Are you talking about a technical problem, or a conceptual problem? By technical, I mean something like "The service keeps crashing roughly once a week and we

Are you talking about a technical problem, or a conceptual problem? By technical, I mean something like "The service keeps crashing roughly once a week and we can't figure it out." By conceptual, something like "How do I hire the best engineers within my tiny budget?" or "How do we get SOC2 certification even though we do this or that poorly?" In either case, I always start with building an understanding of the problem. Gather what you do know about it. Write down why it is hard. Note any contradictions. Note things you don't know, such as terminology, concepts, or data that is not available. Next, research the things you don't know. Look up the terminology. Read up on the concepts. See what you need to do to get the data. At this point, you don't need perfect answers. Anything that improves your knowledge and understanding is a step in the right direction. At this point you should be able to talk about the problem in clear terms. It should be clear what you know and what you don't know. Tackle the unknowns now. Take steps to make them known. Read. Talk to people. Observe. At some point you might run into contradictions. Think long and hard if they are really contradictions, or if you are making some assumptions incorrectly which are leading to a contradiction. What if you change some assumptions? Does the contradiction go away? If the contradictions seem firm, see if you can redefine the problem ina different way that removes the contradiction. Solving this other problem might not solve the original problem, but it might make the situation better. It might mitigate the problem. This is not a straight path. It's applying the learning loop iteratively. Assess where you are. Decide on a next step to change something. Take those steps. Repeat.