There is a number in this year's job-search data that reads like a mistake.
Nearly a quarter of job seekers say they have significantly reduced the number of jobs they apply to. Not because they gave up. Because applying more stopped working, and they noticed before the advice did.
If you have been hunting this year, you already feel why. The median time from starting a search to a first offer is now past 100 days. Somewhere in there, most people cross a line where the search stops being a project and starts being a second job that pays nothing.
The maths that broke
The old advice was volume. Apply to everything, something lands, it is a numbers game.
That advice assumed each application costs you five minutes. It does not any more. Here is the honest cost of one real application in 2026:
- Read the posting properly. Four minutes.
- Adjust the CV so the first third matches what the role actually screens for. Ten minutes if you are fast.
- Upload it. Watch the parser eat it and hand back a form with your dates wrong. Fix it by hand. Twelve minutes, and this is the part nobody counts.
- Answer the extra questions. Notice period, salary expectation, work authorisation, the essay box. Six minutes.
- Press submit. Find out nothing.
That is over half an hour. Twenty applications is a working week. And a chunk of those postings are not real, which is its own tax on top.
So people did the sensible thing. They cut the number down and put the saved hours somewhere with a better return.
What the people cutting back are doing instead
Three things, and none of them are clever:
They apply to fewer, better-matched roles. Not "roles I could do" but "roles where the first screen will not reject me on the four things it checks": the title, the years, the location, and whether you can legally work there. Everything else is decided later by a human. If one of those four reads wrong, nothing you wrote below matters.
They spend the saved time on one warm route per week. One message to a person who actually works there beats six cold applications, and it takes less time than one of them. Not "networking". One message.
They keep a record of what they sent. Sounds like admin. It is the only way to tell a rejection from an application that quietly never arrived, and those two things look identical from where you are sitting: silence.
The part I find genuinely interesting
The volume advice never died because it was right. It died because volume got expensive. When an application cost five minutes, spraying was rational. At thirty-five minutes, it is the most expensive possible way to be ignored.
And the tooling response to that has mostly been to make spraying cheaper rather than to make it unnecessary, which is how you end up with a market where everyone applies more and nobody hears back sooner.
I build in this space, so treat that as an interest declared. The thing I would defend is narrower than a product pitch: the expensive part of an application is not the writing, it is the retyping and the silence afterwards. Fix those two and the volume question mostly answers itself.
If you are in it right now
Cut your list. Check the four screening facts before you spend thirty minutes on anything. Keep the record. Send one warm message a week.
And do not read your own silence as a verdict on you. Most of it is a req that got frozen, an offer waiting on a signature, or a form that failed.
What actually moved the needle in your search this year? I am collecting honest answers rather than advice, and the honest ones have been more useful.
I build AI Applyd - it fills in and submits job applications across the twelve systems companies actually use to receive them.
Top comments (0)