To answer in sort of a roundabout way, I think the problem domain matters a lot for how much you relatively learn when switching languages. With web app backen

To answer in sort of a roundabout way, I think the problem domain matters a lot for how much you relatively learn when switching languages. With web app backends, for example, most of the problems somehow relate to difficulty maintaining state over widespread, always-on networked systems - things like authentication, sessions, concurrency, large-scale, flexible datastores. So the systems architected gradually get more and more tuned towards serving this particular goal with little programmer effort. And because those problems are a good match to functional programming - because it eliminates state - functional has become popular. But if you were building an embedded system, statefulness is less of a problem: there's usually only one thread, and all memory can be mapped in an orderly way. Your performance constraints are reliant on a single fixed platform, so the methods of inspecting and debugging code change. So to get back to the original question, think about reframing your familiar problems in a way that would simplify the solution with a functional technique. Preferably, look for simplifications that are really big changes and throw out traditional assumptions - the use of relational databases in webapps is one common target for this. Even if you get it wrong you'll get a better understanding of where the functional solution can and can't work.