Abstract
Algorithmic services promise to empower their users in areas like health management, personal finance and wellbeing. Users assess the legitimacy of novel algorithmic features compared to a status quo, while balancing concerns about uncertain performance expectations and properties of algorithms like low explainability. The paper introduces a research project that aims to conceptualise and empirically explain users’ micro-legitimacy assessments of algorithmic services. We present preliminary findings from a dataset of reviews for popular wellbeing mobile applications that deploy algorithmic features. Our natural language processing approach applies a series of steps to filter and analyse text to extract relevant user assessments. We discuss potential contributions from this work on technology legitimation studies beyond traditional explanations of user adoption.