But the human brain is very good at talking about specifics. Questions like "can you walk me through what you do everyday" or "are there any repetitive tasks that you do day-to-day" will lead you down a much more interesting path. You’ll find problems that your software actually can solve.

Great advice. I like to take this even deeper:

  - What is the next thing you do?
  - Who do you call / talk to / pass this to?
  - Show me that form with real data on it.
  - What's the worst thing that happens? 
  - How often does that happen? (hourly, daily, etc.)

And my 3 favorite questions (which are often very revealing):

  - What % of time does that happen?
  - On a scale of 1 to 10, how good is that?
  - What is the #1 thing needed to take that to a 10?

Also, I try to ask these questions to > 1 person. The discussion of 2 or more people coming up with the answer is often more educational than the answers themselves.

I have at least 10 years of real work in my pipeline thanks to these techniques. I never worry about "make something people want".

Great post, Dan and Jason. Thank you!