I had one resume and zero replies after three weeks of applying. Not rejections. Silence. So I tried something I should have tried on day one: I split the search in half and ran an experiment.

Thirty applications. Fifteen with my original resume, fifteen with a rewritten one. Same person, same skills, same job boards, same two-week window. The only difference was how the resume was built.

Resume A was skills-first. A big skills section at the top, every tool I had ever touched, then my projects, then my work history at the bottom. It was the resume everyone tells you to write when you are early in your career: lead with what you can do.

Resume B was experience-first. Work history at the top, each role with two or three lines about what I actually did and what changed because of it. Skills shrank to a single line near the bottom. Projects stayed, but they moved below the jobs.

I applied to similar roles with both, junior developer and support engineer posts, mostly at small companies and startups, and I tracked every reply in a spreadsheet like a nerd. Here is what came back.

The numbers

Resume A, skills-first: 15 applications, 1 reply. A generic rejection that arrived 11 days later.

Resume B, experience-first: 15 applications, 6 replies. Four screening calls, one take-home task, one rejection with an actual human sentence in it.

That is 7 percent against 40 percent. Not close. And the replies to Resume B came faster too: the first one landed in 36 hours, and three of the six arrived within four days.

What the replies actually said

The replies to Resume B taught me more than the count did.

Two recruiters mentioned the same line: I had written that I cut our support ticket backlog from 120 open tickets to under 20 in six weeks. One of them quoted it back to me on the call and asked how I did it. Nobody has ever quoted my skills section back to me.

Another reply came from a hiring manager who said my resume "read like someone who had actually done the work." I laughed when I read that, because of course I had done the work. But that was the point: Resume A never gave him a chance to see it. It buried the work under a wall of tool names.

The single reply to Resume A was a template. "Thank you for your interest. We have decided to move forward with other candidates." Eleven days later. That is the sound of a resume that never got read by a person.

Why I think it worked

Resume A looked like every other junior resume. Python, JavaScript, Git, HTML, CSS, React, SQL, Docker, and so on down the page. A recruiter scanning that list learns nothing about me except that I know how to type tool names. Every applicant has the same list. It is the one part of a resume where being honest and being generic look identical.

Resume B gave a stranger something to picture. A ticket queue going down. A small internal tool I built that saved the team about four hours a week. Those are small numbers. They are also specific, and specific is the only thing that stands out when someone is reading their fortieth resume of the day.

The honest downside

I need to say the things that weaken my own result, because a 30-application experiment is easy to oversell.

First, the sample is small. One field, junior developer and support roles, mostly small companies. A big company with an automated screening system might read these two resumes completely differently, and I have no data on that.

Second, I applied to the Resume B jobs second, a week after the first batch. I was better at picking roles by then, and I wrote slightly better cover notes. That could account for some of the gap. Not all of it, I think, but some.

What I would do differently

But I am not running it again right now, because Resume B is doing its job. The lesson I took from it is simple, and it changed how I think about applying: a resume is not a list of what you know. It is a short story about what you did with what you know, told in numbers a stranger can picture.