The Hiring Market Isn’t Broken. It’s Working Perfectly (And That’s Terrible)
Everyone says the hiring market is broken. In reality, it’s simply doing exactly what is required of it.
Over the last few weeks and months, I’ve read a hell of a lot of articles, posts, and comments about the modern hiring market. Recruiters, job seekers, businesses—everyone is howling at the ceiling in unison that the hiring market is broken. And what pisses me off the most is that everyone is howling from their own sandbox, without even trying to lift their head and look at the problem as a whole. Okay, not everyone is howling. Some are preaching about “how to do it right.” But again, from their own little house.
Candidates say that finding a job is impossible, and their resumes are simply ignored. Recruiters write sad posts about candidates flooding them with flat and completely identical resumes. Business suffers because they have to hire completely compromise people, who later cause more problems than they solve.
So, here’s the thing. The market isn’t broken. It works exactly the same way it worked before. It’s just that in the last year or two, two key events occurred that made the problem glaringly obvious. First, the market shifted, and now dozens, if not hundreds, of candidates are hunting for every single vacancy. And second, Artificial Intelligence has entered widespread use.
How, in general, does hiring work? A company needs an employee for a specific position. A job description is written, detailing duties and required skills. The description is given to a recruiter, and they start selecting a shortlist of suitable candidates to speak with in more detail. And this works quite well when there are a dozen or two people on the list. And it works quite well when recruiters are proactively sourcing candidates by profile. But everything breaks when the scale of the problem grows exponentially.
We have crashed loudly into an era where not 20, but 200 resumes arrive for a single vacancy. And a recruiter usually has not just one active vacancy, but several. This is literally thousands of candidates. And manually processing this squall of information with any quality is absolutely and categorically unrealistic. You have to filter. Filter fast, as fast as possible. Manually—by quickly skimming diagonally with your eyes, snatching keywords out of the text. When a human stops coping, the same thing happens, but automatically. In more advanced systems, the filters are more complex; in the coolest ones, they use AI. As a result, the recruiter gets a list of a more-or-less sane length, while the vast majority of candidates are never dignified with human attention, receiving at best a template rejection, and at worst—silence.
Upon receiving yet another rejection, the candidate starts wondering, “What exactly am I doing wrong?”, looks for information, and logically arrives at resume optimisation. A simple, general-purpose resume no longer works. It has to pass the filters. Highlight keywords so they are visible during a quick eye-scan. There shouldn’t be too few or too many keywords—otherwise, the automated system decides it’s either a lack of experience or spam. Format it so that, God forbid, you don’t break sensitive filters. Use AI to tailor the resume to the job description.
And then everything unfolds quite predictably. Many rejections — candidates optimise resumes — more resumes pass filters — recruiters see “spam” — filters become stricter — more rejections — resumes are optimised even harder. The spiral tightens even further.
We discussed a very vivid analogy in the comments recently. Imagine a cooking competition. There are rules: you must cook only using eggs and milk, use only salt for spices, and use only a frying pan as a tool. Obviously, if there are 100 chefs at this competition, we will likely get a hundred versions of an omelette, and 95 of them will be practically indistinguishable from one another. Well, maybe someone brave will just make a fried egg.
Are the chefs at this competition bad? No, they aren’t bad; they are cooking. Some better, some worse, but more or less everyone handles the task. And the judges? No, they are judging by the rules that exist. And if someone uses a microwave (the AI analogue) and cooks faster and more than others? Also, not a problem, if theirs turns out better than the rest. But what about the rules? If we tighten the rules that strictly, then getting, for example, a fried chicken will be impossible no matter what. Even though it would seem—you have a pan, you have eggs—that’s almost the same thing as a chicken…
And as a result, we get exactly what we complain about. Candidates can’t find work, recruiters complain about spam, and businesses get a stream of compromise candidates who seem to fit the job description but are somehow always “not quite right"—they have to hire not those who fit best, but someone whose resume fits into the Procrustean bed of the filters. There are many candidates, but those who make it to the interviews are very identical. Roughly the same level, not bad and not good. Average.
And no, this didn’t happen suddenly when AI appeared. The market has been playing by these rules for many years, but what changed is the large number of available candidates and the cost of optimisation dropping by orders of magnitude. And neither fancier ATS filters nor magic prompts for resume optimisation will solve this problem.
And the scariest part is that every step in this hell is absolutely logical.
