Applying K-Means Clustering to Analyze the Contextual Impact of Failure Severity on Dissatisfied E-Consumers

Résumé

Abstract

This research offers insights into the dynamics of online service failures and recovery. Failures occur frequently, yet consumers generally do not anticipate them. Their baseline expectation is that no failure will occur, making failures particularly disappointing and, at times, shocking when a purchased product does not meet expectations. While failures create a sense of imbalance, recovery services aim to restore fairness by compensating dissatisfied customers.

However, companies must proceed with caution, as unhappy customers closely observe how their concerns are addressed. Research has shown that the severity of a failure negatively impacts customer satisfaction during the recovery process. This study further explores how the context of a failure influences this relationship, employing the k-means clustering method. Data were collected from 355 online consumers who experienced dissatisfaction following a problematic purchase, allowing the identification of distinct consumer profiles based on their reactions to failure and recovery strategies.

Keywords

Failure Severity; Satisfaction Recovery; Negative Effect; Contextual Effect; K-Means Clustering