Finding out what the customer really wants
IT Conversations has a great speech by Macolm Gladwell (author of "The Tipping Point") about marketing and customer feedback, and how people don't tell you what they really want. This destroyed some of my preconceptions and really made me think. He starts by telling the story of the Aeron chair (apparently the most successful office chair in history) and how all the focus groups universally hated it.
I've long struggled with a variant of this problem: how to know that a public seminar will have a certain number of attendees before deciding to hold it. If I could do that I'll bet I could give a lot more seminars. I think it requires some kind of upside-down inversion-of-control thinking to solve the puzzle. Such as: "groups form around various ideas, and when they reach a certain number of committed members they can commission a seminar." A big problem is that it's not a casual commitment since it's a traveling-and-full-immersion experience, and people have date conflicts etc. But the upshot is that if there's a way that a group of people can form and say to me "we have this many people willing to pay this much for a seminar on one of these dates," we'd work something out. I would imagine that the early committers would get the benefit of lower prices, as incentive to commit, and those that came after the seminar was officially announced would have to pay more.