Analysis of participant satisfaction in the Madrid marathon: a case study using a data-driven and an unsupervised machine learning approach

Purpose: This study aimed to present a data-driven framework for analyzing participant satisfaction in marathon events. The specific objectives were to establish a procedure for cleaning, preprocessing, and analyzing survey data; to assess satisfaction with event components and intentions to re-participate; and to identify participant segments based on experiences and motivation.Design: A quantitative approach was applied using 3,669 online survey responses from Madrid Marathon participants. The analysis included descriptive statistics, bivariate analyses to explore relationships between satisfaction dimensions and behavioral intentions, and multivariate techniques – factor analysis and K-means clustering – to identify runner segments.Findings: Participants reported high satisfaction levels, with a moderate correlation (r = 0.64) between overall event satisfaction and the race experience. Two main runner profiles emerged: “performance-oriented” and “entertainment-oriented,” differing significantly across 32 satisfaction variables.Practical Implications: Results provide actionable insights for event organizers, highlighting the need for tailored engagement strategies that reflect runners’ distinct motivations and experiences.Research Contribution: The study advances sport event management by integrating unsupervised machine learning with survey-based satisfaction analysis, offering a replicable methodological approach for participant segmentation.Originality: By combining data-driven analysis and clustering techniques, the study offers a novel, scalable model for understanding participant satisfaction in international running events.