Survival determinants and prediction for Airbnb listings

This study assesses these events’ effects on business survival and alleviation strategies in the context of peer-to-peer (P2P) accommodation sharing. We specifically examined the COVID-19 pandemic's impact on Airbnb in Shanghai, China. Results indicated a decline in listings’ average survival probability from 56.84 % to 50.08 % after the outbreak, with this downturn being more severe in 2020 than in 2021. Based on dynamic capabilities theory and Aristotle's rhetorical theory, trust was found to play a significant moderating role: higher response rates and review ratings could attenuate the observed negative effect, whereas larger accommodation capacities may intensify it. We also developed models to predict Airbnb listings’ survival probability and identify factors influencing longevity. The best-performing prediction model (i.e., a random forest model) highlighted the listing's age and COVID-19 incidence as key in predicting a listing's survival. This study enriches theories in tourism and hospitality and points to techniques for fortifying businesses’ resilience.