We describe an approach to constructing monad–theory correspondences which are, in a sense to be made precise, the best possible ones for a given situation. The scaffolding around this result turns out to be itself interesting, giving a sensible framework to talk about presentations and sketches. This is joint work with John Bourke.