Nudges have been critiqued for being too blunt of a tool. For instance, a retirement savings default may be helpful for a group of employees on average, but subgroups, say under-savers or over-savers, might be helped or harmed by this one-size-fits-all approach. As such, there have been calls to develop a more personalized approach to nudging (see here in our collection: “Imagining the Next Decade of Behavioral Science”). This paper outlines two dimensions that behavioral scientists could consider when designing personalized nudges: choice personalization and delivery personalization. Think of choice personalization as “personalization within nudges”—the method of nudge has been set (say, a default) but is tailored to specific individuals (different default leves of retirement contributions, for those over-savers and under-savers). Think of delivery personalization as “personalization as across nudges”—understanding the most effective method to nudge a certain individual. Personalizing nudges does come with data privacy and legal concerns, but these can be overcome, the paper argues.