Predicting overall customer satisfaction: Big data evidence from hotel online textual reviews

• Published in: International journal of hospitality management 2019-01, Vol.76, p.111-121
• Main Authors: Zhao, Yabing, Xu, Xun, Wang, Mingshu
• Format: Article
• Language: English
• Subjects: Big data; Hotel industry; Online textual reviews; Overall customer satisfaction; Technical attributes
• ISSN: 0278-4319; 1873-4693
• DOI: 10.1016/j.ijhm.2018.03.017

[Display omitted] •Customer satisfaction is predicted by the linguistic characteristics of reviews.•Review diversity and polarity have a positive effect on customer ratings.•Review subjectivity, readability, and length have a negative effect on ratings.•Customer review involvement has a positive effect on customer ratings.•The importance of technical attributes on affecting customer ratings is compared. Customer online reviews of hotels have significant business value in the e-commerce and big data era. Online textual reviews have an open-structured form, and the technical side, namely the linguistic attributes of online textual reviews, is still largely under-explored. Using a sample of 127,629 reviews from tripadvisor.com, this study predicts overall customer satisfaction using the technical attributes of online textual reviews and customers’ involvement in the review community. We find that a higher level of subjectivity and readability and a longer length of textual review lead to lower overall customer satisfaction, and a higher level of diversity and sentiment polarity of textual review leads to higher overall customer satisfaction. We also find that customers’ review involvement positively influences their overall satisfaction. We provide implications for hoteliers to better understand customer online review behavior and implement efficient online review management actions to use electronic word of mouth and enhance hotels’ performance.