Companies are using automated technologies customer support to elevate customer experiences. However, the question arises as to whether this strategic move is beneficial for companies. Though several studies have demonstrated the use of anthropomorphic chatbots could positively impact customers, studies still lack comparing this benefit with real-time human interaction. Thus, this study aims to examine user experience between AI and human-based chat support and understand whether AI-based chatbots necessitate replacing human interaction in customer support. The results of the study indicated that perceived benefit and relevance among human-based chats are relatively high, and perceived risk was only limited to delayed response and negative attitude. However, chatbots evidenced their perceived risk was high in anthropomorphic chatbots. Risk factors such as ethical concerns, lack of human touch were commonly seen in anthropomorphic chatbots. On the other hand. their perceived benefit and relevance were seen in chatbots acting as substitute of search queries, web navigation. or product demonstration.
TopIntroduction
In recent years, the growing popularity of live chat has interested marketers to improve their relationships with customers. Scholars and practitioners are becoming more focussed on improving their chat services with automation such as chatbots and AI. Chatbots and AI are believed to carry huge advantages of mimicking human traits, retrieving necessary information in conversations, acting as best personal assistant (Khanna et al., 2015), reducing operation cost (Adamopoulou & Moussiades, 2020) and cater to wider audience (Lu et al., 2022). Thus, this has spurred interest in among marketers to invest in chatbots and AI to elevate customer experiences during chat. A study (Grand View Research, 2022), forecasted that chatbot market size is expected to reach USD 2,485.7 million by 2028.
Despite these unquestionable advantages studies still suggest that automation like AI lacks human touch (Lu et al., 2022; Rapp et al., 2021), fails to meet customer expectations (Adam et al., 2021) and thereby highlighting the dark side of technology (Etkin, 2016; Mick & Fournier, 1998; Wilcox & Stephen, 2013). Thus, a lot of scepticism has arose among scholars and researchers about the efficiencies of chatbots.
With this growing scepticism, there is a need to understand whether chatbots can effectively replace human interaction in customer support. In other words, whether AI has the advantage over real time human communication? Although existing literature (Araujo, 2018; Chen et al., 2017; Rapp et al., 2021; Sheehan et al., 2020) argues that perfectly programmed anthropomorphic chatbots could give positive customer experiences, studies still lack comparing this benefit with real time human interaction, albeit with exceptions (Hill et al., 2015). Thus, this study aims to compare and investigate the user experience between Human and AI in chat support and propose the best communication forms to practitioners.
Consequently, this chapter adopts qualitative semi-structured interviews for its flexibility and carries the advantage to conduct in-depth examination of user experience in chat support. With the focus to investigate user perceptions between AI and human agents, the authors were inspired to use some of the approaches of Technology Acceptance Model (TAM), for its popularity to study factors influencing computer usage (Davis et al., 1989), and technology-based contexts, such as online shopping (Reimers et al., 2016). One of the crucial elements to examine a man’s experiences is through perception. Rookes and Willson (2000) define perception as the process of gaining awareness or understanding of sensory information. Consequently, perceptions build a belief system of an experience. Thus, considering the importance of studying perceptions among users and guided by TAM, this study uses constructs such as perceived benefits, perceived relevance and perceived risk to examine user experiences of chatbots and human agents. The above constructs will help in understanding both positive and negative impacts of chat experience among users.
Thus, guided by TAM while adopting semi-structured interviews, the study conducts 10 interviews among participants aged 22-45 belonging to India, to meet following research objectives: (i) to explore the antecedents of chat based user experience, (ii) to analyse the perceived usefulness, relevance and risk associated with human chats and AI based chats, and (iii) to propose an appropriate communication platform for marketers and researchers.