Machine learning breakthrough sheds new light on hotel customer satisfaction

A research team has crafted an innovative machine learning model that delves into the intricate dynamics between service attributes and customer satisfaction. This groundbreaking study is poised to arm hoteliers with actionable insights, empowering them to refine their services and elevate the guest experience to unprecedented levels.

Customer satisfaction in the service sector, particularly within hospitality, has long been a focal point for both academic research and practical application. Traditional analyses, such as the Kano model and importance-performance analysis (IPA), have offered valuable frameworks but often fall short in capturing the intricate and non-linear nature of the attribute performance-customer satisfaction (AP-CS) relationship.

A study (DOI: 10.1016/j.dsm.2024.01.003) published in Data Science and Management on January 11, 2024, employs a novel machine learning approach to reveal the complex relationship between hotel service attributes and customer satisfaction, providing actionable insights to improve the guest experience.

This study advances beyond conventional analysis by introducing a machine learning-based framework that unravels the intricate interplay between hotel service attributes and customer satisfaction. Through the analysis of 29,724 TripAdvisor reviews of New York City hotels, the research team has formulated an interpretable machine learning-based dynamic asymmetric analysis (IML-DAA) model. This pioneering method integrates extreme gradient boosting (XGBoost) with SHapley Additive exPlanations (SHAP), achieving unparalleled accuracy in predicting customer satisfaction and elucidating the impact of specific service attributes on overall guest contentment. Distinct from prior models, IML-DAA skillfully captures non-linear relationships and the changing influence of these attributes over time, providing a detailed insight into customer preferences. The model's capability to adapt dynamically to shifting customer expectations offers actionable insights, empowering hotel managers to strategically refine service attributes, prioritize enhancements, and navigate market fluctuations.

According to the study's lead researcher, Prof. Shaolong Sun, "Our approach leverages the power of interpretable machine learning to not only predict customer satisfaction more accurately but also to provide actionable insights into how various service attributes contribute to overall satisfaction."

The methodology empowers stakeholders to make informed decisions on service improvement, resource allocation, and strategic planning, adapting proactively to changes in consumer expectations. This study represents a pivotal advancement in harnessing machine learning to refine customer satisfaction strategies in the hospitality sector.

Reference

Title of the original paper	The relationship between attribute performance and customer satisfaction: an interpretable machine learning approach
Journal	Data Science and Management   Data Science and Management (DSM) is a peer-reviewed open access journal for original research articles, review articles and technical reports related to all aspects of data science and its application in the field of business, economics, finance, operations, engineering, healthcare, transportation, agriculture, energy, environment, sports, and social management. DSM was launched in 2021, and published quarterly by Xi'an Jiaotong University.
DOI	10.1016/j.dsm.2024.01.003

Image title: Framework of this study design and model development.

Image credit: The authors

Image link:https://ars.els-cdn.com/content/image/1-s2.0-S2666764924000031-gr1_lrg.jpg    

License type: CC BY-NC-ND 4.0

Funding information:

This research was partially supported by the National Key R&D Program of China (Grant No.: 2022YFF0903000) and the National Natural Science Foundation of China (Grant Nos.: 72101197 and 71988101).

Contact the author:

Name: Shaolong Sun 

Email: sunshaolong@xjtu.edu.cn