This study investigates the impact of recommendation systems on user satisfaction through a moderated mediation approach, focusing on the interaction between recommendation types (accuracy and diversity) and user shopping goals. The research explores how these factors influence user satisfaction through psychological mechanisms such as feeling right and psychological reactance. The findings reveal that user satisfaction is significantly affected by the alignment between recommendation types and shopping goals. When a user's shopping goal aligns with the recommendation results in terms of accuracy and diversity, user satisfaction is enhanced. Conversely, when there is a mismatch, user satisfaction decreases. The study also highlights the mediating roles of feeling right and psychological reactance. For goal-directed users, accurate recommendations lead to a feeling of right, increasing user satisfaction. However, for exploratory users, accurate recommendations may trigger psychological reactance, reducing satisfaction. The results indicate that user shopping goals moderate the indirect effects of recommendation types on user satisfaction through these psychological mechanisms. Specifically, accurate recommendations enhance satisfaction for users with specific goals, while diverse recommendations improve satisfaction for exploratory users. Theoretical implications of the study include the integration of user shopping goals into the analysis of recommendation systems, providing new insights into the psychological mechanisms underlying user satisfaction. Managerially, the findings suggest that e-commerce websites should tailor recommendation systems to users' shopping goals, providing accurate information for goal-directed users and diverse options for exploratory users. Additionally, the study emphasizes the importance of balancing personalization with privacy and ethical considerations in recommendation systems. Limitations include the simplified categorization of recommendation types and the potential for recommendation pages to not fully represent real-world shopping experiences. Future research should aim to enhance the realism of recommendation scenarios to better reflect actual user experiences.This study investigates the impact of recommendation systems on user satisfaction through a moderated mediation approach, focusing on the interaction between recommendation types (accuracy and diversity) and user shopping goals. The research explores how these factors influence user satisfaction through psychological mechanisms such as feeling right and psychological reactance. The findings reveal that user satisfaction is significantly affected by the alignment between recommendation types and shopping goals. When a user's shopping goal aligns with the recommendation results in terms of accuracy and diversity, user satisfaction is enhanced. Conversely, when there is a mismatch, user satisfaction decreases. The study also highlights the mediating roles of feeling right and psychological reactance. For goal-directed users, accurate recommendations lead to a feeling of right, increasing user satisfaction. However, for exploratory users, accurate recommendations may trigger psychological reactance, reducing satisfaction. The results indicate that user shopping goals moderate the indirect effects of recommendation types on user satisfaction through these psychological mechanisms. Specifically, accurate recommendations enhance satisfaction for users with specific goals, while diverse recommendations improve satisfaction for exploratory users. Theoretical implications of the study include the integration of user shopping goals into the analysis of recommendation systems, providing new insights into the psychological mechanisms underlying user satisfaction. Managerially, the findings suggest that e-commerce websites should tailor recommendation systems to users' shopping goals, providing accurate information for goal-directed users and diverse options for exploratory users. Additionally, the study emphasizes the importance of balancing personalization with privacy and ethical considerations in recommendation systems. Limitations include the simplified categorization of recommendation types and the potential for recommendation pages to not fully represent real-world shopping experiences. Future research should aim to enhance the realism of recommendation scenarios to better reflect actual user experiences.