Below is based upon a reply I made to the ixda list awhile back re: listening to customers (or when not to!)…
I always take user research findings (quantitative or qualitative) with many grains of salt. It’s supplementary data to help a designer understand, empathize, interpret, and then make a “good” decision. Designers are informed visionaries, not “short-order cooks” doing simply what the user asks b/c often users often don’t know what they want, nor how to express exactly what they want. If they did, we wouldn’t have jobs :-) Designers must exercise their best judgment (comes with years of experience, I fully realize) to use or dismiss that data accordingly per the project needs.
Also, what’s the project goal?
1) Breakthrough innovation: It’s important to note that no user specifically asked for a Wii, iPod, Prius, Dyson or flickr, but once manifested, then users wanted them. Discovery activities, asking users their motives/reasons, assessing broader social/tech trends, defining various scenarios might help…but again, take salt with what you find!
2) Incremental clean-up: But if it’s minor tweak for the next point release, listening to those 500 complaints on your forum about the wrong button label might be good :-) Then again, if there’s a valuable opportunity to introduce an innovative UI or behavior, then try it, get a pulse on your users’ reaction (with beta testing or other approaches), and then decide how to proceed.
For other resources on this issue, I’d suggesting googling the following folks: Larry Keeley (Chicago ID), Vijay Kumar (Chicago ID), Roger Martin (Rotman), and Craig Vogel (Cincinatti)–all academics but with practical understanding how breakthrough innovation happens in large corps like Motorola, Ford, Whirlpool, etc. while leveraging customer feedback and testimonials.
But I’d say: Gather some user info, question what’s said, reflect on it, create a solution per your design abilities, evaluate and iterate. (and don’t forget the salt!)