Searching, be it academic or commercial, can be distracting.
A good searcher, like a good writer, has a point - an object, that he or she follows through the mess of documents, references and other material. He carefully refines and filters information until he gets down to a set of references/data points/criteria that match the objective of the search.
For a historian, this could mean analyzing previous scholars' work to ensure they're not rewriting someone else's argument. This is why graduate students spend years reading the work of previous historians with an eye for understanding what has been done. The history of history.
For a prior art researcher, its not very different. He has a concept and he sifts through references with similar subjects and themes until he's reasonably confident that he's found matches or that there aren't any. The trick, often, is tenaciously holding to your concept and not getting sidetracked by a reference that initially seemed interesting, but actually isn't.
It takes practice and I think, a sense of confidence in your understanding of that initial concept.
