Your Resume Gets 7 Seconds. Here’s What Recruiters Actually Look At.

TheLadders ran an eye-tracking study with professional recruiters and found they spent an average of 7.4 seconds on each resume before deciding yes or no. That’s not a metaphor. That’s the median time a trained human being allocates to a document you probably spent three hours on.

The study covered dozens of recruiters tracked over multiple sessions, watching exactly where their eyes landed and for how long. The result was six fixation points, in roughly this order: your name, your current title, your current company, your employment dates, your previous title and company, and your education section. Your bullet points – the things most candidates spend the most time writing – barely register on the first pass.

That doesn’t mean bullets don’t matter. It means they’re read in a second pass, if you survive the first one.

What recruiters are actually deciding in those 7 seconds

The first pass isn’t about qualifications. It’s about trajectory. A recruiter scanning for a senior engineer isn’t reading your bullets to check if you “improved system reliability” – they’re checking whether your title sequence looks like someone who’s been growing, not stagnating. They’re checking whether your employer names signal the right tier of experience. They’re checking whether your employment dates show gaps that’ll require explanation.

Think of it like this: the first pass is a quick sanity check against a mental model of what the right candidate looks like. You’re being pattern-matched, not evaluated.

If you pattern-match correctly, you get a second pass. That’s when your bullets get read. That’s when your scope, your metrics, and your impact statements matter. The recruiter is now building a case to move you forward – or not.

If you fail the first pass, the second pass never happens.

The top third is the only real estate that matters

Eye-tracking research shows that recruiters follow an F-shaped scanning pattern: a horizontal sweep across the top (your name and title), a second horizontal sweep across your current role, then a rapid vertical scan down the left margin. The bottom third of page one gets low attention. Page two gets almost none on an initial screen.

This creates a simple rule: anything you need a recruiter to see should live above the midpoint of page one.

A few things that often violate this rule, in practice:

• Internal titles that don’t translate. “IC5” means something at Meta, but a recruiter at a 200-person fintech startup won’t know that instantly. Put “Senior Software Engineer (IC5)” and let them move on.
• Company names without context. If you worked at a Series B startup that didn’t get press coverage, a one-line descriptor – “B2B SaaS, 120 employees, $18M Series B” – costs you three words and buys you immediate recognition of the tier.
• A dense objective statement taking up six lines. A two-line positioning headline (“Staff Engineer, distributed systems, 9 years at fintech and healthtech companies”) does more work than a paragraph explaining your goals.

I’ll admit I don’t know exactly how much this varies by industry. The TheLadders research used a broad recruiter population, not a sample narrowed to, say, quant finance or biotech. Pattern-matching behavior probably shifts with domain expertise. But the structural principle – top-left, scannable, titled correctly – holds across most tech hiring I’ve seen documented.

The ATS problem that happens before a human even sees your resume

There’s a stage that runs before the eye-tracking even applies. As of 2025, 97.8% of Fortune 500 companies use a detectable applicant tracking system – and more than 90% of employers at those companies filter or rank candidates before a recruiter views their application.

ATS systems parse your resume into structured fields: title, company, dates, education, skills. They do this by reading your document’s raw text. Multi-column layouts, tables, and text boxes that look clean in a PDF viewer frequently corrupt in parsing. A two-column resume where your current title sits in a text box can end up in an ATS database with a blank title field. You passed the formatting eye test with a human reader, but the ATS recorded you as having no current job.

The practical implication is boring but real: single-column layout, standard section headers, no graphics in the body. Save design creativity for a portfolio site. Your resume’s job is to parse cleanly and position you in the top of recruiter search results, not to look interesting in isolation.

What the bullet points actually need to do

Assuming you’ve passed the first-pass scan and the ATS filter, your bullets are now being read by someone building a case. They’re asking: did this person do work at the right scope? Did they have real impact, or were they a passenger on someone else’s project?

Three things distinguish bullets that answer those questions from bullets that don’t:

• Scope before action. “Led a team of 6 to migrate a 47-service monolith to microservices, reducing deploy frequency from biweekly to daily” tells you scope (6 people), the problem (47 services, a real number), and the outcome. “Worked on microservices migration” tells you nothing.
• Impact that a hiring manager can care about. Engineering managers care about reliability, speed, and cost. “Reduced P99 latency from 340ms to 90ms for the checkout service” is concrete. “Improved system performance” is not.
• Technology mentioned where it’s specific, not everywhere. One bullet mentioning Kafka in the context of a real architectural decision is more credible than a skills section listing 23 technologies alphabetically.

The bullets don’t need to cover everything. They need to make the recruiter’s job easier – give them a phrase they can drop into the “why I’m recommending this candidate” note they’ll write if they move you forward.

A note on length

Two pages is fine for anyone with more than 7-8 years of experience. One page is fine for anyone with less. The “one page only” rule is, in my opinion, a holdover from a pre-digital era when resumes were faxed and printing cost mattered. Modern ATS systems don’t care about page count. Recruiters at most tech companies don’t either.

What they do care about is whether your most relevant experience is buried on page two. If you have 12 years of experience and your current role – the one most relevant to the job you’re applying for – starts on page two because you front-loaded a lengthy summary, that’s a structural problem. Not a page-count problem.

Where practicing your pitch changes what you write

One pattern we’ve noticed from how candidates use mock interview sessions on LastRoundAI: the candidates who struggle to articulate their impact in the first 90 seconds of a behavioral question are often the same ones whose resumes have weak bullet points. Not because they lack the experience – but because they haven’t had to put it into words under any kind of pressure before.

Practicing “tell me about a project you’re proud of” in a low-stakes setting forces you to identify which parts of your work are actually interesting to an outsider. That answer, once you’ve refined it a few times, usually tells you exactly what your bullet point should say. The resume and the verbal pitch end up calibrating each other.

It’s worth running that loop before you finalize a resume, not after you’ve already sent it to 30 places.

If you want to get your resume’s structure right before the next round, the LastRoundAI Resume Builder runs through the top-third positioning and ATS formatting issues that the eye-tracking research points to. And for getting your verbal pitch sharp enough that your bullets actually reflect what you can say out loud, behavioral interview preparation is the part most people skip until it’s too late.

Whether the 7.4-second figure is exact or slightly off for your specific recruiter doesn’t really matter. The structure of how resumes get evaluated – pattern match first, detailed read second, ATS before either – is consistent enough that it should change how you format and prioritize. That’s the part worth getting right.