We discuss recent work, including joint with with Emmanuel Kowalski and Philippe Michel, and possibly others, where new bounds on complete exponential sums were proved, which were then applied to analytic number theory problems. We will also discuss some general heuristics for how useful complete exponential sum estimates will be for a particular problem