One of my pet peeves is when people say 'we do it this way because that's the way it's always been done'.
Apart from traditional recipes, I don't think it makes sense to approach problems this way. Instead, I try to rely on first-principles thinking, which I first learned about from Elon Musk. First-principles thinking is particularly useful because it can help you look at the problem from new perspectives and helps to remove prior misconceptions you might have.
Step 1 - What are your current assumptions?
Write a list of the current assumptions you have; even this process itself can be illuminating.
Example (to take a recent real-life problem I was tackling, where I wanted customers to get their orders earlier): 'The lead time for shipping this product out from the warehouse is 5-7 days. It cannot be any faster.'
Step 2 - What are the fundamental principles of this problem?
Example 'What is the actual process of picking and packing from the moment the goods arrive at the warehouse? Which specific aspect takes the longest? Why does each stage take as long as it does? Which parts are currently automated and which are manual? Which manual parts can we actually automate?'
As it turns out, the bottleneck in our process was the hand-stamping of each product, which our fulfillment partners had thought was a non-negotiable for us.
Step 3 - Create the new solution from the ground up
Example: If we switched from hand-stamping to directly printing the stamps, the entire process could be automated. This shortened the lead time from 5-7 days to 1-2 days.
Step 4 - Realizing when it's not the right time to employ First Principles Thinking
Recently, an exasperated friend of mine said to me: "Look, not every aspect of human emotion can be reasoned with from first principles. I understand there are solutions - but right now, I just want someone to listen."