I just picked up a nice story in the book Blink from Malcom Gladwell. He tells the story of the chair designer Bill Stumpf who designed - while working for the Herman Miller company - the most successful chair model Aeron. Now what was surprising is that all market research results at the time showed that the new chair scored high in comfort but very low in aesthetics. So a failure for sure if the product were to be launched. Nevertheless they launched it because the managers in the company liked it, and to the market researchers surprise it became subsequently the most successful model ever designed in their history.

What this story shows is that people have problems judging radically new things, and that's the lesson for any revolutionary new product in technology as well. One can only know whether it is a success or not when it is out there and people have got to know it. Expensive market research programs therefore are better suited for incremental new product improvements rather than radically new ideas.