Customers often struggle to describe the solution they need, even when they can clearly describe their pain; Henry Ford's famous line that customers would have asked for "a faster horse" captures why discovery focuses on problems, not requested features. A good user interview asks open-ended questions about past behavior ("walk me through the last time you tried to do X") rather than hypothetical future behavior ("would you use a feature that does Y"), because people are unreliable predictors of their own future actions but far more accurate describing what they've actually done. You will learn to avoid leading questions, which subtly suggest the answer you want to hear, and to recognize the difference between a stated preference ("I want a dashboard") and the underlying need it may be masking ("I need to know quickly if something is going wrong").