A team shipped fixes for months to stop users from dropping off. Nothing moved. I see this a lot: the instinct is to redesign. New screens, new components, new colors. Ship it and hope the numbers change. Before touching anything, I pulled their data, watched session recordings, and mapped where users actually left versus where the team thought they left. The answers did not match. They had been redesigning around a problem that was not the problem. That is the step most teams skip. They jump from "something is off" straight into a redesign. The review is unglamorous, but it keeps you from redesigning around assumptions. If your product is not performing, the first move is not Figma. It is looking honestly at what already exists.