Authentic or Counterfeit Service? A Framework on the Complexity of AI Enabled Service Encounters

Advancements in technology continually transform service encounters. As companies move from traditional interactive voice response (IVR) systems to artificial intelligence (AI) powered chatbots and virtual assistants, service encounters are reshaped, as the technology powers multiple customer service channels, and employees, or customers, are replaced.

Examples of AI enabled, non-face-to-face, service encounters are abundant across multiple industries, with banks, hotels, and retailers employing the technology. Customers checking into a hotel might interact with a chatbot via a mobile phone text, or AI might read and respond to a retailer’s customer complaint emails. Innovations in AI (i.e., Google Duplex) provide a service whereby a customer’s “virtual assistant” may call a restaurant to make a reservation.

The increasing infusion of AI in service encounters suggests it must be beneficial for customers and employees. However, a closer look reveals a more complicated picture in which the benefits and the drawbacks of AI coexist. Our research focuses on this complexity. We propose a framework that examines the consequences on customers, on employees, and on the service company of infusing the service encounter with AI agents. We examine encounters in which the customer or the employee are aware vs. unaware that they are interacting with an A.I. employee or customer.

Some preliminary findings of our research show that when customers assess the encounter with a DTMF system, voice recognition, AI or human, they evaluate the efficiency and the quality of the AI interaction as positively as the human interaction; however, the pleasantness of the AI interaction and the perception of being valued by the company is rated well below the human interaction. These findings suggest that consumers have ambivalent feelings towards the service experience when AI is present. We propose that a similar phenomenon exists for service employees. Although popular press argues that by letting AI performing routine tasks, employees can handle complex activities and increase engagement in their daily work, there are likely drawbacks associated with AI presence. Employees might perceive that the AI is doing the core of their job, and that the AI receives credit for the work done with the customer.

The complexity of AI enabled service encounters is bound to increase as AI technologies continue to evolve, and become progressively more humanlike. Customer experiences in which customers cannot tell if they are communicating with a human or a computer are increasing. According to 2016 HubSpot Global AI Survey 63% of customers use AI without being aware of doing it. Efforts to design AI in the form of disembodied agents that are difficult or impossible to delineate from human, and potential lack of awareness regarding the presence of AI in dyadic service exchanges, may be problematic. We advance that humanlike AI creates a counterfeit service encounter, if the consumer or employee, is unaware they are interacting with a non-human partner. Counterfeit encounters might increase and negatively affects employees and customer experience, as well as the level of trust towards the service company.