Don't mention about your idea as far in the conversation as possible!

Rules:

1. Talk about their lives, not about your idea;
2. Ask about the specific things from the past, not about hypothetical opinions about the future
3. Talk less, listen more

Amazon Link Essential to figure out if you have a good idea

Golden rules:

1. Opinions are useless.
2. Any predictions about the future is a lie, usually overoptimistic.
3. People know about the problems they face, but don't know how to solve them.
4. Know the goals of your user
5. Some problems in fact are not problems
6. Watch your users in real life solving their tasks
7. If users didn't try to solve this problem before, they wouldn't pay attention to your product.
8. Ask question: "Who else I should talk to?", "Are there other questions I should ask?"
9. Avoid compliments at all costs.
10. Analyze requests for new features.
11. Ask tough questions! You need the truth! It's wonderful if the response is negative! Thus unemotional answers "well, .. seems not so bad" are better than "WOW!" answers (they give more info, wow-answers are unreliable).
12. With high product risks - you need to start developing, you cannot prove your business idea just from the interviews. You have to embrace them.
13. Prepare top 3 questions for any interview (to make them productive).
14. Better to start talking with clients with informal conversations. Avoid formalities until you have something to show them
15. People that are not going to buy anything from you - are especially dangerous source of unambiguous information.
16. There are no "ok" meetings, all of them are either successful or failed.
17. Conduct interviews until you get new info. If you're getting new info from every customer consider narrowing your target audience.