Introduction and Purpose: The growing integration of artificial intelligence (AI) into digital services has transformed customer–business interactions, particularly through AI-powered chatbots that provide rapid responses, readily accessible assistance, and personalized communication. However, the availability of intelligent technologies does not necessarily ensure customers’ continued engagement with digital platforms. In increasingly competitive online environments, understanding the factors that encourage customers to return to and repeatedly interact with a platform has become essential. This phenomenon, referred to as customer stickiness, reflects customers’ tendency to maintain their engagement with a platform despite the availability of alternative options. The present study examines the emotional and functional pathways through which chatbot characteristics contribute to customer stickiness. Specifically, it investigates the effects of three chatbot characteristics—usability, responsiveness, and anthropomorphism—on two dimensions of perceived value: extrinsic value, which reflects the functional and utilitarian benefits of interactions, and intrinsic value, which represents their experiential and emotional benefits. Furthermore, the study examines how these two dimensions of perceived value affect customer stickiness. By integrating functional and emotional perspectives into a unified framework, this research seeks to identify how chatbot characteristics generate perceived value and how these value dimensions, in turn, contribute to customers’ continued engagement with digital platforms. Particular attention is devoted to comparing the relative contributions of the emotional and functional pathways to explaining customer stickiness.
Methodology: This study is applied in terms of its purpose and adopts a quantitative approach with a descriptive-survey research design. The statistical population consisted of users with prior experience interacting with AI-powered chatbots and intelligent service systems on digital platforms. Given the absence of a complete sampling frame, convenience sampling was employed. A total of 234 valid questionnaires were collected and analyzed. Data were gathered using a structured questionnaire adapted from previous studies in the areas of AI-enabled services, chatbot interactions, perceived value, and customer stickiness. The questionnaire measured six constructs: usability, responsiveness, anthropomorphism, extrinsic value, intrinsic value, and customer stickiness. All items were measured on a five-point Likert scale ranging from strongly disagree to strongly agree. The measurement model was assessed using Cronbach’s alpha and composite reliability to evaluate internal consistency, factor loadings and average variance extracted (AVE) to examine convergent validity, and the Fornell–Larcker criterion to assess discriminant validity. Partial least squares structural equation modeling (PLS-SEM) was conducted using SmartPLS to evaluate the structural model and test the eight research hypotheses. Path coefficients, t-statistics, significance levels, and coefficients of determination (R²) were used to evaluate the proposed relationships.
Findings: The findings supported all eight research hypotheses, indicating that usability, responsiveness, and anthropomorphism were positively and significantly associated with both extrinsic and intrinsic value. Usability exhibited a positive and significant relationship with extrinsic value (β = 0.203, t = 3.589, p = 0.001) and intrinsic value (β = 0.299, t = 2.845, p = 0.004). Responsiveness also demonstrated positive and significant relationships with extrinsic value (β = 0.532, t = 9.389, p < 0.001) and intrinsic value (β = 0.128, t = 2.475, p = 0.013). Among the three chatbot characteristics, responsiveness had the strongest estimated relationship with extrinsic value, underscoring the importance of timely and relevant responses in enhancing customers’ perceptions of functional benefits. Anthropomorphism was positively and significantly associated with extrinsic value (β = 0.362, t = 2.380, p = 0.017) and intrinsic value (β = 0.469, t = 5.524, p < 0.001). Its stronger estimated relationship with intrinsic value highlights the importance of human-like communication and interaction characteristics in shaping customers’ emotional and experiential evaluations. The coefficients of determination indicated that the model explained 64.6% of the variance in extrinsic value (R² = 0.646) and 51.1% of the variance in intrinsic value (R² = 0.511). Furthermore, both dimensions of perceived value exhibited positive and statistically significant relationships with customer stickiness. Extrinsic value had a path coefficient of β = 0.315 (t = 2.455, p = 0.014), whereas intrinsic value had a coefficient of β = 0.448 (t = 2.996, p = 0.003). The estimated coefficient for intrinsic value was numerically higher than that for extrinsic value, suggesting a stronger association between emotional and experiential benefits and customer stickiness within the proposed model. However, this numerical difference alone does not establish a statistically significant difference between the two relationships.
Discussion and Conclusion: Overall, the findings indicate that customer stickiness in AI-powered interactions is associated with the combined contributions of functional efficiency and emotional experience. While responsiveness plays a particularly important role in enhancing extrinsic value, anthropomorphism is more strongly associated with intrinsic value. Moreover, the results underscore the importance of both dimensions of perceived value in explaining customers’ tendency to continue interacting with digital platforms. These findings suggest that customer stickiness should not be understood solely as the outcome of efficient service delivery; rather, it also reflects the quality of the emotional and experiential benefits generated through interactions with AI-powered systems. From a managerial perspective, digital businesses should adopt an integrated approach to chatbot design and implementation. Improving usability, response speed, and information relevance can strengthen customers’ perceptions of functional value. At the same time, developing natural, accessible, and appropriately human-like interactions can enhance the emotional and experiential quality of customer engagement. Accordingly, organizations should regard functional performance and emotional experience as complementary dimensions of customer retention strategies rather than treating them as independent priorities. This study contributes to the literature on AI-enabled customer interactions by proposing an integrated framework that explains customer stickiness through functional and emotional pathways. Nevertheless, the findings should be interpreted in light of the study’s limitations, particularly its use of convenience sampling and a cross-sectional, survey-based design. Future research could employ more diverse sampling methods and longitudinal or behavioral data to investigate how perceived value translates into actual patterns of repeated customer engagement.