Airdna Data Methodology

Airdna analytics and reports are based on Airbnb data gathered from information publicly available on the Airbnb website. Our database currently tracks the performance of 2,000,000 Airbnb listings around the globe each day and automatically generates over 2,000 free web pages and 5,000 premium market intelligence reports on a monthly basis. Airdna is the only trusted source for short-term rental data that provides occupancy rates and revenue data.

How can Airdna determine what is a booking versus a date blocked from booking in the Airbnb calendar?

Airdna has developed advanced artificial intelligence and machine learning technology that allows for accurate identification of blocks of unavailable dates observed on Airbnb’s platform as either booked by a customer or blocked by the host. This ability to discern between booked and blocked days is core to any analysis of Airbnb data.

Airbnb originally did not obscure booked vs blocked information and only started that practice in Q4 2015. Airdna’s ability to develop such a model is possible because of the extensive historical data set which captured actual booked and blocked data for 18 months prior to Airbnb’s implementation of the practice of obscuring data types, as well as institutional knowledge on Airbnb host behavior and smart application of modern artificial intelligence technology. Airdna utilizes statistical pattern-recognition techniques, which define a mathematical relationship between what is known about a property and actual classification of genuine reservations or dates blocked by the host. These are similar in concept to the algorithms that enable Amazon to recommend new products that you might be interested in, Netflix to recommend new movies and OkCupid to recommend potential partners.

The accuracy of Airdna’s prediction model is tested by setting aside a portion of Airdna’s known booked/blocked historical data, hiding it from the model training data and then asking the system to classify the blocks of unavailable dates once it's been trained. This output is then compared against the actual known booked/blocked status of each grouping of days to assess the degree of predictive accuracy.

True to its machine learning classification, Airdna’s artificial intelligence model continues to learn and improve as time goes on. This is important as Airbnb booking trends are likely to continuously evolve over time as evidenced by past behavior. Airdna’s AI continues to observe behavior, extract patterns from new information and historical knowledge and, as a result, predict Airbnb booking information with accuracy.

Why does the free city analysis report fewer total properties than what Airbnb itself reports?

Airdna only reports "active" properties that are actually located in the geographic boundaries of each city. A large percentage of listings on Airbnb are no longer being actively rented, haven't updated their calendar in many months and haven't accepted a reservation for an extended period of time. We remove these listings from our analysis to provide a more accurate picture of the current competing properties in each area. In addition, Airbnb displays many listings that are located outside of the queried location. Airdna’s reports only displays properties actually located within the boundaries of each city, postal code or neighborhood.

How is Airbnb Listing Revenue Calculated?

We are constantly reviewing the calendar information of Airbnb properties to determine when a place was booked and for how long. When a new reservation is recorded, we calculate the advertised daily rate of each of those days directly before the booking occurred and then add in the cleaning fee for each unique reservation. At the end of the month we sum up exactly how many days have been booked and at what rate and add in cleaning fees to calculate monthly revenue. Fee’s for additional guests, or last minute discount are not visible.

How do you determine the exact location of properties?

We create a much more exact property location than is shown on the Airbnb website by taking the center point of the average displayed latitude and longitude coordinates. Airdna has built out one of the most robust GIS databases in the world to report on postal codes, neighborhoods, cities, MSA’s, states and countries.

Can I get the source data for the market intelligence reports?

We do provide property specific information to allow our customers to conduct their our analysis. Contact us for further details.

When will Airdna be adding analytics for my city?

Our team is working around the clock to add new cities around the world. Airdna is able to quickly add analytics on new cities upon request. If you would like data on a city that you do not currently see listed, please contact us to get Airdna enabled for that city.