The product backlog is one of the hardest parts of using an Agile process in my opinion. It quickly becomes a daunting list of tasks that can be unmanageable. What I hate most is the subjectivity of picking what items should be worked on. How do you really weigh what it is important and what isn’t?
Mike Cohn gave a sweet presentation on just this topic. Listen up for Kano analysis, which I think is the simplest and probably most powerful way to weigh new features.