Improving peer-to-peer accommodation service based on text analytics

Abstract

Purpose

This paper aims to identify key service attributes in peer-to-peer (P2P) accommodation from online reviews and formulate service improvement strategies based on the unsatisfactory service encounters mined from the reviews.

Design/methodology/approach

The methodology involves topic modelling using latent Dirichlet allocation, sentiment analysis and process analysis based on process chain network (PCN).

Findings

The text analytics results showed that negative P2P accommodation experiences are caused by the lack of hot water for shower, poor sleep quality and unpleasant check-in.

Research limitations/implications

The PCN analysis shows that the surrogate interactions of the P2P accommodation platform with both the guest and the host impact consumer experiences. This highlights that the key to managing consumer experiences lies in the non-human resources such as information, rather than direct interactions between process entities.

Practical implications

The information on the P2P accommodation platform should be in a more interactive format such as video and 360 degrees camera. Hosts should ensure a good condition of the physical products such as water heaters and beds before guests' arrival. Professional videography and handyperson services should be provided by the platform to help hosts deliver a preferred consumer experience. Flexible and strict check-in polices should also be introduced to smoothen the check-in process.

Originality/value

This study is built on multi-attribute utility theory. It is also one of the first to study P2P accommodation services from an operations management perspective. It demonstrates how text analytics serves as an additional supplement for service improvement.

Citation

Lee, C.K.H. and Tse, Y.K. (2021), "Improving peer-to-peer accommodation service based on text analytics", Industrial Management & Data Systems, Vol. 121 No. 2, pp. 209-227. https://doi.org/10.1108/IMDS-02-2020-0105