What Is an ATS? How Resume Scoring & Keywords Work
If you've applied to more than a handful of jobs online, your resume has already been read by software before any human saw it. That software is called an ATS, or applicant tracking system, and understanding how it actually works is the difference between a resume that gets seen and one that quietly disappears.
What an ATS actually does
An ATS isn't one thing. It's a category of software (Workday, Greenhouse, Lever, iCIMS, and dozens of others) that companies use to collect, organize, and filter applications. Every ATS does three things: it parses your resume into structured data, it stores that data so recruiters can search and filter candidates, and in many cases, it scores or ranks candidates against the job posting before a person ever opens the file.
How resume parsing works (and why formatting breaks it)
Parsing is the part most candidates get wrong. The ATS extracts your resume's text and tries to sort it into fields: name, contact info, work history, education, skills. It does this using position on the page and simple pattern-matching, not real understanding.
That means:
- Tables, text boxes, and multi-column layouts often parse incorrectly or out of order, because the ATS reads left-to-right, top-to-bottom in a way that doesn't respect visual columns.
- Headers and footers are sometimes skipped entirely, which is a bad place to put your contact information.
- Images, icons, and graphics (including a skills-bar graphic showing "Excel: 80%") aren't read at all. The ATS sees nothing there.
- Unusual fonts or symbols used as bullet points can occasionally be parsed as garbled characters.
The safest formatting choice is a single-column layout, standard section headers ("Experience," "Education," "Skills"), and a common font. It's less visually interesting than a template with a sidebar and icons, but the ATS has to actually be able to read it first.
How ATS scoring and keyword matching work
Once your resume is parsed, many systems compare it against the job description, looking for overlap in skills, tools, job titles, and qualifications mentioned in the posting. This is where keyword matching comes in: if the job description says "Google Analytics" and your resume says "web analytics tools," a literal keyword match might miss the connection even though you mean the same thing.
This is also why tailoring your resume per application matters more than most people think. A single generic resume, however well-written, will consistently score lower against specific postings than one that mirrors the specific language of that job's description, not by lying, but by using the same terms for skills you genuinely have.
What this means for your resume
- Keep formatting simple: one column, standard headers, no tables or text boxes for anything you need read.
- Match the job description's actual terminology where it's honest to do so.
- Don't over-rely on scores alone. A resume that "beats the ATS" but reads awkwardly to a human recruiter afterward has just moved the problem downstream instead of solving it.
Once your formatting and keywords are solid, the next lever is how you actually write your bullet points; see how to quantify resume bullet points with real metrics.