13-05: Miaolei (Liam) Jia
Chair: Paula Dootson
Robot or Human? The Effect of Robot-versus-Human Caused Service Failure on Firm Evaluation
The use of robots in frontline services has been rapidly changing service industries (Wirtz et al. 2018). For instance, hotel chains such as M Hotel Singapore and Hilton Hotel have introduced service robots to provide room and concierge services (Scott and Gugliocciello 2016; Street 2017). The benefits of using robots are widely acknowledged—consistent quality and performance, virtually endless memory and access, and high economies of scale and scope (Wirtz et al. 2018). However, could there be potential drawbacks in using service robots?
In this research, we study and found that when a service failure is caused by a service robot (vs. a human employee), consumers will evaluate the firm more negatively, experience more anger at the firm, blame the firm more, and are more likely to write a bad review of the firm. We propose that this effect could be explained by accountability. When a service failure is caused by a human employee, consumers tend to hold the employee accountable; however, when a service failure is caused by a robot, consumers cannot hold the robot accountable, and instead hold the robot’s owner (i.e., the firm) accountable. We tested our proposition in three studies.
Study 1 used a 2 (robot vs. human) × 2 (light vs. severe service failure) between-subjects design. Participants imagined that they were dining at a restaurant. In the human (robot) condition, a waiter (robot) served the food, and spilled a bowl of hot soup. In the light-failure (heavy-failure) condition, the participants were told that they were not scalded by the hot soup but their outfit was lightly stained (severely scalded by the hot soup and their outfit was heavily stained). Participants were asked to rate “How angry are you at the restaurant right now?”, “To what extent do you blame the restaurant?”, and “How likely are you to write a bad review of the restaurant online?” These three items (α = .844) were averaged to form an evaluation index. We found that the participants rated the restaurant more negatively when the failure was caused by a robot versus a human (Mrobot = 4.80, Mhuman = 3.30; p < .001).
Study 2 conceptually replicated the effect in a hospital setting. We found that the participants evaluated the hospital more negatively when the service failure was caused by a robot versus a human (Mrobot = 4.87, Mhuman = 3.96; p < .001).
Study 3 tested the accountability explanation in a hotel context. We found that the participants evaluated the hotel more negatively when the failure was caused by a robot versus a human (Mrobot= 4.35, Mhuman= 3.65; p = .003), and that this effect was mediated by accountability.