The Effect of Digital Empathy on Purchase Intention: An Integrated Behavioral MCDM Study of Virtual Influencers and Chatbots

Özet

Purpose – This study examines how digital empathy influences consumer preferences in AI-mediated interactions by comparing virtual influencers, e-commerce chatbots, low-empathy agents, and hybrid AI agents
Design/methodology/approach – A qualitative thematic analysis based on interviews with 25 participants was conducted to identify core dimensions of digital empathy. These dimensions were subsequently modeled using Fuzzy DEMATEL to reveal causal relationships and Fuzzy TOPSIS to rank digital agents according to their overall performance. Sensitivity analysis was employed to assess the robustness of the rankings under alternative weighting scenarios.
Results – The thematic analysis identified trust (frequency = 31) and emotional response (25) as the most salient components of digital empathy. Fuzzy DEMATEL results showed that trust emerged as the strongest causal factor (total influence score = 3.91), exerting significant influence on empathy-related dimensions and purchase intention. Fuzzy TOPSIS rankings indicated that the hybrid AI agent outperformed all alternatives (closeness coefficient = 1.0000), followed by the virtual influencer (0.5837), chatbot (0.1472), and low-empathy agent (0.0000).
Discussion – By integrating behavioral theory with fuzzy MCDM techniques, this study demonstrates that hybrid AI agents achieve superior performance by simultaneously activating socio-emotional and functional drivers of consumer decision-making, offering a structured explanation of why certain digital agents are more effective than others.