My thesis both develops and applies representation learning to three problems in marketing strategy. In the first chapter, I train transformer-based representations of individual donors to solve an identification problem ...