Customer experience in the age of AI-powered service agents
Niilo Vakula · Aaltodoc (Aalto University) · 2025
This thesis examines how AI-powered service agents, such as chatbots and virtual assistants, influence customer experience in digital services. These systems are becoming common in many different industries, but the research around them is still quite scattered. Because of that, it is not always clear what we actually know about how customers experience these tools. To get a clearer view, two methods were combined: a bibliometric analysis of 66 peer-reviewed Q1 articles and a thematic analysis based on the same material. The bibliometric results showed that the field has expanded rapidly since 2019 and that most studies revolve around three areas: design features that shape how the interaction with the agent feels, the technical performance of the systems, and different trust- or risk-related factors that influence whether people want to use these agents in the first place. The thematic analysis helped bring these threads together. It pointed to three central themes: the theoretical foundations used in the field, the role of design choices such as empathy and human-likeness, and finally the importance of trust factors, privacy concerns, responsiveness, and performance quality. Across the reviewed studies, customer expectations appeared to play a major role in how these systems are judged. In many cases, expectations shaped the whole experience before the interaction even happened. The thesis ends with implications for both research and practice, and notes that the fast development of AI will likely change these patterns in the near future.