What do you think about your results? Are you seeing the type of information you were hoping for? Are there too many results to look through? Too few? By looking at your results, you get instant, empirical feedback on whether the keywords you chose to describe your concepts are used that way in research.**
Harvesting terms in my example, search yielded almost 1,400 more articles to look through! And each article will have its own set of subjects to continue guiding me forward.
**Note that just because a keyword is not producing results does not mean it is a wrong word. It may just mean that in the context of this particular research database, your concept is described in a different way.