All articles
Data · Job SearchJanuary 14, 20265 min read

We Tracked 47 Job Applications Over 8 Weeks. Here's Where They Actually Got Lost.

Most job seekers assume rejection happens at interview. Our tracking data tells a different story — and the real drop-off point is earlier than almost anyone expects.

The Assumption That Costs People Interviews

When applications go unanswered, the natural assumption is that something went wrong in the process — a weak cover letter, a poorly timed follow-up, an interview that didn't land. We tracked 47 applications submitted across marketing, operations, finance, and software roles over an eight-week period to map exactly where in the funnel candidates were losing ground.

The results shifted our thinking. The majority of applications weren't being evaluated and rejected — they were disappearing before evaluation began.

61%
of applications in our tracked cohort showed no sign of human review. They were filtered before a recruiter opened the file.

Mapping the Drop-Off Points

We categorised each application outcome by the furthest stage reached: ATS filtered, recruiter screen, hiring manager review, interview, or offer. The distribution was striking. Sixty-one percent never progressed past automated filtering. A further nineteen percent reached recruiter review but were deprioritised before a screen call. Only twenty percent generated any direct human interaction.

This means that for every five applications submitted, four were effectively decided by an algorithm — before a single person read a word of the CV.

"The question isn't whether your CV impresses a recruiter. It's whether your CV survives long enough to reach one."

What Changed When We Optimised for ATS First

In the second half of the tracking period, we applied ATS optimisation — keyword alignment, format standardisation, section header correction — before submitting. The shift was significant. ATS pass-through rate improved from 39% to 71%. Human review rate increased from 20% to 44%.

The experience and qualifications didn't change. The vocabulary and structure did.

✗ Before — Pass rate: 39%

Generic headers, creative section titles, skills listed in a table, summary with no role-specific keywords.

✓ After — Pass rate: 71%

Standard headers, single-column layout, keywords mirrored from job descriptions, role title in summary.

What we changed

  • Replaced all creative section headers with ATS-standard versions (Work Experience, Skills, Education)
  • Removed tables and columns — converted to single-column plain text
  • Added the exact job title from each posting to the CV summary before applying
  • Pulled the top 8 keywords from each job description and verified they appeared in the CV
  • Ran each CV through an ATS checker before submission — targeted 65%+ keyword match

Frequently Asked Questions

How do you know applications were ATS filtered and not just ignored?

The signal is timing and pattern. ATS rejections typically arrive within hours and use identical templated language regardless of role. Human-reviewed rejections take longer and occasionally include role-specific feedback. The volume and speed of non-responses in our tracked cohort was consistent with automated filtering.

Does ATS optimisation help for all roles?

For any role where applications are submitted through an online portal — which covers the majority of mid-to-large employer hiring — yes. Direct outreach, referrals, and some small employer hiring bypasses ATS entirely.

Ready to optimize your CV?

Get an instant ATS score, AI-powered rewrites, and land more interviews — in minutes.

Get Started Free

14-day free trial · No credit card required

Everything inside CV Optimizer AI

Every tool you need — in one place.

From CV score to shareable link. See exactly how each feature works.

Know exactly why you get filtered out.

Paste your CV and the job description. Get a score out of 100 in seconds — with every weakness ranked by impact, so you fix the right things first.

61%of CVs are rejected before a human ever reads them

ATS Compatibility

38/100
highMissing keywords from job description
highNon-standard section headers detected
mediumMulti-column layout may break parsing
lowContact info inside header block
Joined by 20,313+ professionals14-day free trialNo credit card required