The Gatekeepers & Applicant Tracking Systems

Note: Updated after a conversation with a recruiter. I'd oversimplified how auto-rejection works, and the nuance matters. The base ATS really is mostly a digital filing cabinet—but some companies now run AI tools on top of it that score and rank candidates before a human looks, and I've added sources on the lawsuits unfolding around exactly that. These features evolve constantly. This was originally posted in June 2026 and reflects where things stand now.

You’ve rewritten your resume. You’ve learned how to translate your production work into corporate language. You’ve included keywords from the job description.

You uploaded it to the online portal.

And then nothing.

You feel it in your gut: Your resume probably never got a real human review at all.

This may or may not be true. If you’ve applied at a mid-large size organization, how (or if) your resume lands with a human is first determined by software called an Applicant Tracking System (ATS).

This is the (not so) hidden gatekeeper of corporate job searching. And if you don’t understand how to translate your skills and present them in your application, even a perfectly good application hits roadblocks before any human even sees it.

When you apply to a job through a company website, LinkedIn portal, or any online application system, your resume probably doesn’t go directly to someone’s inbox.

It goes into an applicant tracking system. An ATS is software that companies use to manage hiring at scale. When I say “at scale”, I’m talking up to 50,000 applicants! They don’t just hold and organize these documents—they enable AI-powered screening, advanced analytics, automated communication, and they integrate with other programs that large businesses use to manage their workforce.

Think of an ATS like a scanning machine at airport security. It’s looking for very specific things in a very specific format. The machine has to be able to read you before you get where you need to go.

It’s collecting specific information about you. Usually 3-5 questions that are the first real deal-breakers.

Are you legally allowed to work in Canada? Are you PMP Certified? Etc.

It’s parsing your resume—pulling out details like your contact information, work history, skills, education. The extent to which this happens varies by software.

Does it have a clear work history section? Can it actually read your formatting, or does fancy design confuse the software?

It makes recruiters lives easier by presenting the information in a consistent way and creating a searchable database of all the applications. Imagine having 500 applications in your email inbox and trying to review and compare them. Each one has different formatting so you’re hunting for basic critical information. You remember a candidate you interviewed for a different job last month and think they might be a great fit for this new job you just posted. You’re searching your email but the details you are referencing in your search are in the attachment and won’t come up when you search your inbox. It’s a lot more work and takes a lot longer!

Organizations that use ATS platforms have one primary goal: help recruiters narrow their focus to the applications that are worth their time for detailed reviews.

Some companies layer AI scoring tools on top of their ATS that rank candidates before a human ever looks by matching the elements of your application against the job requirements. These are the features that give ATS the bad rap as some obstacle to beat or get beaten by.

I can’t definitively tell you how much this affects your odds (nobody can) but I can tell you it’s real enough to be argued in court. In Mobley v. Workday, a judge let age-bias claims proceed against Workday, finding its AI screening acted as the employer’s agent, not just software. In Kistler v. Eightfold, applicants allege the platform built profiles on over a billion workers and scored applicants zero to five on “likelihood of success” and fed those rankings to employers—Microsoft, Starbucks, and Chevron among the clients—before anyone reviewed a file. Both cases are unresolved; both companies deny wrongdoing. The point isn’t the verdict. It’s that “this isn’t really happening” as a counter argument is losing credibility.

So here’s where I’m landing:

Can your ATS “optimized” resume give you a fast pass to a job offer? No. Can a poorly formatted resume put your application at a disadvantage from the moment you hit “apply”? Yes.

To what degree depends on the ATS platform, how it’s configured, and the individual recruiter using it.

If the recruiter has configured filters to auto-disqualify candidates based on certain answers to the screening questions or other criteria, your application could be rejected without human review. If it can’t properly read your resume, the ATS may make the wrong determination.

If not automatically disqualified, your application may then be given an automated score or ranking. Again, the specifics vary by system and how it’s configured, but moral of the story is: you want a resume that lands you in the bucket that tells the recruiter you’re a good candidate. If the system cannot read your resume, you may find yourself in a bucket that tells a recruiter you’re not a good candidate, until proven otherwise.

Changing someone’s assumption from “probably no” to “yes” is a lot harder than getting from “probably yes” to “yes”. Imagine trying to convince your friends to watch the latest film you worked on. One of them has read press that the film is good. One of them has read press that the film is bad. The film is still the film. But there’s a bias from square one that’s harder to undo.

Your resume moves to the next stage—a human actually reads it.

This isn’t a slam dunk either. You can optimize your resume perfectly for ATS, use keywords from the job description, and still not get the job. Not because the system failed. Because you got through the filters by making yourself sound like everyone else.

I wrote about this a few months ago in Make the Protagonist More Likeable—what bad network notes and AI resumes have in common. If you missed that one earlier, pause here and come back after reading it.

In the rest of this article, I’ll take you through:

• Outdated advice that’s hurting your applications

• Practical steps to give the ATS what it needs

• How to Goldilocks your way to the right balance of keywords and specificity

  • The Keyword Strategy: How to get past the first gate

  • The Specificity Principle: How to Write Bullets that Survive Both Stages

• My ultimate tip that’s worth the extra effort

You’ll learn how to do more than just hit keywords. You’ll learn how to show why your specific background makes you valuable in a way that generic corporate experience doesn’t. Most of the details are for paid subscribers only. Upgrade to get instant access to the whole thing, the full translation system, and all the other benefits that the membership has to offer.

For years (and not that many years ago) the standard advice was: make your resume stand out. Use design. Show your creative sensibility. If you work in a visual industry, your resume should reflect that. Custom section headings. Interesting layouts. A PDF that looks like you understand what good taste looks like.

That advice made sense when a person was reading your resume first.

It now actively works against you.

An ATS is not reading your resume the way a human does. Interesting layouts can draw the human eye to read a document in a certain way. When an ATS is parsing your resume, it only reads linearly. For all of the marvels of AI (and the OCR technology that preceded it), it’s still really bad at reading tables or columns. Humans use tables and layouts to encode meaning through position, not just text—the full meaning only makes sense relative to what surrounds it. AI is good at extracting text but falters when it’s unable to preserve these two-dimensional relationships.

When it hits a table, it gets confused about column order. When it hits a text box, it skips it. When it hits a non-standard font, the extracted text can come out as scrambled characters. When your contact information is nested in a designed header or footer, some systems skip that section entirely.

So: the visually distinctive resume that used to help you stand out is now likely setting you back before your name is even read correctly.