The resume was never the point
A resume is a document optimized to get past software. It was never designed to show what someone can actually do.
The resume has been optimized for ATS keyword matching over decades, selecting for resume-writing skill rather than job performance. Research shows it amplifies racial bias at scale — identical resumes get starkly different outcomes based on name alone. The actual assessment has to happen in conversation, not on paper.
A document designed for a different era
A resume is approximately one page of words chosen to match a job description that was itself written to match a budget approval that went through three rounds of editing before it was posted. By the time a candidate reads it, the job description is a legal document and a wish list in equal measure. By the time a recruiter reads the resume, it is a keyword-matching exercise. The actual human on both sides of this transaction has largely been edited out.
The resume as a format made sense when it was a letter a person wrote to another person before a meeting. It has not made sense for a long time. Today, most resumes are filtered first by an applicant tracking system, then skimmed by a recruiter in under ten seconds, then passed to a hiring manager who reads it looking for reasons to say no. At each stage, the document is being evaluated for pattern-matching — does this person look like the people we've hired before? — rather than for signal about what they can actually do.
Thirty years of optimizing the wrong thing
The industry has spent thirty years optimizing the resume rather than replacing it. Candidates learned to keyword-stuff. Recruiters learned to filter for institutions and logos. LinkedIn automated the mutual performance. The result is a system that is extremely good at finding people who look right on paper and extremely bad at finding people who are actually right for the job.
Bias baked in
Research from the University of Washington published in 2024 made the pattern-matching problem concrete. When identical resumes carried names associated with different racial groups, white-associated names were preferred in 85.1% of cases, Black-associated names in just 8.6%. The resume didn't introduce that bias. It amplified it, because a document with no other signal leaves readers to fill the gaps with whatever patterns they already carry.
What a better system looks for
What would a better system look for? The same things a good manager looks for when watching someone actually work: how they think through a problem they haven't seen before, what they do when their first approach doesn't work, how they explain something complicated to someone who doesn't have their context. These are assessable. They are just not assessable from a piece of paper.
The resume is not going away. It has too much institutional inertia, too many systems built around it, too much of the hiring process dependent on it as a coordination tool. But it is worth being clear-eyed about what it is and is not. It is a document for getting in the room. The actual assessment has to happen somewhere else — in a conversation, in a task, in a demonstration. Companies that treat the resume as the assessment are not hiring well. They are hiring for resume-writing skill, which is a different thing entirely.