Employers: Cut 95% waste from your screening process without buying another tool.
Last week we were looking for a machine learning professional who could actually do DevOps. You’d think at this point, with all the modern tech in hiring, someone would have fixed the basics. But screening tools? Useless. LinkedIn’s hiring interface? Even dumber—“Pay us to flood your inbox with irrelevant and unqualified people who aren’t tailored for your problem.”
The people who came in? A cascade of wrong titles, wrong skills, and frankly, some overwhelmingly written by generative AI.
So the parade begins: we write the job description. See a few that seem promising, screening literally hundreds… interview a few—three or four of which didn’t even know what we do, hadn’t even skimmed our website, and were basically there to fulfill some weekly self-imposed “send out resumes” quota. In those cases? Out. I told them they were wasting our and their time and hung up in less than three minutes. No love lost.
On a whim, I run a little experiment. I posted a revised job. Seconds—literally, within seconds—my inbox lights up with two resumes with AI-generated cover letters. Would sound damn near the same, despite having different resumes. Both are so cookie-cutter that, if you swapped out my company for a fish cannery in Nebraska, nobody would’ve noticed. So, we’re PAYING for this. Shelling out real dollars to attract people who put in less effort applying than it takes to pick a bad Netflix show.
And then it hit me: I’m the idiot here. Time to flip the script.
I rewrite the job posting. This time, at the very bottom, after the requirements, I add: skip the fake cover letter. Go to our website. Learn a bit. If you’re genuinely interested, sign up for a trial and use the onboarding agent we just developed. Give me feedback on that agent—tell me what works, what doesn’t, and demo your skills and see if you want to work on this. If you can’t spare five minutes, neither of us needs to waste our time. I put a custom required question in the application that just says, “How many steps?” Those that did the work know; those that don’t can’t/won’t apply.
Results? Magic. The time-wasters, the serial appliers, and the GPT-powered resume writers poof—gone. I’m not sifting through hundreds of resumes, dodging the generative equivalent of junk mail. I get five applicants. Five. Every one of them knows who we are, what we do, and (holy hell) actually cares. Zero wasted dollars. Zero wasted time. Ninety-five percent less effort, and every single conversation is with someone motivated on both sides.
Here’s the real lesson: We can all learn from this, and I’m as guilty as anybody else. You have to use the AI. But the reward comes from investing in the relationship. Content is now free; time to prove you invested the time. Content alone, isn’t it… Quit worshipping the altar of more… More applicants, more tech, more “next-gen” filtering. What’s needed isn’t a shinier AI; it’s a smarter zoomed-in process. Lower the volume and raise the bar of who is worth your time. If you want people who are actually interested in you, stop incentivizing applicants to treat you like a lottery. Invest a bit more in your process. Ask them to invest a bit of energy up front—to repel the riff-raff, but not enough to kill genuine curiosity.
You want to fix hiring? Stop looking for the latest magic bullet. Zag when the world keeps zigging. Build a process where people actually have to care, or they don’t get in the door. That’s it. The rest is noise.