I distinguish between work and problems.

Work can be done by following well-known, accepted rules. It may not be easy, but you know from the start how hard it is and how much effort it takes. Work just takes time, energy and commitment. So, just do it already!

There are no predefined rules for problems. You have to figure them out as you go along. It's not even clear how to know which solution is best, because problems usually have multiple perspectives with (possibly conflicting) interests. Problems may even have multiple stakeholders. So you have to analyze the problem to come up with possible solutions, evaluate the solutions by all different criteria, and than choose on basis of incomplete information. And that's all before you can start actually solving the problem itself.

The obvious trap here is that you can get stuck in the analysis/evaluation, because choosing is hard and often involves conflict. So my approach often is to just choose one option, fix the problem and deal with the consequences. Which may require me to get back to the drawing board to come up with a whole new solution. I am happy to solve a problem more than once, because it means you always have a solution in place (as imperfect as it may be), and it will gradually get better.