Look for real stories
Instead of asking “Would you use this feature?”, ask about the last similar situation. What did the person do, what tools did they use and where did they stop? A specific story usually reveals more than polite approval of an idea.
Do not defend the product
When someone does not understand a screen, it is tempting to explain the intention. During research, first find out what they expected to see. Otherwise, the conversation becomes a presentation and hides the very difficulties it was meant to uncover.
Do not generalise too quickly
One answer does not describe the whole audience. Compare observations, record differences and acknowledge the limits of your sample. A useful conversation produces both findings and questions for the next round of testing.