Private Rooms in Peer-to-Peer Accommodation: Employing Machine Learning to Determine Revenue Drivers
Private Rooms in Peer-to-Peer Accommodation: Employing Machine Learning to Determine Revenue Drivers

Summary/Abstract: Purpose: The goal of this study was to identify the performance determinants of private rooms offeredvia peer-to-peer accommodation (P2PA) platforms.Design/methodology/approach: Based on property level data of Airbnb, the applications of machinelearning clustering and artificial neural networks methods were proposed to identify the performancedeterminants of private room performance defined as yearly revenue.Findings: The results indicate that the most important revenue determinants were maximum number ofguests, number of photos, number of bathrooms, location, and cancellation policy, as well as high availa-bility of the property throughout the year. The highest revenues were achieved by 2–3 person rooms withat least two bathrooms, 4-person rooms, or rooms located around the Central Railway Station. However,properties with at least 2 bathrooms tend to deliver high revenue, regardless of location. Furthermore,a sufficient number of photos and a flexible cancellation policy could offset location disadvantages anddeliver higher revenues in distant vs. central locations.Research limitations/implications: The limitations of this study are that it covered only one destinationduring a 12-month period, which encompasses yearly seasonality, but might include some exceptionalevents. Replication for other destinations (urban or non-urban) or timeframes would be valuable futureresearch recommendations.Originality/value: This study adds threefold to the P2PA performance determinants stream. First, it offersan application of machine learning methods to identify property and host features and behaviours thatcontribute to high performance. Second, as opposite to the majority of available research, it definesperformance as revenue rather than price. Third, building on the hedonic price theory, it sheds light onhow to manage a complex offer for private rooms in P2PA, in contrast to entire homes/apartments thatresearchers have focused on to date.

• Details
• Contents