A candidate surfaced a number worth sitting with. Google studied its own hiring data to ask the obvious question: does adding more interviewers actually make the final call better?
Four interviews, the study found, predicted the hiring decision with roughly 86% confidence. A fifth interviewer added less than one percent more accuracy. Past that point the company was adding meetings, not signal.
So Google capped most loops at four. Time-to-hire fell by about two weeks, and the hires did not get worse.
Meanwhile, the candidate wrote, the processes they keep meeting run six rounds, seven, sometimes nine. The question they left hanging: who, exactly, benefits from the extra four?