Skip to content

BLOG · ATS

How to make an ATS-friendly résumé

What applicant tracking systems actually reject, which formatting advice is outdated, and how to write a résumé that parses cleanly and still reads well to a person.

8 min read · 19 August 2026

ATS 8 min read

Most advice about applicant tracking systems is a decade out of date, and some of it will actively make your résumé worse. This is what is still true, what stopped being true, and what to do.

The claim you should stop believing first

You have read that ATS "rejects 75% of résumés before a human sees them". That figure has no traceable source. It circulates because it is alarming, and it has convinced a lot of people to write for software instead of for the person doing the hiring.

What an ATS mostly does is parse and store. It reads your document into fields — name, employer, dates, titles, skills — and puts them in a database a recruiter searches. The rejection, when it happens, is usually a person filtering that database, or a knockout question you answered on the form. Not a robot deleting you.

This matters practically: if you assume the enemy is a parser, you optimise for keywords. If you assume it is a busy person searching a database, you optimise for being findable AND worth reading. The second is the one that works.

What actually breaks parsing

Parsing failures are real, they are just narrower than the folklore suggests. These are the ones that still bite:

  • Text inside images. A résumé exported as a picture, or a header rendered as a graphic, contains no text to read.
  • Multi-column layouts where reading order is ambiguous. Some parsers read straight across the page and interleave two columns into nonsense.
  • Contact details in the document header or footer. Several parsers skip those regions entirely, so your email disappears.
  • Tables used for layout rather than data. Nested tables in particular.
  • Dates written only as a range with no year, or in a format the parser cannot recognise.
  • Unusual section names. "Where I have worked" is charming and will not be recognised as Experience.

Notice what is not on that list: PDFs. Modern systems read PDFs fine. The advice to always submit .docx is left over from a time when they did not. Submit whatever the application asks for, and if it gives you a choice, PDF preserves your layout.

Structure that parses and reads

The dull structure is the right one. Not because software demands it, but because it is also what a person scanning for eight seconds needs.

  1. Name and contact details in the body of the document, at the top, not in a header.
  2. A short summary only if you have something specific to say. Three lines of "results-driven professional" is worse than nothing.
  3. Experience in reverse chronological order, with employer, title, and dates on their own line.
  4. Education, with the degree and institution recognisable.
  5. Skills, as a plain list. This is the section a recruiter search actually hits.

Standard section headings — Experience, Education, Skills. This is one place where being conventional is the whole point.

Keywords, without keyword stuffing

Recruiters search the database by skill. If the posting says "Kubernetes" and your résumé says "container orchestration", you do not appear. That is the real mechanism, and it is why matching the posting's vocabulary matters.

It is not a reason to paste the job description into white text at the bottom of the page. That trick is old, it is detected, and being caught doing it ends the application.

The honest version: use the words the posting uses, for the things you have actually done.

Instead of
Responsible for improving the deployment process

Write:

Moved deployments from manual releases to a Kubernetes pipeline,
cutting release time from two days to under an hour

The second version contains the keyword, the outcome and the number. It is better for the database search and better for the person reading it, which is the test any change should pass.

Where the score comes from

ORQILO reads your résumé across seven measured dimensions — section completeness, quantified impact, action verbs, active voice, readability, length and vocabulary variety — plus grammar. The measurement is deterministic: no model produces the number, so the same résumé scores the same way twice.

That is deliberately not a prediction of whether you will get an interview. Nothing can promise that, and a tool that claims to is selling you a number it invented.

A short checklist

  • Contact details in the body, not the header.
  • One column, unless you are certain the layout is unambiguous.
  • Standard section headings.
  • No text locked inside images.
  • The posting's vocabulary, used honestly.
  • A number in as many bullets as genuinely have one.
  • A file name that is your name, not "resume_final_v3".

If you want the underlying mechanics rather than the checklist, how applicant tracking systems actually work covers what happens to your document after you upload it.

ATS in ORQILO

What applicant tracking systems do, what they do not do, and which of the things people fear about them are actually true.

check your résumé against ATS with ORQILO