Organizations require good performance from individuals to achieve their objectives. In view of the growing presence of technology, it becomes necessary to understand performance in the context of information systems. Previous research shows that knowledge and perceived usefulness factors have direct effects on performance. However, the literature also recognizes that there may be different man–machine arrangements to carry out the tasks (level of automation). This study, using a multi-disciplinary approach, evaluates empirically whether the level of intervention moderates the effects of knowledge and perceived usefulness on performance. A questionnaire was used to collect data from 201 users in different organizations and different functional areas. The structural equations model was used for analysis. The results show that the degree of automation moderates the direct relationships. Thus, in structured and proceduralized environments, at high levels of automation, the relevance of knowledge of the task may decrease, and at low levels of intervention, the relevance of perceived usefulness may fall.