Not sure whether it's the ATS, your resume, or something else keeping you out of the interview room? Book a free strategy call and we'll tell you honestly.
You've rebuilt your resume twice. You've tailored it to the ad. You've applied at 6am because somebody said early applications get read first. You've switched from PDF to Word and back again. You're doing everything you were told to do.
And nothing comes back.
So somebody tells you it's the ATS. That a robot is shredding your resume before a human sees it.
That explanation is wrong. And believing it costs you months.
The number everyone repeats
You've seen it: 75% of resumes are rejected by an ATS before a human ever reads them.
Career blogs. LinkedIn posts. YouTube. The sales page of every resume service going.
There is no research behind it.
Researchers went looking for the source. They traced it to a company that no longer exists. No study. No data. Nothing.
A 2025 study interviewed 25 recruiters across tech, healthcare and finance, working on Workday, iCIMS, Greenhouse, Bullhorn and SuccessFactors. Two of the twenty-five had set their system to auto-reject on content.
Two.
of the recruiters interviewed had not set their ATS to auto-reject on content. They read applications manually, or used simple eligibility filters instead.
The same study asked recruiters where they'd first heard the 75% claim. Sixty-eight percent said job seekers on social media. Another 20% said career coaches and resume services repeating old advice.
So the number circulates between the people it frightens and the people selling the fix. That is the entire ecosystem.
What actually happens when you hit apply
Most applications here start on SEEK. Some on Indeed, Jora, CareerOne, or APSJobs for government roles. From there they land in whatever system the employer runs: PageUp, JobAdder, Workday, Bullhorn.
Then three things happen:
- The system parses your file into fields. Name, contact, employers, dates, education, skills.
- It may score you against the job requirements.
- A recruiter opens a list and starts reading.
Step three is the one nobody mentions.
The software is a filing cabinet with a search box. Not a bouncer. The "T" stands for tracking.
The one filter that is actually automatic
This is the part that matters if you're on a visa, and I almost never see it written about properly.
Knockout questions are not content screening. They're the eligibility questions you answer before you upload anything. Do you have the right to work in Australia. Do you hold a driver's licence. Will you consent to a police check.
Every recruiter in that study used them. All 25.
This filter is real. It is automatic. It runs before your resume exists to anyone.
Which means the highest-stakes moment in your application is often not your resume. It's a dropdown.
And people get it wrong constantly. Not because they're careless. Because the question is ambiguous and the stakes aren't obvious. "Do you require sponsorship?" means one thing to a recruiter and something else entirely to someone who has never been asked it before.
Get it wrong and you're gone. Not ranked low. Gone. No formatting fix recovers it.
So: know your own work rights before you answer. Check your conditions directly with the Department of Home Affairs, or with a registered migration agent, not a forum, not a friend, not us. Then state them plainly. If there's a free-text field, one clear line does it.
Ambiguity reads as complication. Complication gets filtered.
What actually breaks the parsing
The formatting advice isn't useless. It's just far smaller than people sell it as.
Parsing genuinely fails on these:
| What breaks it | Why | Instead |
|---|---|---|
| Tables and text boxes | Content reads out of order, or drops entirely | Plain paragraphs, simple bullets |
| Headers and footers | Many parsers skip them. Your phone number vanishes | Contact details in the body, at the top |
| Two-column layouts | Text interleaves into nonsense | Single column, top to bottom |
| A PDF exported as an image | No text to read. Just a picture of text | Text-based PDF, or .docx |
| Graphics, icons, skill bars | Carry no readable information | Write the skill and the evidence for it |
| Creative section headings | "My Journey" maps to no known field | Work Experience. Education. Skills. |
Ten minutes. Then stop thinking about it.
What you should not do is spend your evening stuffing keywords to chase a "78% match score" from a free scanner. Those scores are invented by the tools that display them. No employer sees that number. It does not exist anywhere in the hiring process.
There's more on structure in our Australian resume format guide.
So why does nobody reply?
Volume.
One in-demand role pulls 400 to 2,000 applications in days. A recruiter with six roles open is not reading yours carefully. They're scanning, filtering on two or three criteria, shortlisting from the top.
Your resume was probably seen. For about eight seconds. Next to four hundred others. And nothing in those eight seconds made anyone stop.
That is not a parsing problem. That is a competition problem, and it needs a completely different fix.
More applications won't solve it. Four hundred people are already doing that.
The 15-minute habit
Inside the application channel, this is the highest-return thing you can do:
- Read the ad. Note the five or six capabilities it names most often.
- Check each one appears in your resume, in the employer's words, not yours.
- Move the most relevant role or project higher up the page.
- Rewrite your top line so it says what you are, not what you're looking for.
Fifteen minutes an application. Which means fewer applications, done properly, instead of forty in an afternoon.
Most people hate that trade. It feels less productive.
It isn't.
Questions I get asked
Does the ATS reject my resume automatically?
Usually no. Research from 2025 found 92% of recruiters do not enable content-based auto-rejection. Eligibility questions about work rights or licences are the genuine exception, and they run before your resume is read.
Should I submit a PDF or a Word document?
Either, as long as the PDF contains real text rather than being an exported image. If the job ad specifies a format, follow it.
Do free ATS scanners work?
They'll catch genuine parsing problems, which is useful. Ignore the percentage score. It's not a number any employer sees.
Do keywords still matter?
Yes, but as a matching signal, not a magic trick. Use the employer's terminology where it honestly describes what you did. Stuffing keywords you can't defend in an interview costs you at the next stage.
Is this different for smaller Australian employers?
Often there's no ATS at all. Applications land in an inbox and a person opens them. All the more reason to make the first few lines count.
The honest summary
The ATS is not why you're not getting interviews.
Fix the parsing basics: ten minutes, done. Know your work rights and answer that question clearly, because that filter is real.
Then stop optimising for software and start on the actual problem: how do you reach the person making the decision, instead of queuing behind four hundred applications that all look like yours?
That's the harder question. It's also most of what we work on.
Stuck in the application black hole? Book a free strategy call. We'll look at where you're actually stuck and tell you honestly whether we can help.
See how the program works →Sources
- Enhancv, ATS Rejection Study, 25 recruiter interviews, September–October 2025, reported via HR.com and HR Gazette
- The Interview Guys, analysis of ATS auto-rejection claims, 2025
- CVMate Australia, Australian resume format and ATS platform overview, 2026
This article covers general job search practice. Ark Abroad is a career accelerator, not a migration agent. For anything about your visa conditions, work rights or sponsorship eligibility, check the Department of Home Affairs directly or speak to a registered migration agent.
Book a free strategy call
We'll look at where your applications are actually getting lost, and tell you honestly whether it's the ATS, your resume, or the channel you're applying through.
