7 ATS myths that make CVs worse — and what actually helps
· 7 min read · by the JobsFleet team
Do applicant tracking systems really reject 75% of CVs? Do two-column layouts always fail? What's true about ATS software, and a checklist that helps.
An applicant tracking system (ATS) is the software an employer uses to collect applications and move candidates through hiring — Workday, Greenhouse, Lever, iCIMS, SAP SuccessFactors, Oracle Taleo and many others. Candidates almost never see inside one, so advice about “beating the ATS” is often guesswork. Some of it makes CVs worse. Here's what's actually true.
Myth 1: “75% of CVs are rejected by an ATS before a human sees them”
This number is everywhere, but no study behind it has ever been published. It's usually traced back to marketing from a CV-optimisation company more than a decade ago.
What's true: most applications that get filtered out automatically fail knockout questions the employer set up — work authorisation, location, a required licence, notice period or salary range. Answer those accurately; they matter far more than your font.
Myth 2: “The ATS reads your CV and rejects it if the formatting is wrong”
An ATS parses your CV: it pulls out your name, contact details, jobs, dates, education and skills, and stores them as a profile. Recruiters then search, filter and read.
Bad parsing hurts in a quieter way: a missing email, a job with no dates, or skills that never made it into the profile make you harder to find and make your application look incomplete. That's worth fixing — but nothing is rejecting you for using bold text.
Myth 3: “Two columns, tables and icons always fail”
It depends on the parser and on your file. The real risk is reading order: a parser that reads straight across a two-column page can splice your sidebar into the middle of your job history. Text inside images (a skills chart, a logo with your name) is never read at all.
A quick test: open your PDF, select all the text, and paste it into a plain text editor. If it comes out in a sensible order with your contact details, titles, employers and dates intact, a parser can usually follow it too.
Myth 4: “Never send a PDF”
A text-based PDF — exported from Word, Google Docs or a CV builder — parses well in most modern systems. A scanned PDF is a picture of text, and without text recognition there's nothing to read. If a posting asks for a .docx, send a .docx.
Myth 5: “Every ATS gives your CV a score”
Many tracking systems don't rank candidates at all: recruiters search, filter and read. Some newer systems and job boards add AI matching or ranking, but there is no single, universal “ATS score” that decides your fate.
Any score a tool shows you — ours included — is an estimate of how well your CV's text covers a posting. It's useful for spotting gaps, not a pass mark.
Myth 6: “Hide keywords in white text”
Parsers extract hidden text just like visible text, so the recruiter who opens your parsed profile sees a strange block of keywords. Long keyword lists with no context read badly to the person who actually decides. Put the posting's terms where they're true, inside real bullet points about real work.
Myth 7: “Exact wording doesn't matter — the AI understands synonyms”
Partly true for newer search tools, but recruiters still type exact terms. If your CV says “Postgres” and they search “PostgreSQL”, use their spelling — when it's accurate. Spell an acronym out once: “search engine optimisation (SEO)”.
What actually helps
- Answer knockout questions accurately — they're where most automatic filtering happens.
- Use plain section headings: Experience, Education, Skills.
- Put your email and phone in the body of the page, not only in the header or footer — some parsers skip those.
- Give every job a clear title, employer and dates (month and year).
- Use the posting's wording for the skills you really have, in context.
- Send one clean, text-based file — PDF, or .docx if they ask for it.
- Tailor your CV to each job, but never add a skill you don't have. It gets found out in the interview.