ATS in plain language
An applicant tracking system (ATS) is the software an employer uses to receive applications and manage them. When you apply online, your resume lands in one. The system stores your file, pulls its text into fields, and lets recruiters search, filter and move candidates through stages.
It is a database with a workflow on top, not a robot that reads your resume and decides. People choose which questions the form asks, which filters to use and whom to call. The system makes you easy or hard to find. A person still decides.
Common platforms include Workday, Greenhouse, Lever and iCIMS. They differ in detail, and each employer sets theirs up differently. Any advice that claims to know exactly how "the ATS" scores you is guessing.
What happens after you click apply
Most applications go through the same few steps, whatever the vendor.
- You submit a file and a form. The file is your resume. The form holds your contact details, a few questions and sometimes a cover letter.
- The system extracts the text. It reads your file as plain text, without the layout, and tries to sort it into fields: name, email, phone, employers, job titles, dates, education, skills.
- A profile is created. The recruiter usually sees both the parsed fields and your original file. Many portals also prefill the rest of the application from the parsed text, which is your chance to see what the parser read.
- Recruiters search and filter. They type terms such as a tool or a job title, filter by location or by answers to screening questions, and get a list of candidates.
- Some systems sort or score. Several platforms can rank applicants against the job. Whether that is switched on, and how much a recruiter relies on it, is the employer's choice and is rarely published.
- People move you through stages: applied, reviewed, interview, offer, rejected. Each move is usually a person's decision. The rejection email is often a template the system sends when someone closes your application, sometimes in bulk when the role is filled.
Where automatic rejection does happen
There is one place where the system can reject you without a person reading anything: screening questions, sometimes called knockout questions. The employer writes them and decides which answers disqualify.
Typical examples are "Are you legally authorized to work in this country?", "Do you hold a valid driver's license?" or "Can you work on site three days a week?". If the employer has marked "no" as disqualifying, a "no" can move you straight to rejected.
This filter works on your answers, not on your resume. No font, layout or keyword changes it. Two things follow. Read every screening question slowly, because a careless click costs more than any formatting choice. And answer truthfully. A false "yes" gets you past the filter and into a conversation where the real answer will come out.
A worked example: one application, three moments
This example is fictional. Daniel is applying for a Customer Success Manager role at a mid-sized software company. His resume is a two-column PDF made from a design template: his experience in the main column on the right, and a sidebar on the left with his contact details and a list of skills.
Moment one, the form. The application asks whether he can work from the Chicago office two days a week. He can, and he answers yes. In this case the employer has set "no" to disqualify, so this one click matters more than anything on his resume.
Moment two, the parse. The system pulls the text out of his PDF. It reads the sidebar first, line by line, so his skills end up mixed into his name and phone number, and the skills field in his profile stays empty. His job titles, employers and dates come through cleanly, because they sit on plain lines in the main column.
Daniel would never know this, except that the application form prefilled itself from his resume and left the skills box blank. That blank box is the parser telling him what it read.
Moment three, the search. A week later the recruiter searches the applicant pool for "renewals" and "Salesforce". "Renewals" appears in two of Daniel's bullets, so he shows up in the results. "Salesforce" was only in his sidebar. Whether the search finds it there depends on whether this system searches the whole file or only the parsed fields, and Daniel has no way to know which.
The recruiter opens his PDF, which looks fine to her, and starts reading. From this point the resume is judged by a person, on what it says.
What the example shows
Nothing in Daniel's story is a robot rejecting him. The screening question was the only automatic filter, and he passed it by answering truthfully. The parse made one part of his resume harder to find. The search found him anyway, through words he had used to describe real work.
The fix for the parse is small. He can move the skills into the main column, or, better, name each tool inside the bullet where he used it: "Ran quarterly renewal reviews with each of my accounts in Salesforce." The tool then appears in the text whatever the parser reads first, and the recruiter sees how he used it, not just that he lists it.
That is the usual pattern. What helps the parser rarely hurts the person reading, and what helps the person reading usually helps the parser.
Myths worth ignoring
You may have heard that "75% of resumes are rejected by ATS before a human sees them." That figure is widely repeated but not well sourced. It traces to old vendor marketing, not independent research. What is well documented is that large employers use ATS software heavily. What is not documented is a universal auto-reject rate.
"Stuffing keywords beats the system." A list of terms might put you in a search result. It does not survive the person who opens the file, and some systems now search by meaning as well as by exact words.
"Hide keywords in white text." Parsers drop colors. White text becomes ordinary text in the parsed profile, where a recruiter can see it sitting in the middle of your resume, making no sense.
"An ATS cannot read PDFs." Most current systems read text-based PDF and Word files. What they cannot read is text that is not really text, such as a scanned page or a picture of words. If an employer asks for a specific format, use it.
"You need a score of 80 to get through." There is no shared score across employers. Third-party match scores, including ours, measure your resume against one posting. They are not the number an employer's system uses, if it uses one at all.
What usually goes wrong
When an application stumbles at the system stage, the cause is usually one of these. None of them is about your experience.
- Layouts that scramble reading order. Columns, sidebars and text boxes can come out of the parser in the wrong sequence.
- Text inside images, icons or charts. A skill drawn as five dots, or a name set inside a logo, is not text.
- Contact details only in the header or footer. Some parsers skip those areas, so put your email and phone in the body of the page.
- Unusual section titles. "Where I've been" instead of "Experience" makes the parser guess.
- Tables used for layout. Cells can be read in an order that is not the one you see.
- A generic resume. If the words a recruiter searches for describe work you did but appear nowhere on the page, the search cannot find you.
- A careless screening answer. One wrong click on a knockout question outweighs everything else on this list.
What you can control
Use standard headings, a simple layout and the same terms the posting uses for tools and skills you genuinely have. Tailor for each role, so the words a recruiter is likely to search for are on the page wherever they are true.
Test your file before you send it. Open the PDF, select all the text, paste it into a plain text editor and read it from top to bottom. If it reads in the right order and your name, email and phone are there, a parser will most likely manage too. When a portal prefills a form from your resume, read the prefill. It is the parser's view of you.
And keep a sense of proportion. A match score against a specific job description is a fit check. It is useful for spotting missing keywords and weak structure, but it is not a promise of an interview. Treat it as feedback, then improve from your real experience.
How Verlane handles it
Verlane does not know how any particular employer has set up its system, and nobody outside that employer does. It works on the parts you control.
- The match score is a fixed formula, not a model's opinion. Most of it is keyword coverage: how many of the posting's key terms your resume uses, counting the usual other ways of saying the same thing. The rest is structure: contact details, a summary, dated roles, bullets, measurable results, skills and education.
- Matched and missing terms are listed, so you can see which words the posting repeats and your resume does not use. A missing term is a question about your experience, not an instruction to add it.
- The free health check includes an ATS file readability check. It reads your uploaded file the simple way a basic parser would, and reports whether your name, email, phone and the Experience, Education and Skills headings come through.
- It runs the same read on the PDF and Word files Verlane produces, so you can see whether the download is easier to read than your original.
- Sections you hide on the resume are left out of the score, because they will not be on the file you send.
What a score cannot tell you
None of this predicts an interview. It tells you whether your resume says what the job asks for, in a form a machine can read. That is the part of the process that belongs to you. The rest belongs to the people on the other side, and no tool can promise what they will decide.