In addition to the individuals that recruiters find while searching a system for job candidates, there are people that could be a good fit, but aren't found because of the lack of a specific keyword in a text box or missing value from a dropdown. Clues to a great candidate are hidden in the freeform cover letters and text resumes that might get ignored. I’m guessing that a rough estimate of this class of data would be at least 50% of each data source that is searched. Using the Natural Language Processing power of Watson streamlines the search process and eliminates the need to create difficult and complex Boolean searches that rely on heavily formated data. By also including the "intelligent search and match capabilities" of Watson ensures that a recruiter will consider other qualified candidates that might otherwise have been missed.
Thoughts and Insights from Trillium's Practice Leaders