Given the complex characteristics of city structure, the analysis of tourists’ activities in urban space is challenging. Previous studies on urban tourists’ activities were mostly limited to official administrative units, such as attractions, blocks, and administrative districts, and rarely considered the multi-scale characteristics of urban tourism space. This study offered insights into urban tourism by depicting multi-scale tourism spaces from the perspective of tourists’ activities. Based on a spatial conceptual model, the urban destination was divided into multi-scale spaces, including tourist areas of interest (areas of intensive tourist activity), attraction-community complexes (comprising tourist attractions and supporting service facilities), and tourism districts (subregions with different tourism types). The Geographic Information System field and spatial interaction models were combined to identify multi-scale tourist activity spaces using digital footprints from geolocated social media data. The proposed framework was verified in Suzhou, a typical urban destination in China, which confirmed that the systematic and comprehensive methodology could be used as a tool for smart urban destination management and planning based on tourists’ demands.