New data: 87% of hiring managers say their AI screening software reads plain, text-only resumes more accurately than fancy visual ones. So that slick template with icons and skill bars might be getting you filtered out before a human ever sees your name. I came across this breakdown from a savvy professional who spent weeks going through hiring reports and studies from MIT and Oxford, covering more than 4,000 hiring managers and nearly 2 million applications. His bottom line is simple: resumes still matter, but the rules have changed now that AI sits on both sides of the hiring table.
I’ll be honest, I expected this to be another “AI killed the resume” take. It’s the opposite. The data shows exactly where humans still decide, and how to use AI without looking lazy.
The five findings that matter
The author boiled the research down to five numbers, and each one has a direct consequence for how you write your resume:
- 87% of hiring managers say their AI tools read simple, text-based resumes more accurately than visual ones. Readability beats appearance.
- Tailored resumes had 84% higher interview rates (5.71% vs 3.09% for untailored ones). But resumes with the highest keyword coverage got 21% fewer interviews than those with moderate coverage. Tailoring works, stuffing backfires.
- 59% of hiring managers see AI use as a good sign, yet 28% reject AI-heavy resumes that show little effort. Both are true at once.
- Resumes that quantified impact saw 75% higher interview rates than those that only listed responsibilities.
- 60% of hiring managers want candidates to prove AI skills, not just list them. Only 6% say AI is the final decision-maker, so a human still reads what passes the filter.
The creator frames this as two stages. Stage one, pass the AI: your file has to be easy to parse and clearly match the role. Stage two, pass the human: a real person has to want to move you forward. Rules one and two handle the machine. Rules three through five handle the person.
Three practical applications
- Make the machine’s job easy The expert’s advice here is delightfully boring. One column. Conventional headings like Summary, Experience, Education, Skills. No skill bars, no big images, no headshot unless local norms call for it (he points out that headshots are standard in China, and some employers there even ask for your zodiac sign). Export as a PDF with selectable text, ideally under 2.5 MB, since some systems openly state they can’t parse anything larger. Quick test: open the PDF and try to highlight and copy a line. If you can’t, your text is trapped inside an image and the AI can’t read it either.
- Map keywords, don’t stuff them This is where the 84% vs 21% contradiction resolves. Keyword mapping means using the job description’s language to describe work you actually did. Keyword stuffing means cramming every phrase in whether or not you can back it up. The author shares a stuffed bullet he genuinely read while at Google: “Delivered customer support, customer service, and customer-focused communication to improve the customer experience.” Compare that to a mapped one: “Reduced customer complaints by 31% using AI to create an onboarding email sequence from help center documents.” His workflow: give AI the job description plus your base resume, ask it to identify the problems the employer needs solved and the skills that matter most, then ask for stronger bullets. Keep only what you can prove in an interview.
- Brain dump first, polish second An MIT experiment with nearly half a million job seekers found that using AI for spelling, grammar, and wording raised hiring probability by 8%. A separate study found that giving applicants ChatGPT made pitches so similar that evaluators had a harder time telling who was qualified. The difference is the input. The original poster suggests writing messy notes for every role, covering three things: the final result, what you specifically contributed, and how you did it. His example note is pure rambling about conflicting refund answers between sales and support. After AI polishing, with strict instructions not to change facts, it becomes: “Decreased duplicate tickets by 50% by creating a shared refund guide for sales and support so customers received consistent answers.”
Prompt of the day
Two prompts the creator uses for the numbers step. First: “Help me identify the most relevant metrics for each experience on my resume. Ask me clarifying questions if needed.” Then, once you’ve supplied verified numbers: “Rewrite each bullet using Google’s XYZ formula: accomplished X as measured by Y by doing Z. Do not add or change any facts or numbers.” Revenue isn’t the only metric. Time saved, speed, scale, and accuracy all count.
Tips and pitfalls
🔹 Lead with an AI achievement. The author recommends making the first bullet under each role an AI-related win that’s relevant to the job. Example: “Cut weekly customer feedback reporting from 2 hours to 30 minutes by using Claude Code to build a shared database of recurring issues.”
🔹 No workplace example yet? Build one. Do a small project for the job you want and list it under a Projects section. Then make it inspectable: a GitHub repo, or simply a Google Doc explaining the situation, how you used AI, and what changed. Hiring managers said they want proof through tasks (26%), work examples (19%), or certifications (15%).
🔹 Watch for generic AI filler. Phrases like “results-driven professional with a proven track record” are exactly what tired reviewers skip. If you couldn’t say a line naturally out loud in an interview, cut it.
🔹 Experience above education, unless you’re a fresh grad. 86% of hiring managers value relevant work experience over formal education.
What I loved about this one is that none of the advice is brand new, but seeing the numbers behind it makes it much easier to follow with confidence. Check out the full video for the complete walkthrough, the exact prompts, and the bullet point examples he shares along the way.