Who measures short-term rentals and how

Three ways the figures are made

Hawaii’s Department of Business, Economic Development and Tourism (DBEDT) put the occupancy of the state’s vacation rentals at 46.8% in August 2026, against 72.7% for its hotels, using data compiled by the analytics firm Lighthouse¹. Figures like these come from producers that work in one or more of three ways.

• Reading the booking sites. A firm collects what Airbnb, Vrbo or Booking.com show for each listing: the nightly price, the calendar of open and closed nights, reviews and host details. The Airbnb calendar does not show whether a closed night was booked by a guest or blocked by the host, so these firms estimate which closed nights were bookings³⁴.
• Receiving booking records. Property managers’ software, known as a property management system or PMS, sends confirmed reservations and their revenue to the data firm, which pools them by market. KeyData calls this direct data and describes it as records of actual reservations and the revenue tied to them⁷.
• Agreements and registers. Platforms share aggregated data with statistics offices, and governments register rental units. Under an agreement the European Commission announced on 5 March 2020, Airbnb, Booking, Expedia Group and Tripadvisor agreed to share the number of nights booked and the number of guests with Eurostat, aggregated by municipality⁸.

Several firms combine the first two methods, as the sections below show⁹¹⁷¹.

AirDNA

AirDNA, headquartered in Denver, Colorado, sells short-term rental data and analytics to hosts, investors and businesses¹⁰. Its methodology page says it analyses “100% of listings on Airbnb, VRBO, and Booking.com every single day”, collecting daily prices, calendar availability and host information, and that with its direct data this comes to “over 16M listings every day”⁹. Property managers and hosts from “100k+ properties” share reservation data with the company, which says it uses that verified data as a quality-control layer on its analysed figures⁹.

Its help centre, in an article dated 19 June 2026, calls scraped data its primary source. AirDNA says it has gathered Airbnb and Vrbo data globally since 2015 and Booking.com data since 2019, and that Booking.com data was added to its platform in June 2026¹¹. Listings found only on Booking.com have data from March 2025 onward, which the article says may affect year-over-year trend comparisons, and the main metrics combine Airbnb and Vrbo data by default¹¹.

AirDNA defines a listing’s occupancy as reserved days divided by active listing nights over the past twelve months. A night counts as active if it was not blocked and was reserved or available, provided the listing had a reservation within the past 28 days¹². To tell guest bookings from host blocks, the company says it uses machine learning and “16 different booking signals”, such as length of stay, lead time and reviews¹². A listing whose calendar was blocked for a whole month is not treated as truly active, because no reservation was possible¹²¹³.

AirDNA publishes its own test against Airbnb’s reported results. For the third quarter of 2023 it lists Airbnb’s reported 113.2 million nights booked beside its own estimate of 113.2 million, Airbnb’s reported 19% year-over-year growth in active listings beside its estimate of 18.1%, and Airbnb’s reported 3.3% year-over-year rise in booked average daily rate beside its estimate of 4.5%¹⁴. The comparison and its framing are AirDNA’s.

KeyData

KeyData (KeyData Dashboard, Inc.) says it was founded in 2018; its help centre also writes the name as Key Data¹⁶¹⁸. Its homepage lists “80+ PMS integrations”, “17,000+ property managers” and “10 million+ properties tracked”, and says it covers hotels as well as vacation rentals¹⁵.

Its how-it-works page says the platform combines scraped data from Airbnb and Vrbo with direct data from property management systems and hotel reservation systems, all updated at least once a day¹⁷. The help centre separates the two: direct data comes from customers’ PMS and records actual reservations and revenue, while scraped data shows the rates and “blocks” that Airbnb and Vrbo display, which are not reservations fed in directly⁷. The homepage describes the platform’s decisions as built on confirmed bookings, “never scraped or modeled”¹⁵.

KeyData publishes two occupancy measures. Paid Occupancy % is guest nights divided by all nights in the period, regardless of owner or hold nights; Adjusted Paid Occupancy % is guest nights divided by the nights available for guests to book¹⁸¹⁹. In its own example, if owners booked 10% of the nights and 5% were holds, guest occupancy could reach at most 85%¹⁸. To protect the managers who contribute, its benchmarking report shows market data only where at least 3 other partners have properties in the market and no single manager holds more than 70% of the market’s property share²⁰. Its reports page lists a quarterly KeyData Index and a monthly Short-Term Rental Pulse²¹.

Lighthouse and Transparent

Lighthouse is the new name of OTA Insight. On 10 March 2022 OTA Insight announced it had bought Transparent, a Madrid-based firm whose data, the announcement said, covered “more than 35 million listings across the globe” from public and proprietary sources including Airbnb, Vrbo, Booking.com and industry partners²². On 9 November 2023 OTA Insight said it had rebranded as Lighthouse and that data from acquisitions including Transparent was now part of the Lighthouse platform²³.

Lighthouse’s data-solutions page gives one combined figure, “16.4 million” hotels and short-term rentals profiled daily, and offers data through APIs, data feeds, custom dashboards and scheduled reports, priced on the volume and scope of data required²⁴. The page does not set out its short-term rental method. The fullest public description found for this guide comes from a client, DBEDT, whose monthly report says the data cover properties listed on Airbnb, Booking.com, Vrbo and TripAdvisor, and that Lighthouse enriches its occupancy and price calculations with reservation data from vacation rental software providers, online travel agencies and local property managers¹.

The same report records three method notes. Lighthouse now reports prices as a “Total Rate” that includes fees such as cleaning and service fees, with all history from January 1, 2019 reprocessed; in July 2026 it upgraded the algorithm that removes listings duplicated across platforms and restated history; and reservation data from its partners represent “about 37.3 percent” of the estimated unique vacation rental properties in the state¹.

PriceLabs Market Dashboards

PriceLabs’ Market Dashboards report on the listings around an address the user supplies, covering 1000, 5000 or 10,000 listings depending on the plan bought⁶. The help centre says the dashboards use scraped data from Airbnb and Vrbo, and that a booking taken on another channel is still captured, because a synced calendar blocks the same dates on Airbnb; new bookings usually take 1 to 2 days to appear⁴⁶.

When consecutive dates close between two scrapes, PriceLabs first records a single booking, then applies block-removal logic that weighs factors such as length of stay and booking window; it removes any stay longer than 60 days⁴⁵. Revenue is estimated from the last price seen, without cleaning or extra-guest fees⁵. The help centre states that there is no guaranteed way of telling whether unavailable dates were booked⁴.

Beyond and Wheelhouse

Beyond, which says it has helped short-term rental owners and managers with dynamic pricing since 2013, describes its market intelligence as coming from “direct integrations with booking channels, consumer search demand and property management systems”²⁵. Its Insights product benchmarks a portfolio that the operator connects, and can be used without Beyond’s pricing tool²⁵. Wheelhouse’s market insights page claims “1,500+ markets, 100+ countries” and offers a free trial²⁶. Neither company’s pages read for this guide set out how listings are counted or how booked nights are told apart from blocked ones.

Airbtics and AllTheRooms

Airbtics says it collects publicly available Airbnb listing information, including calendars, advertised prices, property details and reviews, and generally scans listings weekly²⁷. Its homepage claims “30,000+” markets and “15M” Airbnb listings tracked, with data since January 2021, and offers a free account to analyse a first market²⁷. It says its model uses calendar changes, reviews and stay patterns to pick out likely reservations, and that these checks “cannot identify every night perfectly”²⁷.

AllTheRooms, a short-term rental analytics firm based in Wilmington, Delaware, was bought by Deckard Technologies, which runs the Rentalscape short-term rental management and tax compliance platform; the deal database PrivSource dates the announcement to August 7, 2025²⁸. The AllTheRooms analytics address now opens a page headed “Deckard Technologies - Analytics and Tourism Reporting”, and Booked News could not read a current methodology there.

Inside Airbnb

Inside Airbnb was founded in February 2015 by Murray Cox, whom the project describes as an artist, activist and technologist, and since 2023 it has been sustained by a not-for-profit, the Housing Justice Data Lab²⁹. It provides data and advocacy about Airbnb’s impact on residential communities and states that it is not associated with or endorsed by Airbnb or its competitors²⁹³. Its data are snapshots of Airbnb listings at a particular time; quarterly data for the last year for each region can be downloaded free under a Creative Commons Attribution 4.0 International License³³⁰.

To estimate occupancy it uses a modified “San Francisco Model”. It assumes a review rate of 50% to turn reviews into estimated bookings, multiplies those bookings by an average length of stay, using 3 nights where a city has no published figure or the listing’s minimum stay if that is longer, and caps occupancy at 70%³. It chose 50% because it sits almost exactly between a review rate of 72%, which it calls unverifiable, and one of 30.5%, which its own analysis found more fact-based but probably not conservative enough³.

The assumptions page lists the limits: the Airbnb calendar does not distinguish a booked night from an unavailable one, listings can be deleted, map locations sit within 0-450 feet (150 metres) of the real address, and the accuracy of information compiled from Airbnb is not Inside Airbnb’s responsibility³. A paper by Abdulkareem Alsudais, first posted in July 2020 and published in Decision Support Systems, reported incorrect data added to the dataset, linked it to a feature Airbnb introduced and to systemic errors in data collection, and examined why two releases of the dataset are hard to compare³¹.

STR, the hotel benchmark

STR measures hotels, not rentals, and CoStar Group now sells its data as CoStar with STR Benchmark. Its figures come from the hotels themselves: CoStar cites STR’s “directly-sourced top-line sample of 94,000 hotels and 12 million rooms”³². Reports sometimes set hotel and rental figures side by side. Hawaii’s vacation rental report, which prints both, notes that rental units are not necessarily available year-round or every day of the month and often hold more guests than a hotel room¹.

Government statistics and registers

Eurostat publishes figures built from data the platforms share with it. It reported that guests spent 951.6 million nights in short-stay accommodation booked online through Airbnb, Booking or Expedia in 2025, an 11.4% increase compared with 2024³³. Tripadvisor, the fourth platform in the 2020 agreement, announced its exit from that line of business in late 2024 and no longer provides data⁸³⁴. Eurostat publishes the figures for countries and regions, in the datasets tour_ce_omr and tour_ce_omn12³³.

New EU rules add a second official source. Regulation (EU) 2024/1028 on data collection and sharing for short-term accommodation rentals applies from 20 May 2026. It provides for registration numbers that identify each rented unit and requires platforms to send information on rental activity to authorities through a single digital entry point³⁵.

City registers are a third. New York City adopted Local Law 18 on January 9, 2022. It requires hosts to register with the Mayor’s Office of Special Enforcement and requires booking services to use the city’s verification system and stop processing unverified transactions; the office’s first phase of enforcement, from September 5, 2023, focused on the booking platforms³⁶. The office publishes its registration dataset, the latest dated January 7, 2026, giving each unit’s registration number, status, address, booking service and listing URL or number, without the applicant’s name³⁷.

DBEDT publishes Hawaii’s vacation rental report every month, with a PDF and an Excel data file². Its August 2026 report covered 36,079 units with 62,270 bedrooms. It excludes units counted in the state’s hotel and timeshare reports and does not determine or differentiate between permitted and unpermitted units, since legality is decided by each county¹.

What the platforms publish

Airbnb’s newsroom, last updated May 2026, lists “5.5M+ hosts on Airbnb”, “9M+ active listings worldwide” and listings in “220+ countries and regions”³⁸. These are global totals. Platform data also reach the public through official channels such as the Eurostat series, whose headline totals combine Airbnb, Booking and Expedia³³. Outside firms test their estimates against what the platforms report: AirDNA says it benchmarks its data against Airbnb’s financial filings¹⁴.

Why two providers report different numbers for the same market

The providers themselves name the main reasons their figures differ.

• Different listings. PriceLabs says its reports are localised around an address, while other services report by city, zip code or other boundaries, so the listings tracked can differ⁶.
• Different platforms. PriceLabs’ reports use Airbnb or Vrbo data, and it notes that others use other booking sites or add hotels⁶. AirDNA says providers relying only on Airbnb data or on property managers miss “up to 50% of listings in most markets”¹⁴. KeyData says most market datasets estimate occupancy by analysing listing calendars¹⁵.
• Different treatment of closed nights. PriceLabs says every provider that uses scraped data must identify and remove blocked dates, and that each one’s techniques are unique and can produce differences⁶. Inside Airbnb’s availability figures count a booked night and a blocked night alike, because the Airbnb calendar does not tell them apart³.
• Different definitions. AirDNA divides reserved nights by active nights, counting only nights that were not blocked, in listings with a reservation in the past 28 days¹². KeyData’s Paid Occupancy % divides guest nights by all nights, and its adjusted measure by the nights open to guests¹⁸¹⁹. Inside Airbnb estimates occupancy from reviews and caps it at 70%³. Within one provider, AirDNA notes that its total active listings count and its 12-month listings count will in most instances differ¹³.
• Different prices and revisions. PriceLabs estimates revenue from the last price seen, without cleaning or extra-guest fees⁵. Lighthouse reports a Total Rate that includes such fees and reprocessed its history back to January 1, 2019; it also restated history after its July 2026 de-duplication upgrade¹. AirDNA added Booking.com data to its platform in June 2026¹¹.

Producer	Method	Coverage as stated	Public outputs	Source
AirDNA	Daily scraping of Airbnb, Vrbo and Booking.com listings, checked against reservation data shared by managers and hosts	“over 16M listings every day”	Free tools listed on its site, including a revenue calculator	AirDNA methodology page and help centre
KeyData	Reservation records from property management systems, plus scraped Airbnb and Vrbo data and hotel data	“80+ PMS integrations”; “17,000+ property managers”; “10 million+ properties tracked”	Quarterly KeyData Index and monthly Short-Term Rental Pulse listed on its reports page	KeyData homepage, how-it-works page, help centre and reports page
Lighthouse (formerly OTA Insight; bought Transparent in 2022)	Listings on Airbnb, Booking.com, Vrbo and TripAdvisor, enriched with reservation data from software providers, travel agencies and property managers, as DBEDT describes it	“16.4 million” hotels and short-term rentals profiled daily	Sold through APIs, feeds, dashboards and reports; published by clients such as DBEDT	Lighthouse data-solutions page; DBEDT August 2026 report
PriceLabs Market Dashboards	Scraped Airbnb and Vrbo calendars and prices, with block-removal logic	1000, 5000 or 10,000 listings around an address, by plan	Dashboards sold by plan	PriceLabs help centre
Beyond	Direct integrations with booking channels, consumer search demand and property management systems	n/a	Insights for operators who connect a portfolio	Beyond Insights page
Wheelhouse	n/a	“1,500+ markets, 100+ countries”	Free trial offered	Wheelhouse market insights page
AllTheRooms (bought by Deckard Technologies)	n/a	n/a	n/a	PrivSource deal record
Airbtics	Publicly available Airbnb listing data, generally scanned weekly	“30,000+” markets; “15M” Airbnb listings tracked; data since January 2021	Free account for a first market	Airbtics homepage
Inside Airbnb	Snapshots of public Airbnb listings; occupancy estimated from reviews	Snapshots by city or region	Last year’s quarterly data, free to download under CC BY 4.0	Inside Airbnb data pages
STR (CoStar)	Performance data supplied directly by hotels	“sample of 94,000 hotels and 12 million rooms”	n/a	CoStar STR Benchmark page
Eurostat	Aggregated data shared by Airbnb, Booking and Expedia; Tripadvisor until its exit	EU countries and NUTS 2 regions	Datasets tour_ce_omr and tour_ce_omn12; news releases	Eurostat news releases
Hawaii DBEDT	Lighthouse data for listings on Airbnb, Booking.com, Vrbo and TripAdvisor, excluding hotel and timeshare units	36,079 units, 62,270 bedrooms (August 2026)	Monthly report (PDF) and data file (Excel)	DBEDT August 2026 report and reports page
New York City OSE	Register of hosts under Local Law 18, checked by booking services	Registered units, with booking service and listing URL or number	Registration dataset, latest dated January 7, 2026	OSE registration pages
Airbnb	The platform’s own counts	“9M+ active listings worldwide”; “5.5M+ hosts on Airbnb” (May 2026)	Newsroom fast facts	Airbnb newsroom

Coverage and outputs as each producer describes them on the pages cited, read on 8 or 9 October 2026. n/a means the pages read for this guide do not state it. Units differ (listings, properties, units, rooms), so the counts are not comparable with one another.

Source: Hawaii Department of Business, Economic Development & Tourism (DBEDT), Inside Airbnb, PriceLabs, AirDNA, KeyData, Hospitality Net (OTA Insight release), Lighthouse, Beyond, Wheelhouse, Airbtics, PrivSource, CoStar Group, Eurostat, NYC Mayor's Office of Special Enforcement, Airbnb^{9111517212412264252628273303233342363738}

Download the tableCSV file, 14 rows, with the source links and the date checked

This table was compiled by Booked News and may be reused with credit and a link (CC BY 4.0). The figures themselves belong to the bodies that published them, named in the sources.

How Booked News treats these sources

Every Booked News story that uses one of these figures names the producer and its method: scraped listings, booking records from property managers’ software, a platform’s own disclosure, or an official statistic or register. Each figure is reported as the producer published it, with the period and comparison basis it used. Booked News does not compute new figures from these sources, does not combine one provider’s numbers with another’s, and gives both figures when two producers disagree. A company’s description of its own coverage or accuracy is reported as that company’s claim. The rules are set out on the standards page.

Common questions

Sources

1. Hawaii Department of Business, Economic Development & Tourism (DBEDT), August 2026 Hawai‘i Vacation Rental Performance Report. Primary source, read 2026-10-08.
2. Hawaii Department of Business, Economic Development & Tourism (DBEDT), Vacation Rental Performance. Primary source, read 2026-10-08.
3. Inside Airbnb, Data Assumptions. Primary source, read 2026-10-09.
4. PriceLabs, Market Dashboard Methodology: Data Sources and Processing. Primary source, read 2026-10-08.
5. PriceLabs, Market Dashboards: Methodology & Terminology. Primary source, read 2026-10-08.
6. PriceLabs, PriceLabs Market Intel Dashboards Frequently Asked Questions. Primary source, read 2026-10-08.
7. KeyData, What is the difference between direct and scraped data?. Primary source, read 2026-10-08.
8. European Commission, Commission reaches agreement with collaborative economy platforms to publish key data on tourism accommodation (2020-03-05). Primary source, read 2026-10-09.
9. AirDNA, The AirDNA Data Model 5.0. Primary source, read 2026-10-08.
10. AirDNA, About AirDNA. Primary source, read 2026-10-08.
11. AirDNA, Where is our data sourced from? (2026-06-19). Primary source, read 2026-10-09.
12. AirDNA, How does AirDNA calculate occupancy rate? (2026-08-20). Primary source, read 2026-10-08.
13. AirDNA, Total active listings (2026-06-12). Primary source, read 2026-10-08.
14. AirDNA, AirDNA Data Accuracy. Primary source, read 2026-10-08.
15. KeyData, KeyData homepage. Primary source, read 2026-10-08.
16. KeyData, About KeyData. Primary source, read 2026-10-08.
17. KeyData, How it works. Primary source, read 2026-10-08.
18. KeyData, Paid Occupancy %. Primary source, read 2026-10-08.
19. KeyData, Adjusted Paid Occupancy %. Primary source, read 2026-10-08.
20. KeyData, Who are the other PM's in the Benchmarking Report?. Primary source, read 2026-10-08.
21. KeyData, KeyData industry reports. Primary source, read 2026-10-08.
22. Hospitality Net (OTA Insight release), OTA Insight acquires Transparent to form the world's first cloud-based commercial platform to serve both hotel and short-term rental industries (2022-03-10). Secondary source, read 2026-10-09.
23. Hospitality Net (Lighthouse release), OTA Insight Rebrands as Lighthouse to Illuminate New Capabilities and Launch of a Unified Commercial Platform (2023-11-09). Secondary source, read 2026-10-08.
24. Lighthouse, Data Solutions. Primary source, read 2026-10-09.
25. Beyond, Market Intelligence That Keeps You Ahead. Primary source, read 2026-10-08.
26. Wheelhouse, Market insights. Primary source, read 2026-10-09.
27. Airbtics, Airbtics homepage. Primary source, read 2026-10-09.
28. PrivSource, Deckard Technologies Acquires AllTheRooms to Expand STR Data & Compliance Capabilities. Secondary source, read 2026-10-09.
29. Inside Airbnb, About Inside Airbnb. Primary source, read 2026-10-09.
30. Inside Airbnb, Get the Data. Primary source, read 2026-10-09.
31. arXiv (Abdulkareem Alsudais), Incorrect Data in the Widely Used Inside Airbnb Dataset (2020-12-01). Primary source, read 2026-10-09.
32. CoStar Group, CoStar with STR Benchmark. Primary source, read 2026-10-09.
33. Eurostat, Tourism nights booked via platforms hit nearly 1 billion (2026-04-01). Primary source, read 2026-10-09.
34. Eurostat, Tourism bookings via platforms grew in summer of 2025 (2026-01-08). Primary source, read 2026-10-09.
35. European Parliament Legislative Observatory, Data collection and sharing relating to short-term accommodation rental services, 2022/0358(COD), summary (2024-04-29). Primary source, read 2026-10-09.
36. NYC Mayor's Office of Special Enforcement, Registration Law. Primary source, read 2026-10-09.
37. NYC Mayor's Office of Special Enforcement, Registration and Listing Data. Primary source, read 2026-10-09.
38. Airbnb, About Us. Primary source, read 2026-10-09.