The role of empathy in providers’ online customer complaints management

Customers share negative service experiences to complain and warn prospective customers. Our research uses semi-supervised machine learning to fine-tune a BERT (Bidirectional Encoder Representations from Transformers) model, which examines the effects of affective empathy (acknowledging and responding to the complainer’s emotions) versus cognitive empathy (demonstrating perspective-taking to understand the complainer’s situation) in provider responses to complaints. Specifically, we examine the effects of these responses on prospective customers’ reactions. Our two field studies, which cover 12,638 negative reviews on TripAdvisor and 36,478 complaints on Facebook, reveal that cognitive empathy is generally more effective than affective empathy in increasing prospective customers’ likes and purchases, particularly when a complaint is formulated concretely. However, affective empathy is more suitable when a complaint is intensely affective. Four experiments confirm that cognitive empathy (vs. affective empathy) leads prospective customers to view providers as more competent, ultimately enhancing their purchase intentions. These insights offer guidelines to providers on how to respond to online customer complaints.