How to Apply for Software Engineering Jobs

The number nobody tells you

A reasonable application to job ratio in a normal market is somewhere around one interview per ten to fifteen applications, and one offer per five to ten interviews. Work that through and a job search is often a hundred applications.

That number is worth knowing before you start, because the alternative is applying to twelve places, hearing nothing, and concluding something is wrong with you. Nothing is wrong with you. The process has a low signal to noise ratio in both directions, and volume is part of how you get through it.

What follows is how to make each application cost less and land better.

Fix the resume before you send anything

Almost every engineering resume has the same defect: it lists what you were responsible for instead of what happened.

"Worked on the payments service" tells a reader nothing. Every engineer at the company worked on something. Compare it with "cut payment failures by a third by making the retry path idempotent". Same job, same person, completely different signal. The second one says you diagnosed a problem, chose an approach, and it worked.

Go through every bullet and ask what changed because you were there. If nothing did, either find the thing that did or cut the line.

Other rules that matter more than they should:

• One page until you have roughly ten years of experience. Two is acceptable after that. Three is never.
• No skills bar charts. Nobody believes you are seven out of ten at Python, and it wastes the most valuable space on the page.
• Plain formatting. Single column, standard fonts, no tables, no text inside images. Automated parsers mangle clever layouts and a mangled resume gets filtered before a person sees it.
• Put the interesting thing first. A reader gives you a few seconds. If your best work is at the bottom, it does not exist.

Write it once, properly, then adjust the top few lines per application rather than rewriting the whole document each time.

Referrals outperform everything else

An application through a company's careers page enters a queue with several hundred others. A referral goes to a recruiter with a name attached and a reason to look. The difference in response rate is not marginal.

This makes people uncomfortable because it sounds like you need to know important people. You do not. What works is much more ordinary:

• Ask people you have already worked with. This is the highest yield and the most underused, because it feels like an imposition. It is not; most referral schemes pay a bonus, so you are offering something.
• Ask people you have interacted with genuinely online. Not a cold template. Someone whose work you have actually engaged with.
• If you must approach a stranger, be specific and brief: the role, why you are a plausible fit in two sentences, and an offer to send your resume. Make it easy to say yes and easy to ignore.

The reason to publish work publicly, write about what you are learning, and contribute to things is largely this. It converts strangers into people who have a reason to answer you.

Apply to the right number of the right places

Two failure modes here, at opposite ends.

The first is the spray: two hundred one click applications with an unedited resume. Response rates collapse and you burn out around week three with nothing to show.

The second is the shrine: eight applications to companies you have researched for a fortnight each, all of which reject you, after which you have no pipeline and no data.

What works is a middle band. Something like twenty to forty live applications, each with a resume adjusted at the top and a genuine reason you are applying. Enough volume that rejections do not end the search, enough care that each one has a real chance.

Sort them into rough tiers and apply to the lower tier first. Your interview performance improves substantially over the first few, and you would rather that improvement happen somewhere it costs you less.

Cover letters, and when they matter

For a large company with a structured process, largely ignored. For a small company where a founder or an engineering manager reads applications directly, they can decide the outcome.

If you write one, three short paragraphs: why this company specifically and not a generic compliment, the one thing you have done that is most relevant to what they are asking for, and what you want. If you cannot write the first paragraph without a template, that is useful information about whether you want the job.

Track it, or you will lose it

By week three you will not remember which company asked for a take home, who you spoke to, or which of two similar roles you already applied to. A spreadsheet is enough: company, role, date applied, source, stage, next action, notes.

The tracking is not administrative tidiness. It gives you the two numbers that tell you what to change. If very few applications become interviews, the problem is upstream, in the resume or the targeting. If interviews rarely become offers, the problem is preparation. Those are different problems with different fixes, and without the numbers people usually guess wrong and fix neither.

Timing, and the follow up

Apply within the first week of a posting where you can. Many roles fill from the earliest applicants and later ones enter a queue that was already long.

Follow up once, roughly a week after an interview if you have not heard back, briefly and without apology. Once. A second follow up does not help and is remembered.

Silence after an interview is usually process and not judgement. Companies go quiet because a hiring manager is on leave or a headcount decision is stuck, and they very rarely tell you that. Assume nothing until you have a decision, and keep applying while you wait, because a pipeline that stops moving is how a search stretches from two months to six.

Handling the rejections

You will get many, most without explanation, some after you were sure it went well. Some will be for reasons that have nothing to do with you: an internal candidate, a frozen budget, a role that changed shape.

Two things help. First, treat the pipeline as the unit rather than the application. One rejection is noise; the pattern across twenty tells you something. Second, ask for feedback once, politely, and expect not to get it. When it does come it is worth a great deal.

What does not help is rewriting your entire approach after each rejection. Change one thing, send another ten applications, and see whether the number moved.

The offer is not the end of the process

When an offer arrives, the reflex is relief followed by acceptance. Wait.

Ask for the details in writing. Ask for time; a few days is normal and nobody rescinds an offer over it. And negotiate, because the number is almost always somewhat flexible, recruiters expect the conversation, and this figure anchors your next several years including your next offer somewhere else.

The most common expensive mistake in a whole job search happens in the ten minutes after the good news, when someone says yes to the first number out of gratitude.

The short version

Fix the resume so every line says what changed. Get referrals wherever you can. Keep twenty to forty applications live. Track it so you know whether the problem is the resume or the interview. Follow up once. Expect a lot of silence. Negotiate at the end.

It is a slower and more mechanical process than anyone wants it to be, and the people who do well at it are usually not the strongest engineers in the pool. They are the ones who kept a pipeline moving while feeling discouraged.