How an ATS parses your resume, and how to make it readable

An applicant tracking system, or ATS, is the software an employer uses to publish its jobs, take applications and move candidates through its hiring steps. Workday, Greenhouse, Lever, iCIMS, Ashby and SmartRecruiters are some of the common ones. When you press Apply on a company's careers page, the form you fill in usually belongs to one of them.

Your resume does two things inside it. The file itself is kept and shown to whoever reviews your application. And the text in it is read by a parser, which turns it into structured fields: your name, email and phone, each job with its employer, title and dates, each school and degree, a list of skills. Those fields are what the recruiter searches and filters on, and what some application forms fill in for you.

1. Text extraction. The parser takes the text out of the file. A PDF made from a word processor has a text layer it can read. A scanned page or a PDF exported as an image has none, and the parser gets nothing or a poor guess from character recognition.
2. Reading order. The text comes out in the order the file stores it, which is not always the order you see. In a two-column layout the left and right columns can be read line by line across, mixing a job title from one column with a skill from the other.
3. Sections. The parser looks for headings it knows (Experience, Work Experience, Education, Skills, Summary, Projects) and files what follows under each. A heading it does not know, such as “Where I've made an impact”, leaves the lines under it without a home.
4. Entries. Inside Experience it looks for the pattern of a job: an employer, a title, a date range, then lines of work. Inside Education, a school, a degree and dates.
5. Fields. The result is stored as the candidate's profile in the ATS, beside the original file.

Most parsing problems come from a handful of layout choices.

None of this means a designed resume is always rejected. It means the structured profile built from it may be wrong or empty, and a recruiter searching that profile for a skill will not find you.

• One column, read top to bottom.
• Your name and contact details in the body of the page, at the top, as plain text: email, phone, city, and a LinkedIn or portfolio link.
• Standard headings: Summary, Experience, Education, Skills, and Projects or Certifications if you have them.
• Each job in the same order: title, employer, place, dates, then its lines.
• Dates with months, in one format throughout: “Mar 2021 – Jun 2024”.
• Plain bullets, and no tables, text boxes, icons, images or charts.
• A common font at a readable size, and a PDF with a real text layer, or a Word file if the posting asks for one.
• The right paper size: US Letter for jobs in the US and Canada, A4 elsewhere, so the file prints as the reader expects.

A quick check: open your PDF, select all the text and paste it into a plain text editor. If what you see is in the right order, with your contact details at the top and each job's dates beside it, a parser will most likely read it the same way.

Many application forms offer to read your resume and fill in the fields for you. SmartRecruiters builds your experience and education entries from the file you upload. Workday offers to autofill its form from a resume, or to apply manually. Some other systems tick a “fill out the application with my resume” box by default.

This is where a parse problem becomes your problem. A misread resume can put a job title in the employer field, merge two jobs into one, or drop a degree, and it can write over fields you had already filled in. Before you go on to the next page, read every entry the form made from your file. Our guides to applying on Workday and SmartRecruiters walk through those forms page by page.

A common claim is that most resumes are thrown out by an ATS before a person sees them. We have not seen evidence that supports a figure like that, and the systems named above are built for people to review applicants. What can happen is quieter: a recruiter searches the candidate pool for a skill or title, and a resume whose text did not parse, or does not name that skill, does not come up. Some employers also add their own knockout questions, such as the right to work in a country, and an answer there can end an application whatever the resume says.

So the goal is plain: a file whose text comes out whole and in order, under headings a machine recognises, naming the skills the posting asks for in the posting's words. The guide to resume keywords covers that last part.

Every resume Jobapplier makes follows the layout above: one column, standard headings, dates with months, and no tables or icons, as a PDF with a text layer. It is printed on US Letter for a posting in the US or Canada and A4 everywhere else, and every file is named Resume.pdf, so its name tells the employer nothing about which version it is. Technologies are bold inside the lines that used them and listed once under Skills, which keeps a separate “Tech stack” line out of each role.

Your master resume also gets a health score, out of 100 with a grade from A to F, from checks a parser cares about: whether an uploaded file has a readable text layer, your contact details, standard headings, dates with months, length and the shape of your bullets. Each failed check says what cost points.

When you apply with the browser extension, “Apply with tailored resume” attaches that posting's own resume on every page of the form, in place of a file the form already holds. On forms that read a resume and fill fields from it, the extension fills the fields from your profile, and you check them, as with any field, before you press Submit yourself. See the browser extension and resume tailoring for more.