“Crowdworking” (CW) seems to epitomize the highly flexible online world of work. Platforms offer tasks to the crowd, their registered freelancers or crowdworkers, and pay them on a task-by-task basis. But some platforms focus on tasks with considerable skill needs such as text creation and are therefore interested in engaging skilled crowdworkers on a more long-term basis. In this paper, we argue that such platforms incentivize and commit their crowdworkers through an ingenious “rating-based compensation system”, we explore how affective and calculative commitment evolves among crowdworkers on one platform, and we test how commitment is related to participation and intention to stay. Based on survey data and a fuzzy-set qualitative comparative analysis (fs/QCA), we identify six groups of committed crowdworkers, each with a specific combination of needs and satisfaction with the platforms compensation system. Furthermore, we find that affective commitment is associated with a stronger intention to stay with the platform, and that calculative commitment is associated with more work hours on the platform (participation). In sum, this paper adapts the concept of organizational commitment to the crowdworking context, demonstrates how distinct groups of committed expert crowdworkers can be identified via the configurational fs/QCA method, and illustrates how rating-based compensation systems motivate and engage crowdworkers on a long-term basis.