An increasing number of workers turn to digital platforms–such as Fiverr, Freelancer, and Upwork–as an alternative to traditional work arrangements. Digital platforms govern how gig workers join, move through, and leave platforms – often with the help of self-learning algorithms. While digital platforms and algorithms take on HRM practices, we know little about how HRM activities unfold on digital work platforms in the gig economy. The study therefore aims to understand how HRM activities apply to and take shape on digital platforms by studying worker perceptions. We combine supervised text analysis with an in-depth qualitative content analysis, relying on 12’924 scraped comments from an online forum of workers on Upwork. We outline five conversations on HRM practices that pertain to access and mobility, training and development, scoring and feedback, appraisal and control and platform literacy and support. Based on these findings, we build five propositions about how digital work platforms employ HRM activities. Our paper contributes to recent work on HRM on digital platforms by (1) developing a new mixed-methods approach that illustrates how the content of HRM practices may differ from traditional organizations, (2) highlighting the changing role of actors in creating HRM practices by introducing the concept of ‘crowd-created’ HRM practices, and (3) conceptualizing how digital platforms employ a ‘hybrid HRM approach’.