Before you touch a single pixel, you sit down and you listen. Not to be polite — to understand.
Good discovery work isn't about gathering ammunition for decisions you've already made. It's about interrogating your assumptions until they either hold or fall apart. The ones that fall apart? Those are the most valuable things you'll find in this phase.
You talk to founders. You read the journals. You do competitor autopsies. You survey the people who actually use the thing — or who tried to and gave up. And you come in with your mouth shut and your ears open, because the moment you start pitching in a research session is the moment you stop learning.
What you're after is the real problem. Not the one in the brief. The one underneath it.