Short-term rental data, explained.

Short-term rental data describes the supply of homes rented by the night on platforms like Airbnb and Vrbo: which properties are listed, what they charge, how often they are booked, what guests say and who operates them. Four groups buy it for four different reasons, and each cares about a different part of the file.

What a listing row contains

The platform, the listing identifier, the property type and size, the location as the platform discloses it, the amenities, the host identifier, the nightly rate and the cleaning and service fees, the minimum stay, the review count and rating, and the dates the listing was first and last seen. Everything else, occupancy, revenue, and the owner behind the host, is derived.

How occupancy and revenue are estimated

Platforms do not publish bookings. What can be observed is the listing's calendar: which nights are available and which are blocked, and the price on each. Sampling that calendar repeatedly and watching nights move from available to blocked, then checking against review timing, produces an estimate of booked nights. Multiply booked nights by the observed rate and you have an estimate of revenue. The estimate is honest about being an estimate: a night blocked because the owner is using the home looks the same as a night that was booked, and the methods to separate them are probabilistic.

Two rules follow. First, compare providers on the same listings and the same period, because different sampling frequencies produce different numbers. Second, prefer a provider who shows you the calendar observations behind the estimate, because that is the evidence.

Why aggregation matters

The same house is often listed on two platforms, and sometimes twice on one. A market-level count that does not deduplicate overstates supply. Aggregation means matching listings across platforms on location, photos, size and host details, keeping one record per property, and rolling the result up to a market, ZIP or city with listing counts, active share, median rate and occupancy, and the change against the prior period. That rolled-up series is what a municipality, a hospitality analyst or a lender actually uses.

Tracing a listing to a parcel and owner

Platforms disclose approximate locations. Matching a listing to a specific parcel uses the location, the photos, the property characteristics and, where present, fragments of the address, and it produces a probabilistic match with a confidence tier. Once matched, the assessor file supplies the owner of record and the mailing address, which is how an LLC-held rental is traced to a person. Enforcement programs act on the high-confidence tier and verify the rest. Prospecting programs use the owner match to reach the person who decides on management, insurance and services. The match is US only, because it depends on the parcel file.

The four buyers

Investors want to underwrite a purchase against what nearby rentals actually earn rather than a projection. Municipalities and their vendors want to find unregistered rentals and collect occupancy tax. Property managers, insurers, cleaning and smart-home companies want the host and owner to prospect. Hospitality and tourism analysts want the market series. The same file serves all four, cut differently.

History and backtesting

Because listings are observed continuously, the history is a series of dated observations, and a market's supply and rate can be reconstructed as of a past date. That is what makes the data usable for a backtest or for a regulation-impact study: you can see what supply looked like the month before a rule took effect and the months after.

Questions