What an ATS actually does

An applicant tracking system is the software an employer uses to collect applications. It stores them, pulls your details into fields, lets a recruiter search and filter, and tracks where each candidate is in the process.

That is genuinely it. It is closer to a shared inbox with a spreadsheet attached than to anything that judges you. The scoring features that do exist are mostly configured by the employer, used unevenly, and invisible from outside.

Nobody outside a given company can tell you how their system ranks you. Any product that hands you an ATS score has invented a number that looks like one.

Almost every very large employer: 98.8% of the Fortune 500src. Further down the market it drops off fast, at 89% of large orgs, 35% of small onessrc.

So if you are applying to a fifteen-person company, there is a good chance a person opens your PDF and reads it. If you are applying to a bank, assume software touches it first. Both are normal, and neither requires you to write a different resume.

Each of these is repeated constantly, usually by someone with something to sell.

• Myth: three quarters of resumes are rejected by a robot before a human sees them. The figure behind that claim has never been traced to anything checkable. It gets repeated because it is frightening, and frightened people buy things.
• Myth: you must match the keywords exactly or be filtered out. Recruiters search; they do not usually auto-reject. Using their words helps you turn up in that search, which is a reason to be accurate, not a reason to stuff.
• Myth: any design at all breaks the parser. Modern parsers handle a sidebar, a colour and a table perfectly well. What they handle badly is text inside images and text in headers or footers.
• Myth: you need a special ATS resume separate from your real one. You need one good resume that happens to be machine-readable, which costs you nothing in how it looks to a person.

The real cost of these myths is not money, it is attention. Hours spent reformatting against an imaginary robot are hours not spent on the thing that actually moves your odds, which is aiming the application at the job.

Short list. All of it is free, and all of it takes one pass.

• Real, selectable text. If you cannot highlight a word in the PDF, neither can the parser.
• Nothing load-bearing in a header or footer. Your phone number belongs in the body.
• Standard section headings: Experience, Education, Skills. Parsers look for those exact words.
• Consistent date formats, on every role, in one style.
• Their words for the work, where the meaning is genuinely the same as yours.
• A file name with your actual name in it, because a person eventually saves it.

That is the whole list. Anything beyond it is optimising against a system you cannot see, and you would be guessing.

Machine-readability is a floor you step over once. The thing that actually changes your odds is what the resume says and who it is aimed at, and that is where the rest of your hour should go.

There is a guide for that, and the evidence behind it is here.

If you would rather have the floor checked for you than check it by hand, our ATS resume checker runs on every resume built here, free, and is equally direct about the one thing it refuses to guess at.