How to Write an ATS-Friendly Resume

Almost every mid-size and large employer runs applications through an applicant tracking system, or ATS. Around that fact has grown a small industry of myths: that the ATS scores your resume out of 100, that it silently rejects you if you miss a keyword, that a hidden white-text block will trick it into liking you.

The reality is duller and much more useful to know. An ATS is a database. Its job is to take your file, pull structured fields out of it — name, employer, job title, dates, skills — and store them so a recruiter can search and filter later. It is not judging your writing. It is transcribing it. Almost every ATS problem is a transcription problem.

What the ATS actually does with your file

When you upload a resume, the system runs a parser over it. The parser walks the document looking for text, then applies heuristics to guess what each chunk of text means. It looks for a line that resembles a job title, a nearby line that resembles a company, and a date range close to both. It looks for a heading it recognises — Experience, Education, Skills — to know which part of the document it is in.

Those heuristics are the whole story. When your resume parses cleanly, your work history lands in the right fields and a recruiter searching for “React” and “3+ years” finds you. When it parses badly, your experience section might be stored as one undifferentiated blob of text, your job titles might be missing, and you become invisible to exactly the searches you should show up in.

Nobody sends you a rejection for this. You simply never surface. That is why parsing quality matters more than any keyword trick.

The formatting choices that break parsing

In rough order of how much damage they do:

1. 1
   Text inside images or graphics

   A parser reads text, not pictures. If your name, headline, or skills sit inside an exported graphic, they do not exist as far as the ATS is concerned. This includes “infographic” resumes and skill ratings drawn as dots or bars — the bars carry the meaning, and the meaning is lost.

2. 2
   Multi-column layouts built with text boxes

   Two-column designs are not automatically a problem, but the way they are built usually is. When columns are separate text boxes, many parsers read them in creation order rather than visual order, interleaving your sidebar into the middle of a job description. A single-column layout, or a two-column layout built on a real grid, avoids the whole question.

3. 3
   Headers and footers

   Some parsers skip page headers and footers entirely. Putting your phone number and email there is a classic way to become uncontactable. Keep contact details in the body of the first page.

4. 4
   Tables for layout

   Tables used to arrange content — rather than to present actual tabular data — get flattened unpredictably. Cell contents can be concatenated in the wrong order.

5. 5
   Creative section headings

   “My Journey” instead of “Experience” and “What I Bring” instead of “Skills” defeat the heading detection the parser relies on. Boring headings are a feature.

6. 6
   Unusual characters and fonts

   Decorative bullet glyphs, ligature-heavy display fonts, and text set in icon fonts can come through as question marks or dropped characters. Standard bullets and a common sans or serif face are safe.

A structure that parses cleanly every time

There is nothing clever here, and that is the point. A resume in this shape parses correctly in essentially every system:

1. Contact block — name, then one line with email, phone, city, and links. In the body, not a header.
2. Summary — two or three lines positioning you for this specific role. Optional, but useful.
3. Skills — a short, grouped list of concrete tools and technologies. No ratings, no bars.
4. Experience — reverse chronological. Each entry: job title, company, location, date range, then bullets.
5. Education — degree, institution, graduation year.
6. Optional sections — certifications, projects, publications, volunteering, as relevant.

Keep one consistent date format throughout — Mar 2023 – Present or 03/2023 – Present, but not both. Parsers infer your tenure from these strings, and inconsistency is where they guess wrong.

Keywords: relevant, not stuffed

Recruiters search the ATS database using the language of the job description. If a posting says “Kubernetes” and your resume says “container orchestration”, a keyword search will not connect the two. So the words matter — but as accurate labels for things you have actually done, not as decoration.

The reliable method is to read the posting closely, list the hard requirements it names, and check that each one you genuinely have appears somewhere in your resume in the posting's own words. Where you have a real equivalent, name both: “container orchestration (Kubernetes)”. Where you do not have it, leave it out.

Put keywords where they carry evidence. A skill named in a bullet alongside a result is far more persuasive than the same word sitting in a comma-separated list. We go through this in detail in how to pull keywords out of a job description.

File format and file name

Send a PDF unless the posting or recruiter asks for something else. Modern parsers handle text-based PDFs well, and a PDF guarantees the layout a human sees matches what you designed. The exception is agency recruiters, who often want DOCX so they can edit and re-format before submitting you.

Two things to check: the PDF must contain real text, not a scan or an exported image — try selecting a line of it; and the file name should be readable, because a recruiter will see it. Ali-Humza-Resume.pdf beats resume_final_v3(1).pdf.

How to test your own resume in two minutes

You do not need a paid ATS scanner. Two crude checks catch the majority of real problems:

• The copy-paste test. Open your PDF, select all, and paste into a plain text editor. What you see is approximately what a parser sees. If sections are interleaved, dates are detached from their jobs, or whole chunks are missing, you have a parsing problem — fix the layout, not the wording.
• The six-second scan. Hand it to someone for six seconds, then ask what role you are applying for and what your strongest achievement is. If they cannot answer both, the hierarchy is wrong for humans too.

What actually gets you the interview

Clean parsing gets you into the database correctly. It does not get you called. Once a human is reading, the deciding factor is whether your bullets show specific, measurable outcomes in work that resembles the job on offer.

So treat ATS-friendliness as hygiene: get it right once, in a template you reuse, and stop thinking about it. Then put your real effort into writing bullets that carry evidence and tailoring the resume to each posting.