How CASAFARI processes property market data
CASAFARI aggregates real estate advertisements from across Europe and organises them into a clean, deduplicated property graph. Our metrics measure asking prices set by sellers and landlords, providing accurate market indicators for professional reporting.
Where does the data come from?
Every property record is created through a structured five-step methodology.
Collection
We collect advertisements continuously from 30,000+ portals and agency websites as listings appear and change.
Structuring
Every advertisement is structured into one schema covering type, operation, price, area, rooms, features, condition, energy rating, and location.
Cleaning
Values are normalised across sources, locations are resolved to a geographic tree with coordinates, and statistical outliers are flagged.
Deduplication
The same property advertised across multiple portals and agency websites is recognised as one distinct entity.
Merging
All matching advertisements are merged into a single record per property containing its complete historical trajectory.
What do the figures measure?
CASAFARI metrics are computed from asking prices in advertisements rather than final transaction prices from land registries or notary deeds.
The available time-series metrics per place and segment comprise:
- Average asking price: The mean listed price for properties in the selected segment.
- Average asking price per square metre: The mean asking price divided by the property area in square metres.
- Number of properties on the market: The total volume of active listings during the chosen period.
- New properties on the market: The count of listings introduced during the period.
- Properties withdrawn as sold or rented: Listings removed from the market following a transaction.
- Price increases: The count of properties where the asking price was raised.
- Price decreases: The count of properties where the asking price was reduced.
Distributions are available by asking-price band along with the average time on the market, by bedroom count, and by overall time-on-market distribution. Segments cover apartments or houses for sale or for rent, with optional filters including bedrooms, area, condition, features, private sellers, professional agencies, bank-owned properties, and auctions.
Which places and periods can be covered?
Locations are specified using location ids from Location Lookup across 16 countries: Andorra, Austria, Belgium, France, Germany, Greece, Italy, Luxembourg, Monaco, Poland, Portugal, San Marino, Spain, Switzerland, the United Arab Emirates, and the United States (US: entering now, selected states, commercial real estate only, via Claro AI). Queries can select any level of the location tree, combine up to ten places, or define a circle from 50 metres to 50 kilometres around coordinates. Time periods can be set to weeks running Monday to Sunday, months, quarters, or years.
How are outliers and duplicates handled?
Properties with asking prices significantly below or above similar properties in the same segment are excluded from default calculations. Because our property graph deduplicates listings, a home advertised by five different agencies counts exactly once in all market figures.
What are the limitations of the data?
To interpret CASAFARI metrics accurately, please note the following points:
- Figures represent asking prices, which may differ from completed transaction prices.
- Coverage and historical depth vary by country and region.
- Small places and narrow market segments contain fewer properties, causing figures to fluctuate more between periods.
- Recent time periods can adjust slightly as historical listing details are completed.
- United States: CASAFARI is entering now, in selected states, for commercial real estate only, via Claro AI. There is no residential asking-price data.
How to cite CASAFARI figures
Journalists may quote and chart figures published by CASAFARI, provided the citation indicates the period, source, and asking-price basis.
Example citation: "Average asking price per m² for apartments for sale in Lisbon, September 2026. Source: CASAFARI (asking prices)."
For details on republishing full tables or contacting the press office, please visit the main press page.
Frequently asked questions
Direct responses regarding our data source, calculations, and coverage.
Where does CASAFARI's data come from?
CASAFARI gathers real estate advertisements continuously from over 30,000 portals and agency websites across 16 countries. The raw listings are structured, cleaned, and merged into a unified property graph to track the market.
How does CASAFARI handle duplicate listings?
Our deduplication engine analyses property features, addresses, and spatial coordinates across all portal and agency listings. Multiple advertisements for the same property are combined into one master record, ensuring each home is counted only once.
Why can asking prices differ from sale prices?
Asking prices represent the initial or updated figure requested by sellers and landlords during marketing. Final transaction prices agreed upon at closing or recorded in public registries may differ due to negotiations or market adjustments.
How are outliers treated in the figures?
Listings with asking prices that are significantly higher or lower than comparable properties in the same segment are flagged as statistical outliers. By default, these listings are excluded from computed averages to present a reliable view of market trends.
Which periods and places can the figures cover?
Figures can cover weekly periods running Monday to Sunday, months, quarters, or full years. Geographically, data can be retrieved for specific administrative divisions within our 16 covered countries, combined across up to ten places, or defined by a circle between 50 metres and 50 kilometres around a point. In the United States, CASAFARI is entering now, in selected states, for commercial real estate only, via Claro AI.
Can figures change after publication?
Metrics for recent periods may refine slightly over time as property histories are completed. When listing status changes or additional details are confirmed, historical timeline updates ensure total precision across the dataset.
Require custom data for a story?
Our press office can assist with specific market queries and prepare dataset extracts on working days.