I’ve seen you made a mistake: exploring observers’ reactions to robotic service failures and recovery

With the increasing application of frontline robots (FLRs) in public service environments, their service failures are inevitably witnessed by observers. Although valuable explorations have been carried out in robotic failure and recovery research from the perspective of mistreated customers, little is known about how robotic failures and recovery influence observers’ reactions. Based on attribution theory, this research examined the ‘ripple effect’ of FLR failures on observers’ switching intention through three scenario-based experiments. The findings reveal that seeing FLR (vs. frontline employee) failures enhances observers’ perceived stability, thus increasing their switching intention. Providing explanation-focused (vs. apology-focused) recovery, especially outcome-focused explanations, can mitigate the negative effect of FLR failures on observers’ switching intention. These findings contribute to robotic failure and recovery, AI aversion and observer research, and provide practical guidance for better handling FLR failures and increasing observers’ continued usage intention of failed FLRs.