This talk will provide a short overview of our work with operators of smart transportation systems (bike-share systems and an on-demand public transport provider). In each of these engagements, we use large-scale system-use data to estimate commuter responses to aspects of the system’s operational performance. For example, we use data from a bike-share provider to understand commuter responses to system reliability, ease of access and availability. The talk highlights the role of building and estimating detailed models of commuter behavior as a key input to compare and contrast alternate system designs.