You spent an evening making your resume look sharp. Clean columns, a tidy skills sidebar, maybe a couple of small icons next to your email and phone. Then you upload it to a job portal and hear nothing back. Here is the uncomfortable part: before a single human looks at your resume, software reads it first, and that software does not see the document you designed. It sees a flattened stream of text. If the important parts get scrambled or dropped in that step, a recruiter may never see your best work at all.
So let us answer the real question. What does an ATS actually read from your resume, and how do you make sure it reads the right thing?
First, what an ATS actually is
An Applicant Tracking System, or ATS, is the software employers use to collect, store, search, and organize job applications. When you upload your resume, the system runs it through a parser. Parsing is the step that pulls your document apart and tries to sort it into fields like name, contact info, work history, education, and skills. Recruiters then search and filter those fields to find candidates.
The layout you crafted is not what gets stored. The parsed text is. And different companies use different systems, including Workday, Greenhouse, Lever, Taleo, iCIMS, and many others. They do not all parse the same way. That variation is exactly why this question is so slippery, and why nobody can honestly promise your resume will parse perfectly everywhere. What you can do is remove the things that reliably break parsers, so your resume reads cleanly across most of them.
What the parser reads, and what it throws away
Think of parsing as a translation from "how it looks" to "what it says." The parser walks the text content of your file in some reading order and tries to rebuild meaning from it. Plain, linear text in a logical top-to-bottom order translates cleanly. Anything that depends on visual position to make sense can get reordered or separated from the thing it belongs to.
The parser is not out to get you. It is just literal. It reads structure, not intention. So a date that sits to the right of a job title in a table cell, or a skill tucked into a sidebar, can end up detached from its context or dropped entirely.
The design choices that quietly break parsing
These are the usual suspects, roughly in order of how often they cause trouble.
- Multi-column layouts. The single biggest offender. A parser may read straight across both columns instead of down one and then the other, so your sidebar text lands in the middle of a job description. What looks balanced to you can read as noise to the software.
- Tables. Content inside table cells often loses the relationship between cells. Your dates, titles, and companies can get separated. If you used an invisible table to align things neatly, that alignment can work against you.
- Headers and footers. Text placed in the true document header or footer region (the part above and below the page margins) is skipped by some parsers. Putting your name, email, or phone only in the header is a real risk. Keep contact details in the main body.
- Icons and graphics. A little envelope icon next to your email adds nothing a parser can read, and text baked into an image cannot be extracted at all. If your name is part of a logo graphic, the software may not see your name.
- Text boxes and shapes. In word processors, text boxes often sit outside the main text flow. Some parsers ignore them completely, so anything important inside one can vanish.
- Creative section headings. Labeling your work history "Where I've Made an Impact" instead of "Experience" can confuse the mapping step. Conventional headings help the parser know what it is looking at.
- Unusual fonts and special characters. Rarely, certain fonts or ligatures cause character substitution when text is extracted, turning clean words into garbled glyphs. Stick to common, boring fonts and you avoid the whole category.
The 30-second X-ray test you can run right now
You do not have to guess. Open your resume, select everything (Cmd+A on Mac, Ctrl+A on Windows), copy it, and paste it into a plain text editor. Use TextEdit set to plain text on Mac, or Notepad on Windows. What you get is a rough approximation of what a parser extracts.
Now read it top to bottom. Is your name at the very top? Is your most recent job still attached to its dates and bullet points? Did your skills section land somewhere sensible, or did it crash into the middle of a job description? If it reads like nonsense to you, it may read like nonsense to the software too.
This "see the raw text" idea is exactly what Bounce built its free Beat the Bots scan around. Instead of eyeballing a plain text dump, you drop in your resume and it shows you the literal text a parser pulls out, the X-ray view of your document, so you can catch a jumbled column or a missing header before you apply. It is free, and you do not have to buy anything to use it.
How to build a resume the parser reads correctly
None of this requires design skill. It mostly means getting out of the parser's way.
- Use a single-column layout. One column, top to bottom. This alone fixes most parsing problems.
- Keep contact info in the body. Put your name, email, phone, and city in the normal text at the top, not in the document header.
- Use standard section headings. Experience, Education, Skills, Certifications. Plain labels the parser recognizes.
- Avoid tables, text boxes, and columns for anything that matters. If losing it would hurt, do not hide it in a structure a parser might skip.
- Keep icons and graphics decorative, never load-bearing. Never let an image be the only place a fact appears.
- Stick to common fonts. Arial, Calibri, Georgia, Helvetica, Times. Nothing exotic.
- Save as a Word document or a text-based PDF, never an image. A scanned or exported-as-picture resume has no text layer, so the parser gets nothing. A DOCX is the safest default when a job portal accepts it.
- Write dates and terms clearly. Use consistent date ranges, and spell out a term before you abbreviate it, for example "Search Engine Optimization (SEO)," so a search for either wins.
What an ATS does not do
A few myths are worth dropping, because fear leads people into bad habits. Most systems do not auto-reject you the instant one keyword is missing. A recruiter still runs the searches and reads the results. Hiding white keyword text or stuffing a hidden block is a bad idea: it is easy to spot once the text is extracted, and it reads as dishonest to the human who catches it. And parsing cleanly is not the same as getting the job. Clean parsing only earns you the thing you actually wanted, which is a fair read by a person.
The honest bottom line
An ATS reads a flattened, text-only version of your resume, and your job is to make that version say exactly what you meant it to say. Fix the layout, run the copy-paste test, and you have already done more than most applicants.
If you would rather not hand-tune all of this, Bounce Studio builds an ATS-ready resume for you using only your real experience, then tailors it to each job you apply for. It will not invent skills or tools you do not have, so the resume that parses cleanly is also one you can defend in the interview. That is the whole point. Because everyone bounces back, and it is easier when the software actually reads what you wrote.