Observe
Pay attention to patterns. How people use things, where they struggle, what they work around without complaining. The real needs are rarely in the brief. You find them by looking at what people actually do, not what they say they want.
Imagine
Picture a world where that problem just isn't there. Not a feature list. More like, what would it feel like if this worked the way it should? That picture is rough and incomplete, but it gives you something to build toward.
Make
Ideas need material. A sketch, a prototype, something to push against. Not to test what you're thinking, but to figure out what you're thinking. If you're only watching the numbers, you'll miss most of what matters. The feel of the thing tells you more.
Constrain
It's tempting to keep adding. Another feature, another option. But the useful work usually happens when you run out of room and have to decide what actually matters. You iterate, strip away what isn't needed, and at some point what's left starts to feel right. Not because you planned it. Because you stayed with it long enough.