The number that means nothing on its own
“A hundred applications, no replies” is the most common sentence in job hunting and one of the least informative. It says how much effort went in and nothing about where it went out. A hundred applications to roles two levels above you produces the same silence as a hundred applications with a broken PDF attachment, and the two need opposite responses.
Every application passes through four gates: someone replies at all, that reply becomes a screening call, that call becomes a real interview round, and that round becomes an offer. Silence tells you that you did not reach the end. It does not tell you which gate closed.
So the first task is measurement, not improvement. Four numbers is enough.
Find the leak
Put in what actually happened. It names the earliest stage that is underperforming, which is the only one worth working on right now.
What each rate is telling you
These bands come from what a working search looks like, not from a promise. Markets differ, and a downturn moves every number down at once — what matters is which of yours is out of line with the others.
| Stage | Rough healthy range | If yours is far below |
|---|---|---|
| Applications → any reply | 10–20% | Targeting or timing. You are aiming at the wrong roles, or arriving too late. |
| Reply → screening call | 40–60% | Your first response is losing people your CV won. |
| Screen → interview round | 35–50% | The thirty-minute conversation. Almost always fixable by rehearsal. |
| Interview → offer | 15–30% | Depth. You can describe what you did, not why you did it that way. |
Cause one: you are applying to the wrong roles
This is the most common cause and the least popular explanation, because it implies the roles you want are not the roles you can currently get. A CV that is genuinely strong for a mid-level position produces total silence when it is sent to senior ones, and the silence looks identical.
The tell is a reply rate near zero across a large, varied set of applications. If forty applications across eight companies and three cities produced nothing at all, the constant is not the CV — it is the bracket you are aiming at.
The fix is unglamorous: drop one level of seniority, widen the titles you accept, or narrow the stack you claim. All three feel like a step back. All three produce replies within a fortnight.
Cause two: you are arriving late
Response rates fall off a cliff with the age of a posting. A role that has been open three weeks has a shortlist; you are applying to a process that has already moved on. On the aggregators this is systemic — a listing surfaced to you on Indeed or Naukri may have first appeared on the company’s own applicant tracking system a fortnight earlier.
This is invisible in your own numbers unless you record when the role was posted, which almost nobody does. It is also the single easiest thing to fix, because it costs nothing to change *when* you apply rather than *how*.
A quick test for lateness. Take your last twenty applications and split them by how old the posting was when you applied. If the ones under three days old replied and the older ones did not, you do not have a CV problem. You have a speed problem, and it has an entirely different solution.
Cause three: something is mechanically broken
Boring, and more common than anyone admits. A CV whose contact line is inside a header that the parser drops. A phone number missing a country code. An email address you no longer check. A PDF exported from a design tool as flattened image with no selectable text at all — perfectly readable to a human, entirely blank to the system reading it first.
The cost of checking is ten minutes. The cost of not checking is a search that cannot work no matter what else you improve.
The ten-minute mechanical check
Do all of these once. If any fails, nothing else in this guide matters yet.
6 steps
Cause four: volume, honestly
Sometimes every rate is inside the normal band and the total is still zero, because the total is small. At a fifteen per cent reply rate, thirty applications gives you between two and seven replies and could plausibly give you none. That is not a broken search. That is a small sample.
This is the diagnosis people most want to rule out and most often actually have. The counter-intuitive part is that the fix for it is the opposite of the fix for everything else: not more care per application, but more applications at the same care.
Which is where most searches break in practice. Doing this well takes hours a day that most people do not have while also holding down a job.
What to do this week
- Reconstruct the last thirty applications. Company, role, date applied, how old the posting was, and what came back. Your email sent folder has most of it. This is the hour that makes every other hour count.
- Run the four rates. Use the calculator above. Write down the earliest one that is below its band. That is your project for the next two weeks; the others are not.
- Do the mechanical check. Ten minutes, once. It is the only step that can invalidate everything else, so it goes early.
- Change exactly one thing. One variable, thirty applications, then measure again. Changing your CV, your targeting and your channels simultaneously means learning nothing from the result.
- Record as you go from now on. You are about to run an experiment. It only works if the next thirty are logged at the moment you send them, not reconstructed later.
The objections people raise here
Where to go next
- How to track job applications properly — The record this whole guide depends on.
- How many jobs should you apply to per day — If the diagnosis came back as volume.
- Where to find jobs beyond LinkedIn — If the diagnosis came back as timing.
Questions
Why am I getting no response to job applications?
Four causes produce the same silence: you are applying to roles above your current level, you are applying to postings that are already weeks old, something mechanical is broken such as a CV the system cannot read, or you have simply not sent enough applications for a normal reply rate to show up. They need different fixes, so identify which one applies before changing anything.
The fastest way to tell them apart is your reply rate across a varied set of at least fifteen applications. Near zero across varied roles points at targeting; fine on fresh postings and zero on old ones points at timing.
What is a normal response rate for job applications?
Roughly 10 to 20 per cent of applications get a reply of any kind in a functioning search. Of those replies, about half become a screening call, about 40 per cent of screens become a full interview round, and 15 to 30 per cent of interview processes end in an offer.
How many applications does it take to get one job?
Multiplying typical rates gives roughly one offer per 100 to 200 applications, though it varies enormously by field, seniority and market. The number matters less than the shape: if one stage is far below its band, fixing that stage moves the total far more than sending more applications.
Is my CV the problem if I get no replies?
Only if it is failing at the first gate, and only after you have ruled out targeting and timing. If you are getting replies but losing people afterwards, the CV is demonstrably working and the problem is later in the process.
How long should I wait before following up on an application?
Five to seven business days after applying. Early enough that the role is still being screened, late enough that you are not chasing before anyone has read it.
Does applying to more jobs actually help?
It helps when your conversion rates are already normal and your totals are just small — which is a genuinely common situation. It does not help when an earlier stage is broken, because more applications through a broken gate produce proportionally more silence.
How do I know if a job posting is too old to apply to?
Response rates drop sharply after about two weeks and again after a month. Postings under three days old are worth prioritising over anything else in your list, and a listing older than six weeks with no visible update is usually worth deprioritising rather than skipping entirely.