What’s yours is mine: exploring customer voice on Airbnb using text-mining approaches

Abstract

This paper aims to investigate customers’ experiences with Airbnb by text-mining customer reviews posted on the platform and comparing the extracted topics from online reviews between Airbnb and the traditional hotel industry using topic modeling.,This research uses text-mining approaches, including content analysis and topic modeling (latent Dirichlet allocation method), to examine 1,026,988 Airbnb guest reviews of 50,933 listings in seven cities in the USA.,The content analysis shows that negative reviews are more authentic and credible than positive reviews on Airbnb and that the occurrence of social words is positively related to positive emotion in reviews, but negatively related to negative emotion in reviews. A comparison of reviews on Airbnb and hotel reviews shows unique topics on Airbnb, namely, “late check-in”, “patio and deck view”, “food in kitchen”, “help from host”, “door lock/key”, “sleep/bed condition” and “host response”.,The topic modeling result suggests that Airbnb guests want to get to know and connect with the local community; thus, help from hosts on ways they can authentically experience the local community would be beneficial. In addition, the results suggest that customers emphasize their interaction with hosts; thus, to improve customer satisfaction, Airbnb hosts should interact with guests and respond to guests’ inquiries quickly.,Hotel managers should design marketing programs that fulfill customers’ desire for authentic and local experiences. The results also suggest that peer-to-peer accommodation platforms should improve online review systems to facilitate authentic reviews and help guests have a smooth check-in process.,This study is one of the first to examine consumer reviews in detail in the sharing economy and compare topics from consumer reviews between Airbnb and hotels.