The Dual Role of Self-Efficacy and Perceived Privacy Risk in Shaping Consumer Responses to Recommendation Algorithms

Recommendation algorithms have become a routine part of digital platforms, especially in e-commerce and online information environments. Much of the existing literature explains their adoption in terms of technical quality and system performance. Yet users do not respond to these systems on technical grounds alone. Their reactions are also shaped by whether they feel able to use algorithmic features effectively and whether the underlying personalization process feels acceptable from a privacy standpoint. This study examines how self-efficacy and perceived privacy risk jointly shape consumer responses to recommendation algorithms. Drawing on the Technology Acceptance Model, Task-Technology Fit theory, Social Cognitive Theory, and prior privacy research, the paper develops a dual capability-risk framework and tests it using survey data from 515 valid respondents. The results indicate that self-efficacy is positively associated with perceived ease of use, perceived usefulness, satisfaction, and consumer response, whereas perceived privacy risk is significantly related to satisfaction and consumer response. The findings suggest that recommendation systems are more likely to be accepted when users both feel capable of using them and feel reasonably comfortable with the way personal data are involved.