The analytics group at Memorial Sloan-Kettering Cancer Center analytics group uses a variety of methods and tools to leverage a database of more than five million records. This task is approached from two angles: first, prospects are prioritized using a capacity estimate and a predictive model for inclination. Second, custom predictive models are built to produce lists of 500 prospects at a time for personalized telephone stewardship. The presenter will explain how predictive analytics informs MSKCC's dynamic prioritization process, which feeds the pipeline with candidates for research, qualification and potential upgrade at all levels of the operation.