Accounts are classified from their names, and the Cardiff accounts with the most listings were checked by hand. Only property-management businesses are ranked, under the name they trade as; individual hosts who don't trade as a business count in every market figure but are never named, and neither are accounts with a generic name, such as "Property" or "Concierge", that can't be tied to a company. Hotels and serviced-apartment operators that list their own rooms are left out, and so are companies that mostly take stays of a month or more, because they compete with rentals rather than holiday lets.
"Lead vs similar" compares each of a company's homes with the median owner-run Cardiff home of the same bedroom count and room type, and reports the median of those comparisons. The figure is shown where a company's homes earn more than similar ones. A lower figure isn't shown: on Airbnb data alone it mostly reflects listings that are new or re-listed and still building reviews, or homes that differ in ways a bedroom count misses, rather than the management. Only homes listed for the whole 12 months and open to bookings on at least 240 nights, booked on 60 or more, with 5 or more reviews count, on both sides, so a company that added homes mid-year isn't marked down for their short first year, a listing that sits idle or is too new to have reviews doesn't stand for its company, and an owner who rents out their own home a few weekends a year doesn't set the benchmark. Companies with fewer than five such homes show a dash, and those with fewer than ten show "few": too few to compare fairly, so their figures are never quoted. The ranking by revenue needs at least 15. The shortlist by home size needs five homes of a size and marks a figure on fewer than ten as a small sample. The ± bands treat homes as independent; homes run by one company resemble each other, so the true uncertainty is somewhat wider. The profiles split revenue into revenue per booked night and nights booked, on the same homes, each against the owner-run median for that measure. A home's revenue is the one times the other, but the three medians don't multiply: among owner-run homes, the ones that charge more tend to book fewer nights, so the median rate times the median nights isn't the median revenue. A company's rating averages its listings' overall Airbnb ratings, weighted by their reviews. Revenue is gross Airbnb booking revenue, estimated from booked nights and rates, before any management fee. Stays a company or owner sells on other sites or direct aren't counted, so a company that sells much of its calendar elsewhere can look weaker here than it is; that applies to owner-run homes too. A home counts toward the company named on it now, so a home that changed manager during the year brings months run by its previous one.
The fee figures pool the rates management companies publish on their own websites, checked by hand in September 2026 in the 11 markets AirROI researched. No company's rate is shown on its own: most quote each home on request, so the page gives the spread of the published rates instead.
Growth counts the listings a company hosts on its own Airbnb accounts now against 12 months earlier. Companies that mostly co-host on owners' accounts show no growth figure, because Airbnb's co-host records from a year ago are incomplete, and accounts with no listings a year ago are marked new rather than ranked for growth.
The page is rebuilt each quarter from the latest listing data. A company that thinks its figures are wrong can write to AirROI, and corrections are checked against the listing data.