Picture the listing: Junior Developer. Three to five years of experience. Strong React. Node.js. TypeScript. Familiarity with AWS. Docker preferred. Bonus points for ML exposure. Competitive salary. You have eight months of experience and have touched maybe half of those things. You close the tab.
Don't close the tab.
Job listings are not accurate descriptions of what a company needs. They are wish lists, usually written by a committee of people who haven't discussed priorities with each other, often by an HR team that's adapting a template from a hire they made three years ago. The actual hiring manager — the person whose team you'd join — frequently hasn't seen the listing before it goes live.
The Gap Between Listed and Wanted
What's actually on the listing is a combination of: things they genuinely need, things that would be nice to have, things a previous candidate had that impressed them once, and things someone added because they sound important. There's rarely a clear hierarchy.
What the hiring manager actually wants is almost always simpler: someone who can ship working software, communicate when they're stuck, and not introduce chaos into an already complicated system. The five-year experience requirement for a junior role exists because someone once got burned by a candidate who couldn't do those three things. It's a proxy measure, and a poor one.
Understanding this is not about tricking anyone. It's about not self-selecting out of roles that you'd be genuinely good at, because you took a wish list literally.
The 70% rule
If you meet 70% of listed requirements, apply. The 30% you don't is almost always learnable on the job, not actually required in practice, or something the team themselves will teach you. Waiting until you feel ready to check every box means waiting forever.
What Hiring Managers Actually Notice
Portfolio projects that do one thing well. Not five half-finished apps, not a clone of something famous — a single project where the scope was defined, the scope was hit, and you can clearly articulate what it does and why you built it that way.
Cover letters that feel like a human wrote them. The bar for this is genuinely low because most cover letters are generated with minimal thought. One paragraph, one specific connection to the company or role, one reason you'd do this job well. That's all.
GitHub history showing consistent work over months. Not a burst of activity before applying — a record of building things over time. Ten commits a week for six months is more impressive than two hundred commits in the two weeks before an application deadline.
The ability to explain what you built and why the alternatives would have been worse. This is the single most differentiating factor in technical interviews. Most candidates can describe what they built. Very few can articulate why they made the trade-offs they did. Prepare this for every project in your portfolio.
The One Question to Prepare For
"How do you use AI in your workflow?" You will be asked this in almost every technical interview right now. Most candidates either say "GitHub Copilot, sometimes" — which signals they haven't thought about it — or they lie, claiming they don't use it at all, which no one believes.
The right answer demonstrates that you have a genuine workflow, not just a habit. You know when AI speeds you up (boilerplate, documentation, first-pass implementations) and when it leads you wrong (novel logic, security-sensitive code, anything where you don't have enough context to verify the output). You've caught it making a mistake. You have a specific example.
That last part is important. Describe a specific moment when AI produced something wrong, explain how you caught it, and explain what that taught you about how to use the tool. This answer shows judgment. Judgment is what they're hiring for.
How to Write a Cover Letter That Gets Read
Not "I am passionate about software development and would love the opportunity to contribute to your team." That sentence says nothing and has been copy-pasted into approximately forty million cover letters.
- One paragraph only. Hiring managers read quickly.
- One specific thing about this company or this role — not the industry, not 'exciting growth', but something concrete you noticed about what they're building.
- One reason you'd do this specific job well — not that you're a hard worker, but a specific capability or experience that maps to something they need.
- No summary of your CV. They have your CV.
If you can't write that cover letter without researching the company first, good — that's the point. The research is what makes it feel human. And feeling human is the entire competitive advantage, because the templated AI-generated cover letter is the thing you're competing against.
The Numbers Game, Played Differently
Ten genuine applications will outperform one hundred templated ones. This runs counter to the instinct that job searching is a volume exercise, but the instinct is wrong. A templated application that passes screening has a conversion rate close to zero. A thoughtful application that shows evidence of judgment has a much higher chance of generating a conversation.
Identify companies you'd genuinely want to work at. Understand what they're building and why it matters. Apply with something that demonstrates you've thought about it. Follow up once, politely, after a week. Move on if you don't hear back. Start again.
“The best time to apply was before you felt ready. The second best time is now.”
The listing is a starting point for a conversation, not a qualification exam. Treat it that way, and you'll find that a lot more doors are open than they looked from the outside.