What Recruiters Actually Read on Your Resume

Two claims circulate about how resumes get read, and they appear to contradict each other. One says you have six seconds. The other says recruiters spend several minutes on a resume they like.

Both are measured, both are true, and the apparent contradiction is the single most useful thing to understand about resume screening.

The six-second number, accurately stated

In 2012, the job site Ladders ran an eye-tracking study on 30 recruiters over ten weeks and found they spent an average of 6.25 seconds per resume. Ladders repeated the exercise in 2018 and reported 7.4 seconds, attributing the increase to a slacker labor market producing fewer applicants per role.

Before building anything on that: Ladders has never published a sample size, protocol or instrumentation for the 2018 update, and the 2012 study had thirty participants. This is not a robust literature. It is one company's marketing research, repeated twice, and it is quoted far more confidently than its foundations support.

The six seconds are spent in a predictable place

The more durable finding from the 2012 study is where the eyes went. Recruiters spent roughly 80% of that time on six things: your name, your current title and company, your previous title and company, the start and end dates of both, and your education.

Everything else on the page — the summary paragraph, the skills matrix, the carefully phrased bullets — received the remaining fifth of a very short interval. Not because it does not matter, but because it is not what the triage decision runs on.

Why the layout advice keeps repeating itself

The 2018 update found the same layout patterns that every eye-tracking study of screen reading finds. Resumes that performed well had simple layouts, clearly marked section headers, a legible font, and bold job titles supported by bulleted lists. Resumes that performed badly had cluttered layouts, long sentences, multiple columns, little white space, and missing section headers.

This is not resume-specific. Nielsen Norman Group's eye-tracking research, originally covering 232 users across thousands of pages and confirmed on desktop and mobile in a 2017 follow-up, describes the F-shaped pattern: readers scan horizontally across the top, again slightly lower and more briefly, then vertically down the left edge.

The part usually left out is NN/g's own conclusion. The F-pattern is not a target to design for — it is what people fall back on when content is not formatted to be scanned. Headings, bolded first words and short front-loaded lines exist to break it. Applied to a resume: the F-pattern is what happens to a wall of prose, and section headers plus bold titles are what stop it happening.

The second pass is nothing like the first

Now the other half. ResumeGo put 482 people with direct recruiting experience — recruiters, hiring managers, HR staff and executives — through a hiring simulation built from real job openings and real candidates, covering 7,712 resume evaluations.

Time spent per resume: 2 minutes 24 seconds for one-page resumes, 4 minutes 5 seconds for two-page ones. Two-page resumes were chosen 2.3 times as often overall — 1.4× at entry level, rising to 2.9× for managerial roles — and scored 8.6 out of 10 against 7.1.

That is two orders of magnitude away from 7.4 seconds, and it is not a contradiction. It is the second stage. The seconds-long pass decides whether the minutes-long pass happens.

What the reader brings before they read anything

The uncomfortable part of resume psychology is that the first-pass judgment is made too fast to be deliberative, and fast judgments are where bias lives.

Bertrand and Mullainathan's 2004 field experiment sent roughly 5,000 fictitious resumes to help-wanted ads in Boston and Chicago, randomly assigning White-sounding and African-American-sounding names. White names received 50% more callbacks — an advantage the authors quantified as worth about eight additional years of experience. Callbacks were also more responsive to resume quality for White names, meaning the payoff to improving the document itself was unequal.

Kang, DeCelles, Tilcsik and Jun found the same sensitivity to presentation. Across 1,600 entry-level postings in 16 metropolitan areas, Black applicants received callbacks at 25% with racial cues removed versus 10% with them present; for Asian applicants, 21% versus 11.5%. Employers advertising a commitment to diversity discriminated no less than those that did not.

And increasingly the first reader is not a person at all. University of Washington researchers tested three production language models on 554 real resumes and over 500 real job listings — more than three million comparisons — and found the models favored white-associated names 85% of the time against 9% for Black-associated names.

None of this is something a candidate can fix by writing better bullets. It is here because "what recruiters actually read" has an honest answer that includes what they read into it, and pretending otherwise would make this post a worse guide than it should be.

What to do with all of this

• Front-load the six data points. Title, company, dates, education, legible in one glance in the top third of page one. That is the triage surface.
• Use section headers and bold titles. They are how you avoid being scanned in an F and skipped.
• Do not compress to one page reflexively. The only controlled comparison we could find points the other way, more strongly the more senior the role.
• Write for the second pass everywhere below the fold. It gets minutes, not seconds — but only if the first six seconds bought them.
• Match the posting's vocabulary in those six data points. The triage surface is also the part a filter reads, which is where tailoring earns its keep.

The seconds-long scan is real. It is also the cheapest problem on this list to solve, and it is not where your resume is actually being judged.

Sources

• Ladders, TheLadders Reveals That Job Seekers Have Six Seconds To Succeed (2012 eye-tracking study, 30 recruiters over ten weeks) and 2018 update (7.4 seconds; methodology not published).
• Jakob Nielsen / Nielsen Norman Group, F-Shaped Pattern For Reading Web Content (232 participants, 2006; confirmed 2017).
• ResumeGo, Settling the Debate: One or Two Page Resumes (482 recruiting professionals, 7,712 resume evaluations).
• Marianne Bertrand and Sendhil Mullainathan, "Are Emily and Greg More Employable Than Lakisha and Jamal?", American Economic Review 94(4), 2004, pp. 991–1013.
• Sonia K. Kang, Katherine A. DeCelles, András Tilcsik and Sora Jun, "Whitened Résumés: Race and Self-Presentation in the Labor Market", Administrative Science Quarterly 61(3), 2016.
• Kyra Wilson and Aylin Caliskan, "Gender, Race, and Intersectional Bias in Resume Screening via Language Model Retrieval," AAAI/ACM Conference on AI, Ethics, and Society, 2024 — UW News summary.