Machine learning insights into recommendation intention: evidence from the fitness industry
Recommendation intention, or the Net Promoter Score (NPS, in the terminology of fitness centers), is a tool that condenses into a single indicator the willingness of customers to recommend a service. It is important because it has become a widely used proxy for loyalty, brand advocacy, and growth, a...
| Autores: | , , , |
|---|---|
| Formato: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2026 |
| País: | España |
| Recursos: | Universidad de Sevilla (US) |
| Repositorio: | idUS. Depósito de Investigación de la Universidad de Sevilla |
| OAI Identifier: | oai:dnet:idus________::5725041950167ec8377b25aec7314a11 |
| Acesso em linha: | https://hdl.handle.net/11441/185846 https://doi.org/10.1186/s40537-026-01398-5 |
| Access Level: | acceso abierto |
| Palavra-chave: | Machine learning Sports management Intention to recommend Fitness centres Net promotion score |
| Resumo: | Recommendation intention, or the Net Promoter Score (NPS, in the terminology of fitness centers), is a tool that condenses into a single indicator the willingness of customers to recommend a service. It is important because it has become a widely used proxy for loyalty, brand advocacy, and growth, and fitness managers are very familiar with its interpretation. Therefore, it is necessary to examine its validity and determinants with greater academic rigor, particularly in experiential services such as fitness, where competition and customer churn are critical. This study applies explainable machine learning techniques to predict recommendation intention using a sample of 15,822 users from 9 fitness center chains in Spain. Five widely established algorithms were employed (Decision Tree, Random Forest, Logistic Regression, Gaussian Naïve Bayes, and Extreme Gradient Boosting), evaluated with classification metrics such as accuracy, sensitivity, specificity, and AUC. The results show that satisfaction, emotions, and renewal intention are the most relevant predictors, achieving accuracy levels above 80%. Furthermore, the application of interpretability techniques offers a clear ranking of key variables, with practical implications for loyalty management and the design of customer-oriented marketing strategies. This work contributes to the literature by combining large-scale empirical evidence from the fitness sector with advanced data analytics and opens the way for replication in other experience-intensive service industries. |
|---|