Over the past decades, shared micromobility services (SMSs) have experienced rapid growth worldwide. This increase has raised important questions regarding how to enhance the efficiency of service operation and regulation. Addressing these issues requires a clear understanding of how and why SMSs are utilised. This research aims to infer trip purposes of two types of dockless SMSs—dockless shared bikes and e-bikes—and to compare their usage patterns, primarily using an analytical framework based on Latent Dirichlet Allocation (LDA). The proposed framework is applied in Ningbo, China, by integrating a large-scale bike-sharing trip dataset with a Point of Interest (POI) dataset. In our case study, we successfully identified seven typical trip purposes: transport, work, lodging, eating, shopping, education, and others. The results reveal that eating is the most common trip purpose for both bike and e-bike users. Interestingly, the proportions of transport-related trips are relatively low compared to previous findings, for both shared bikes and e-bikes. Furthermore, shared bikes play a more significant role than shared e-bikes in daily trips related to lodging and education, suggesting that the importance of shared e-bikes for these trip purposes may have been previously overestimated. This study offers valuable insights into the trip purposes of SMSs and provides targeted recommendations for planning and operations, contributing to sustainable development of urban micromobility systems.