Abstract or concrete language style? How chatbots of online travel agencies should apologise to customers

Service failure is a common issue in the operations of online travel agencies (OTAs) because of the wide range of heterogeneous services being provided. Many OTAs have now started using chatbots for complaint management. This research investigates the most effective language style (abstract or concrete) for OTAs chatbots to apologize to the customer to gain forgiveness for service failure. Results of three experimental studies with 348, 452, and 470 sample sizes showed that the overall concrete language style of the OTAs chatbot apology was more effective in achieving customer forgiveness. Further, in case of minor (severe) service failure, an abstract (concrete) language style is more effective. Moreover, results showed that the customer's preferred chatbot apology language style was unaffected by the OTAs service type (hedonic vs utilitarian). Finally, results show that perceived firm sincerity and empathy mediate the relationship between OTAs chatbot apology language styles and customer forgiveness.