Jeff Bezos set Amazon’s hiring standard so high that even his own referral didn’t pass. After interviewing someone he vouched for, the team compiled a six-page memo on why the candidate didn’t clear the bar. Bezos glanced at it and said, “Okay,” illustrating that referrals don’t override performance. That story underpins the Bar Raiser program: a trained interviewer from another team who enforces a consistent threshold. Bar Raisers veto hires who wouldn’t outperform half the current people at that level, regardless of resumes or headcount pressure. After nearly a thousand interviews and two decades at Amazon, the author identifies three candidate types that always fail. First, the “half-prepared” candidate nails coding and architecture but flunks behavioral questions. They’ve studied algorithms for months yet can’t cite any risk they took or disagreement they handled. That signals poor judgment or lack of real decision-making power—both red flags, since most Amazon exits stem from behavioral issues, not technical mistakes. Second, the “textbook” candidate memorizes Amazon’s Leadership Principles and tags each answer with names like Bias for Action. Their stories, though polished, describe routine work—no initiative, no real push beyond assigned tasks. Labeling a routine outcome doesn’t prove the principle. The third group earns sympathy: they deliver one standout story so often that it becomes a crutch. Maybe they rebuilt a high-traffic system or launched a successful feature, but when pressed for fresh examples, they fall back on the same narrative. That repetition hides real breadth of experience. Amazon interviews probe multiple dimensions—ownership, inventiveness, customer obsession—and relying on a single anecdote means you can’t demonstrate every trait. Bar Raisers spot that quickly and lean toward candidates who show diverse, concrete examples instead of one shining moment.