http://techcrunch.com/2011/05/25/absolute-must-watch-office-...
http://techcrunch.com/2011/09/12/tc-disrupt-office-hours-wit...
http://techcrunch.com/2011/05/25/absolute-must-watch-office-...
http://techcrunch.com/2011/09/12/tc-disrupt-office-hours-wit...
The solution is to practice beforehand, preferably on people who will tell you candidly when you haven't yet explained yourself clearly.
The test of how well you've explained a startup idea is how accurately the listener could reproduce it from your description.
But I should not be too sarcastic. I really liked these. A lot of the interviewees were very promising even after your well meaning attempts to shoot holes in their business models. And your comments are very informative.
But it seems that your comments and questions also indicate how hard it is to build a successful start-up. It is not merely about making something useful or better than the competition, you have to make something that cures a significant pain point so that your customers are actually motivated to go out and get it. It really does make one take a pause and question things.