Every job application sounds the same now
Small employers are buried in polished, machine-written applications they can't tell apart — while scammers post fake jobs under real businesses' names. A careers page that asks the right things fixes more of both than you'd think.
Put a “help wanted” sign in the window of a café ten years ago and you got a handful of people walking in, a few emails, and a fair idea of who was serious.
Post the same job online now and you get a pile of applications that are all competent, all polite, all enthusiastic about your “dynamic team” — and all strangely alike. Same structure. Same adjectives. Same eagerness to “contribute to your continued success.” You read thirty and cannot remember any of them.
This is now the normal experience
A survey of more than 250 small business leaders in the US this spring found that 54% say machine-written applications have made hiring harder, and 47% say finding skilled people is harder than a year ago. The researchers put the difficulty down to an influx of homogeneous applications that are difficult to authenticate. A separate survey from the same firm found two-thirds of HR leaders saying these applications were slowing their hiring down.
The reason is simple. Applying used to cost an applicant effort, and effort was a filter. A tool can now produce a tailored cover letter for every job in the area in an evening. Your inbox is not measuring how many people want the job. It is measuring how easy it has become to apply.
The other half: someone may be hiring in your name
While employers drown in applications, job-seekers are being targeted with vacancies that don’t exist — and those fake jobs increasingly wear a real business’s name.
This summer a staffing agency in Minneapolis found fraudsters sending out job offers by text using its real logo, fronted by a friendly recruiter whose photo turned out, on forensic analysis, to be generated. The “recruiter” asked applicants for their social security numbers and bank details. The agency had to run a public awareness campaign to warn people away from itself.
US consumer-protection figures put impostor scams at $3.5 billion in losses last year, nearly triple what they were five years earlier, with around $1 billion involving people pretending to be a business. As one academic put it: generative tools have “lowered the cost of producing a convincing fake. Even a website can be put together in a matter of minutes.”
For a small employer, that is a reputational problem you didn’t cause and might not hear about until an upset stranger rings to ask why you wanted their bank details.
A fair word about applicants
Before going further: plenty of good candidates use these tools to tidy up a CV or get past the blank page, the same way people used to ask a friend to read their cover letter. That is not cheating, and penalising it would mean penalising the nervous and the dyslexic along with the lazy.
The problem is not that applications were helped. It is that they have become impossible to tell apart and impossible to check. So the fix is not to ban the tools. It is to ask for things the tools can’t supply.
What a tool can’t fake
A generator is good at fluent, general, plausible text. It is bad at things that require having actually been somewhere, done something, or turned up.
- Specifics about you. “Which of our two shops is closer to you, and have you been in?” There’s no plausible generic answer to that.
- Specifics about the work. Not “describe your customer service skills” but “a customer’s order is wrong and the queue is out the door — what do you do first?” in their own words, two sentences, no essay.
- A real-world step. A ten-minute phone call. Popping in on a weekday morning to say hello. A paid trial shift. The moment someone has to be a person, the pile sorts itself.
None of that is exotic. It is how small businesses hired for a hundred years. The tools have simply made the old ways valuable again.
The fair question
devkoi doesn’t screen candidates, run an applicant system or decide who to hire. It builds the page the job lives on.
That page turns out to matter twice. It is the one place you set the questions, in your own words. And it is the one genuine listing a suspicious job-seeker can find to check whether a vacancy “from you” is real — which only works if your own page exists, is current, and says plainly how you recruit.
What to do this week
1. Put every vacancy on your own website first. Job boards are fine, but your page is the original a job-seeker can check against. If it isn’t there, it isn’t real.
“Add a Jobs page with this vacancy: weekend barista, 16 hours a week, starting in November.”
2. Ask two questions only a real local applicant can answer well. Short ones. Specific to you, your place and the actual work.
“Ask applicants: which of our shops is nearest you, and what would you do first if the coffee machine broke during the morning rush?”
3. Add a human step early. A short call, a drop-in time, or a paid trial shift — said plainly on the page, so nobody feels tricked by it.
“Say we’ll invite shortlisted people for a paid two-hour trial shift.”
4. Publish how you recruit, and what you never do. This is the line that protects people from scams using your name:
“Add: we never recruit by text message or messaging apps, never charge fees, and never ask for bank details or ID numbers before you’ve met us in person.”
5. Take vacancies down the day they’re filled. A stale listing wastes applicants’ time and gives fraudsters a ready-made template to copy.
6. Don’t punish people for tidy writing. Judge on the specific answers and the human step, not on whether the cover letter sounds polished. Polished is free now. Specific isn’t.
The internet made applying for a job frictionless, and then generators made it effortless — so effort stopped meaning anything.
The fix is not a cleverer filter. It is asking, plainly and on your own page, for the small human things no tool can do on someone’s behalf: know where you are, say what they’d actually do, and turn up.