I Sent 2,200 Job Applications and Got Two Interviews. Here’s the Reason Nobody Told Me.

Almost five years and more than 2,000 applications later, I finally learned and understood what was actually the root cause of all the rejections.

Back in 2022, I applied for more than 2,200 jobs, mostly through LinkedIn, one after another, over months.

I got two interviews.

Two.

At the time, I did what most people do when the silence gets loud: I blamed the market. I blamed the economy. I blamed LinkedIn and went so far as to delete my profile entirely.

I knew my skills and experience were aligned with the jobs I was applying for, but still I rewrote the thing a dozen times. Nothing changed.

What I didn’t know, and what almost nobody tells you when you’re job hunting, is that in most of those 2,200 cases, a human being never read my CV at all. Software did. And the software and I did not get along.

The machine that reads your CV before anyone else does

Most mid-size and large companies today use an Applicant Tracking System, or ATS, to manage incoming applications. It’s not optional infrastructure anymore; it’s the default. When you apply for a job at almost any company with a proper HR department, your CV doesn’t land on someone’s desk. It lands in a database, gets parsed by software, and gets scored, sorted, or filtered, before any human being ever opens it. A 2021 Harvard Business School study found that 94% of employers using these systems relied on them to rank or filter candidates for middle-skills roles, and 92% did the same for high-skills roles, during initial screening.

That’s not inherently sinister. Recruiters are drowning in applications, and some kind of triage is inevitable. The problem is what “parsed by software” actually means in practice, and how little most applicants know about it.

I didn’t know, for instance, that the two-column layout I was proud of, the one that made my CV look organised and modern, was actively working against me. Many parsers read left to right, top to bottom, in a single flow. A two-column layout can scramble the reading order completely, merging my most recent job title with a bullet point from an unrelated section, or splitting a sentence in half. To a human, my CV looked clean. To the software, it may have looked like nonsense.

I didn’t know that graphics, icons, and skill bars, the kind of elements a designer instinctively reaches for, are frequently invisible to a parser, or worse, interpreted as blank space where actual content should be.

I didn’t know that a name or job title placed inside a text box or a header/footer sometimes never gets extracted at all, because parsers are built to read the body of a document, not decorative containers.

None of this makes the format “bad” in a design sense. It made it bad in a very specific, very consequential sense: it was bad at getting past the first, invisible reader.

The part that actually surprised me

Here’s the thing I found most uncomfortable to sit with: none of this was a competence problem.

I wasn’t underqualified for the roles I applied to. My CV, read by a person, was fine, maybe even good. The problem was entirely translational. I was speaking a language the first reader didn’t understand, and I never found out until years later.

I think about this a lot now, because I don’t think I’m unusual. Most job seekers have no visibility into what happens to their application after they hit “Send.” There’s no error message. No “your CV couldn’t be parsed correctly.” Just silence, which gets quietly reinterpreted as rejection on merit.

What I did with that realisation

I’m not a recruiter or an HR technologist. I come from twenty years in media, content and communications. But the problem stuck with me long after I stopped needing an answer for myself, because it’s a genuinely solvable, well-defined problem: take a CV, understand how ATS software actually reads it, and fix the specific, mechanical things that get in the way, without inventing a single fact about the person who wrote it.

That’s eventually what became ATS.rip: a small tool that does one thing: it takes your CV, tells you what an ATS is likely to do with it, and restructures it so the content survives being read by software, before a human ever sees it. It doesn’t rewrite your career. It doesn’t invent experience you don’t have. It just makes sure the translation doesn’t get lost on the way in.

If you’re in the middle of a search right now

You don’t need my tool to benefit from knowing this. A few things worth checking in your own CV today, for free:

• Avoid multi-column layouts if you can. A single, linear column is far safer.
• Avoid putting your name, title, or contact details inside a header, footer, or text box.
• Skip graphical skill bars, icons as bullet points, and infographic-style CVs for ATS-heavy applications, even if they look great to a human.
• Save and re-open your CV as plain text at some point. If it reads as a jumbled mess, that’s roughly what a lot of parsers are seeing too.

I wish someone had told me this in 2022, before the two-thousandth application. It wouldn’t have changed how qualified I was. It would have changed how many of those applications a human being actually got to see.

Back in 2019, CNBC covered ATS basics, and it still holds up.

If you found this article useful or maybe even helpful, please support me by:

• 👏 Clapping for the story
• 🔔 Following me on: Medium and LinkedIn
• 🌐 ATS.rip is built by AIKON Media. If you know someone stuck in the same silence I was, feel free to pass it along.

I Sent 2,200 Job Applications and Got Two Interviews. Here’s the Reason Nobody Told Me. was originally published in Bootcamp on Medium, where people are continuing the conversation by highlighting and responding to this story.