250 Applications, No Information

Monday morning. You open the inbox tied to the posting you published eight days ago. Two hundred and fifty applications.

You’ve read twelve. You know you won’t read the rest. And it nags at you a little — because somewhere in that pile there might be the right person, and the only reason you won’t see them is that they landed fortieth.

You’re not being negligent. You’re facing a problem of arithmetic.

It’s not your talent pool that grew

The reflex is to tell yourself the market turned, that people are searching more than before. The numbers say otherwise.

Workday measured a 31% rise in applications in the first half of 2024, while the number of posted jobs grew only 7%. That’s not the same thing. There aren’t more people looking: it’s the same people applying far more often.

The reason is simple. Applying no longer costs anything. It used to take an evening to tailor a résumé and write a cover letter; now it takes four minutes, and you can do it thirty times. LinkedIn now processes 11,000 applications a minute, up 45% in a year. The average posting receives about 250 applications, versus a hundred five years ago.

Your pile has doubled. Your talent pool hasn’t moved an inch.

What you were really sorting

Here’s the uncomfortable part. The résumé never measured anyone’s skills.

What it measured was effort. A polished file, a letter tailored to your company, vocabulary that echoed yours: none of that proved the person could do the job. It proved they wanted that one. It was a crude signal, but it was a signal — and you used it every day without naming it.

That signal is now worth zero.

Employers are already seeing it. A Robert Half survey of Canadian HR leaders in March 2026 reports that 61% of them saw their hiring process slow down because of reviewing AI-generated applications, and that 64% of recruiters feel the volume and uncertainty about authenticity create significant challenges. In other words: more files, all correct, and fewer ways to tell them apart.

The reflex trap

The instinctive response is to automate the sorting in turn. You filter by keywords, let a tool rank them, bring the pile down to a readable size.

That solves a volume problem, not an information problem. A tool sorting generated files looks for words in texts optimized to contain those very words. You don’t get the best candidates: you get the ones whose vocabulary the tool guessed best.

And that race, you won’t win. It costs your candidates almost nothing and you a great deal.

What hasn’t been devalued

So what’s left, when the written file no longer sorts?

What can’t be generated in four minutes. The way someone explains what they’ve done. What they choose to talk about first when you give them the floor. Their energy, their pace, the way they structure an idea with no text in front of them. The question they ask about your company that shows they actually looked.

None of that reads. It’s heard and it’s watched — and two minutes are often enough.

The shift to make isn’t to sort faster. It’s to change what you put first in the process. As long as the entry point is a document, you’ll keep receiving perfect documents that teach you nothing.

You’re not behind

You won’t read your 250 applications. No one will, and it’s no longer a flaw in your method. It’s the normal consequence of a tool that has stopped working.

The right question is no longer “how do I get through the pile.” It’s become: what do I want to see first?