget hotels: outer districts + central hostels
7. Mispricing exists on both extremes
Some 3★ hotels charge $1,800+ with weak reviews
Some high-performing hotels are priced far below peers
Reduce prices to category norms
Avoid premium pricing until trust is established
Align pricing with location quality
Improve service, amenities, and visual presentation
Increase prices 10–30% toward peer median
Use dynamic pricing for weekends & events
Highlight strong reviews to justify higher pricing
Raise prices gradually while monitoring conversion
Python
Pandas (data cleaning & wrangling)
Plotly (interactive visualizations)
Mapbox (location maps)
Jupyter Notebook
Canva (presentation)
Anna Igumnova
Data Analyst — Pricing & Hospitality