A statistical framework for Airbnb hosts and Superhosts

Contu, Giulia

2022-01-01

---

Abstract

We propose a statistical framework in order to investigate the Airbnb hosts activities. We aim to propose an extended model able rstly to comprehend which variables can impact on the hosts' activity; secondly to identify a guide that can support the hosts in the constant eort to reach the best performances and to become a Superhost. The framework use two dierent models, the logistic model and the bivariate odds ratio model. Three groups of variables are taken into account. They are the attributes that Airbnb uses to assign the Superhost badge, the managerial aspects and the characteristics of the accommodations. The analysis is focused on the hosts operating in the Italian most visited cities. Our ndings show the capacity of the framework to identify the variables, as for instance the number of reviews, the services, and the typology of the rented accommodation, that aect the hosts' performance. The results show how the framework can be used as a managerial support for the hosts.