If you could only ask one question to your customers, what would it be?

@hamid_ra I agree about the "find their pain-point" part. I disagree about the "aesthetic" part. In order to "like" a feature a feature doesn't have to be aesthetically pleasing. Aesthetics have little to do with the probability someone will like the feature. The bigger problem with no.2 is that it is vague, unreliable and subjective "metric". It says nothing about the feature and its usefulness. People may like some feature but never really use it, or only sporadically. On the other hand, people may not like some feature and use it extensively. Think about Facebook and feed. In the beginning, people literally hated it. However, they used it more than any other feature. Whether users like some feature is almost irrelevant. It should not be the main reason for keeping or ditching it. A better thing to look for, or more objective metric to measure, is usage. It is more important if users use some feature frequently than if they like it. Usage can also be a good way to find out if that feature solves users' pain. Feature that is used often is very likely to solve some pain your users have. Otherwise, why would they waste their time using that feature? So, don't ask users if they like feature X. Instead, measure how much they use feature X. This will give you a better and more objective answer.