Why Mass Applications are a Logical Response to a Broken System

Overworked recruiters are writing desperate posts: “Applying more isn’t the solution.”

Career consultants advise applying sparingly but strategically, rather than spamming every vacancy in sight.

This advice comes in a market where vacancies are overwhelmed with applications, filters have tightened, and resumes have become increasingly standardised. Let’s examine whether it actually works in reality.

Take a simplified model. A company posts a vacancy. It receives 100 applications. The recruiter somehow forms a shortlist of 10 people, interviews them, and hires one.

In this model, each candidate has a 10% chance of passing the initial filter and getting to an in-person interview, and accordingly, a 90% chance of getting rejected without an interview.

If a candidate applies a second time to another vacancy under similar conditions, the probability of immediate rejection drops to 81%, and the probability of at least one interview invitation rises to 19%. It’s not intuitive, but that’s exactly how probability theory works.

It’s important to state the limitations of this model. We’re assuming independent attempts and a fixed probability of passing the filter. In reality, this isn’t the case: the market is segmented, attempts correlate with each other, and probability depends on the resume and context.

So, with further applications, volume starts working in the candidate’s favour. Already at 50 applications, the probability of at least one in-person interview exceeds 99%.

But there’s an important point. The mass flow of applications and the tightening of filters on both sides systemically push this probability toward a quasi-constant. As resumes get optimised, differences between candidates at the initial filter level are erased. Requirements get averaged out, and format becomes standardised. Therefore, even this crude model provides a lower estimate of the volume effect.

And what do we see? A large number of applications genuinely increases the probability of passing the initial filter. This directly explains why recruiters receive hundreds and thousands of applications. Volume is rational. “Applying more” actually works in this system, whether anyone likes it or not.

Let’s note one more thing. We’re not talking about hiring or candidate quality, but exclusively about passing the initial filter and getting to an interview. This is precisely where the bottleneck of the entire system is located.

Now let’s add optimisation. On the input side, we have opaque selection criteria that boil down to the resume needing to be optimised for the vacancy and ATS. Obviously, an optimised resume increases the probability of passing the filter and reduces the chance of instant rejection. Nothing unexpected.

Now let’s look at the same situation from the hiring company’s side.

A vacancy receives 100 applications. With high volume, filters inevitably tighten. This leads to a well-observed effect: resumes become increasingly similar to each other because they’re optimised for the same ATS signals. Many recruiters themselves complain about the flow of monotonous resumes.

The key point here is correlation. If resumes accurately reflected a candidate’s actual quality, multi-stage interviews wouldn’t be necessary. But they’re conducted precisely because the correlation between “good resume” and “good candidate” is weak. Moreover, a strong candidate may well not know how to write a resume. This is a common situation.

As a result, in this model, the strongest candidate among applicants has about a 90% chance of being filtered out at the first step simply because their resume didn’t pass the filter. Ten people make it to the interview, selected largely randomly from the set of resumes successfully optimised for the filter, rather than necessarily reflecting the actual strength of candidates.

If a vacancy received 1000 applications, but realistically only the same 10 people can be interviewed in person, then the probability that the hired candidate was the best out of those thousand becomes statistically extremely low. Most likely, it won’t be the worst candidate, but not necessarily the best either. It will be someone from the upper tail of optimisation for filters, not for actual quality.

Conclusions.

In the classic “post vacancy — collect applications — filter resumes” model, mass resume submission, including using automated tools and AI, is rational for candidates. It increases the probability of passing the initial filter and reaching the in-person interview stages faster.

On the other hand, high incoming volume inevitably leads to tightening filters. Tightening filters leads to standardisation and averaging of resumes. And this, in turn, reduces signal quality and makes hiring outcomes increasingly random.

In this system, all participants behave rationally. Candidates increase volume. Recruiters strengthen filters. The market optimises for passing filters, not for selecting people. And each step in this process is logical—even if the result satisfies no one.