Methodology

1. Source

All figures come from the listing and review files that Inside Airbnb publishes city by city under Creative Commons Attribution 4.0 International (CC BY 4.0). Inside Airbnb collects the public content of listings at a given date. STR Barometer uses one snapshot per city: Brussels on 20 September 2026; Antwerp on 23 September 2026; Ghent on 22 September 2026; Dublin and London on 17 September 2026. A snapshot is a picture of the listings visible on that date; the review history it contains goes further back and is used for seasonality.

Inside Airbnb covers one platform. Homes offered only on other platforms, or directly by their owners, are not in the files, so none of the indicators is a count of all short-term rentals in a city.

2. Which listings are counted

• All listings: every row of the listing file, whatever its type.
• Entire homes: listings of type "Entire home/apt" — the guest has the whole place. Private rooms, shared rooms and hotel rooms are excluded from every indicator except the entire-home share.
• Active entire homes: entire homes with at least one review in the twelve months before the snapshot and an asking price above zero and below 2,000 in local currency. This is the base of all price, nights and revenue indicators.

3. The indicators

• Nightly rate: the asking price shown on the listing on the snapshot date, in local currency (euro, or pound sterling for London), before cleaning and service fees. It is a price asked for one date, not an average price paid over the year.
• Nights booked: Inside Airbnb's own estimate for the last 365 days. Its published model (data assumptions) assumes that about half of stays lead to a review, multiplies the resulting bookings by an average stay (three nights by default, or the listing's minimum stay when that is longer) and caps the result at 70% of the year — which is why no figure here exceeds 255 nights. We express it in nights and, for readability, as a share of the 365-day year ("estimated occupancy").
• Gross revenue: Inside Airbnb's estimate of revenue over the last 365 days, built from the asking price and the estimated nights. It is before platform fees, cleaning, management, costs and any levy.
• Entire-home share: entire homes divided by all listings in the file.
• Seasonality: the share of each calendar month in all reviews written over the 24 complete months from September 2024 to August 2026. Reviews follow stays, so the profile approximates when guests stay; an even year would give 8.3% per month.

4. Statistics and thresholds

• Every value is a median (half the homes above, half below) or a quartile: the middle half runs from the lower quartile (25%) to the upper quartile (75%). Medians are used because a few very expensive or very busy homes would distort an average.
• A city area is shown only with at least 25 active entire homes; a segment (bedrooms, host size, minimum stay, rating, capacity) only with at least 30. Smaller groups are hidden, never estimated.
• No figure is converted between currencies: London stays in pounds, the other four cities in euros. Compare London with the others through nights booked, shares and seasonality, which carry no currency.
• The structural findings (host portfolios, minimum stay, ratings, capacity, dormant listings) are computed for the three Belgian cities only, because the detailed file used for them covers those three.

5. What the barometer does not say

• It does not measure housing stock, residents or the total number of short-term rentals in a city.
• It is not a time series: each figure describes one snapshot. A later edition will be computed with the same script, so the two can be compared line by line.
• Differences between groups (for example rating bands) are associations in the data, not causes.

6. Corrections

Each page is generated from the source files by a single script; a correction is made in the script and applies to every page at once. Confirmed errors are corrected on the page concerned and described here. To report one, see the about page.