Preliminary findings from an analysis of Airbnb-sourced data in 13 global markets reveal new insights about the platform’s size and scope.
Editor’s note: Airbnb approved the use and presentation of the data and content in this article. Hotel News Now edited the article for spelling and grammar.
GLOBAL REPORT—The following preliminary findings are based on a presentation at the 2016 Hotel Data Conference, which relied on data provided by Airbnb in seven U.S. and six international markets. A full report with finalized findings will be released within the next few months.
Airbnb’s size compared to hotels
Airbnb has a reported 2.3 million listings worldwide. By raw number of listings, Airbnb is more than twice the size of the combined entity of Marriott International and Starwood Hotels & Resorts Worldwide.
However, Airbnb’s listings contain many units that aren’t always available for booking or otherwise are not comparable to hotels. When those are removed, Airbnb inventory is much smaller (see below).
The difficulty of comparing hotels and Airbnb
Comparing hotels and Airbnb listings will always be a case of “apples–to-oranges” because these two accommodation types are so different. To come up with a more comparable data set, we first removed listings that aren’t actually available for rent (some hosts create a listing, but never actually make it available). We also removed shared rooms because it is unlikely most hotel guests would view such a space as a viable alternative to a hotel room. One open question is whether Airbnb “private room” listings (versus “entire home” listings) should be treated as inventory that is comparable to hotel rooms. With private rooms, guests typically share a bathroom or kitchen with a host. For portions of our analysis we chose to include private rooms because these may possibly compete with the lowest-cost hotels.
We reviewed data provided by Airbnb for seven U.S. markets to determine what percentage of Airbnb supply: 1) is not available for rent, 2) are shared room(s) and/or 3) are private room(s). We then applied those percentages to the total number of Airbnb listings globally and estimate that only 981,000 (43% of total Airbnb listings) are likely competitive with hotels.
But even this number may be inflated. Of those 981,000 listings, a (likely small) percentage is comprised of eclectic accommodation types such as treehouses, camper vans, yurts, castles and so forth (which we could not identify from the data provided). Additionally, some Airbnb listings can accommodate families and larger groups. Some listings that are available for booking are only available a small number of nights per year. Finally, because Airbnb listings often include interaction with a host and other unique elements, even for the remaining listings the overall “experience” may not be comparable to hotel stays. So Airbnb listings are often not one-to-one comparable to hotel rooms.
Assumptions required when using scraped data
Most reports released on Airbnb to date rely on scraped data. Pulled directly from advertised listings on the company’s website, scraped data can prove a useful indicator—but one subject to numerous assumptions. The chart below illustrates several of these.
STR has used scraped data in the past and believes it can be a useful, directional indicator of Airbnb’s growth and prices. But just as when benchmarking hotel performance, source data is far more reliable.
Hotel and Airbnb occupancy
Airbnb’s occupancy shows a similar seasonal pattern to hotels. Airbnb sees more sustained occupancy in the summer, a logical finding, since Airbnb is leisure-focused. Hotels have much higher occupancy year-round. This is partially due to hotels’ steady and diverse base of group and contract travel.
The markets with the highest Airbnb occupancy are also the markets with the highest hotel occupancy. One possible interpretation: Hotels in Los Angeles, San Francisco and Tokyo are already full, so Airbnb is accommodating excess demand.
Hotel and Airbnb rates
Hotels saw higher rates than Airbnb listings in all seven U.S. markets analyzed, with the exception of New Orleans.1 On average, hotels charged $43 more than Airbnb. Again, the “apples-to-oranges” comparison of Airbnb listings to hotel rooms makes such a comparison challenging.
Hotel rate growth was also higher in all markets except New Orleans. Five of the seven markets actually saw negative ADR performance for Airbnb. This could suggest that Airbnb is not immune to its own supply growth (or that Airbnb hosts are less versed in revenue management).
Hotel compression nights
Another way to assess the possible impact of Airbnb listings on hotel rate is by analyzing compression nights—high-demand periods in which occupancy is 95% or higher. These nights are important for hotels seeking to maximize revenue because hoteliers are able to charge significantly more when hotels are full. Many hoteliers fear that Airbnb’s new supply might reduce the number of compression nights. Our data shows that compression nights for the seven U.S. markets analyzed haven’t been significantly impacted.
Compression nights are down very slightly from 2014 and 2015—but they are up from 2013. The year of 2013 is arguably a better year to compare to the current year because 2013 was before Airbnb had a significant presence and before hotel supply ramped up. While Airbnb may potentially be impacting compression nights in some markets, it’s also important to consider that these markets are seeing significant hotel supply growth as well.
On compression nights this year in the seven U.S. markets analyzed, hoteliers have charged 35% more than noncompression nights, which is the highest rate premium ever achieved in the recorded history of STR data. It doesn’t appear that additional Airbnb or hotel supply is having an impact on rate premiums at this time.
STR will be releasing a more detailed report later this year. If you are an STR client or Hotel News Now subscriber, you will get an email when it’s released.
1One possible explanation for this is that we have not yet controlled for Airbnb listing size. If New Orleans Airbnb listings skewed larger (e.g. had more multi-room homes), this would explain the higher rates.