Loss of User Control in Personalized Interfaces: How does providing control in streaming platforms affect usability and satisfaction?
Viktoria Kordanowski, Vesta Adomaiyte · Jönköping University Publications (Jönköping University) · 2026
Personalization has become a central feature of user experience in many different platforms, particularly content-based applications, where algorithmic recommendations are used to encourage content discovery. While personalization can improve relevance and usability, it may also reduce the users' own ability to influence the content which is presented to them. This tension creates a challenge when designing a user-friendly interface. The purpose of this study was to examine how providing users with control in web interface personalization affects usability and user satisfaction in content-based applications. To investigate this, a comparative usability study was conducted using two streaming platform prototypes, representing a low and high-control environment. Additionally, a survey was used to gather opinions about usability and satisfaction. Data was collected through task-based usability testing, the System Usability Scale (SUS), and satisfaction questionnaire. The results concluded that both versions of the prototype were rated with high usability and satisfaction. Completion times were similar for the simpler tasks; however, the high-control condition performed better in the most complex task with a difference of 9.8 seconds. Satisfaction ratings also indicated a preference for the high control prototype, although the low-control prototype received a slightly higher SUS score. Findings suggest that user control can improve usability and satisfaction, particularly in situations requiring more complex decision-making. However, increased control could also introduce a learning curve based on individual mental models, which affects the user's initial perception of the platform's usability. This emphasizes the importance of balancing personalization with user autonomy when designing content-based applications.