Inside the call: How customer service agent warmth and competence shape customer reactions

Abstract: Despite the interest and relevance of understanding how firms can address customer problems, limited research exists that investigates adaptive approaches for solving customer problems in real-time voice-to-voice (V2V) interactions. Yet, customer experience is critical in frontline service contexts, and V2V interactions make up a sizable portion of customer-firm contact. Further, technology and artificial intelligence are changing how service delivery and recovery are accomplished. The authors advance the literature on frontline interactions by utilizing newly available machine learning technology to investigate within-call aspects of V2V customer- frontline employee (FLE) exchanges. In drawing an analogy to personal selling, we investigate the role of adaptive FLE problem-solving and relational approaches in customer sentiment (e.g., frustration and satisfaction) and firm outcomes (e.g., interaction duration). The model is tested via data obtained from over 28,000 inbound calls of customers interacting with FLEs to discover the internal processes at play within V2V frontline interactions. Our findings indicate that a one-size-fits-all strategy that either emphasizes relational or task-related efforts is counterproductive and that FLEs need to take call characteristics into consideration when adapting appropriate levels of problem-solving and relational approaches.