When thanks have nowhere to go: emotional indebtedness in the age of service robots

Introduction When hotel service robots deliver meaningful service, customers may experience emotional indebtedness—a moral-emotional state arising from perceived obligation without a clear target for reciprocity. This study examines how social (social presence, enjoyment) and nonsocial (novelty, performance) attributes of service robots influence emotional indebtedness and subsequent brand loyalty. Methods Survey data were collected from 462 hotel guests in China with direct service robot experience. Structural equation modeling (SEM) with maximum likelihood estimation was employed, incorporating second-order latent constructs and multi-group analysis by anthropomorphism level. Results Social attributes significantly increased emotional indebtedness, whereas nonsocial attributes either reduced or failed to elicit such responses. Emotional indebtedness positively predicted brand loyalty. Multi-group analyses revealed that the effect of social attributes on emotional indebtedness was invariant across anthropomorphism levels, whereas anthropomorphism moderated the relationship between nonsocial attributes and emotional indebtedness and the conversion of indebtedness into brand loyalty. Discussion These findings extend social exchange theory to robot-mediated hospitality services, demonstrating that customers’ moral-emotional responses depend more on how robot interactions are socially interpreted than on functional performance alone. Anthropomorphism functions as a selective boundary condition rather than a uniform amplifier of emotional engagement.