In my talk, I will present a family of multi-goal problems in robo-advising. I will discuss the inherent time-inconsistency and approaches to address it, the quantification of target success and the evolution of preferences for these targets as market and idiosyncratic attributes evolve.