Getting the right data is the first and most important thing after which you can work with your instinct when you're not too sure of what to do at certain time.

Sometimes, even after you've gotten some data and you're to an extent sure of what the likely outcome might be, it might not end up giving you the result you're looking for and at other instance, your guts or instinct might just be all you need to get the result you're working on. Instinct works pretty well as much as does whatever analysis you're doing. It's not that you know to know when following your instinct should come in and when relying on just your analysis should come in.