Astrologers in HR and IQ tests: how your funnel scares the best people away

Posted a role, by morning 500 ChatGPT-generated CVs. The exhausted lead asks HR to 'somehow filter this.' Then the creativity starts: Telegram bots, personality tests, unpaid take-homes. You think you're filtering spam. You're filtering the talent.

Modern hiring has turned into a grueling ordeal. Companies are overwhelmed by the volume of irrelevant candidates, while strong specialists burn out from absurd selection stages. Trying to defend against the flood of people with "padded" experience, businesses are damaging their own employer brand.

Familiar picture: posted a role, by morning 500 CVs in the inbox — kindly generated through ChatGPT. The exhausted team lead asks HR "to somehow filter this, because there's no time to work."

And then the workarounds start. Telegram bots with questionnaires get embedded in the funnel. Personality-type tests. Requests to record a video pitch before the first call. Bulky take-home assignments nobody pays for. It seems that this way you'll filter out the low-effort applicants and leave only the most motivated.

Who do you actually filter out? You filter out the normal engineer.

What the strong candidate does when he sees your funnel

A good specialist has a stable job and two or three current offers from sane recruiters. He opens your posting, sees a requirement to take a 150-question psychological test — and quietly closes the tab. He doesn't need to prove his professional fitness through bureaucratic hoops. His GitHub and shipped projects speak for themselves.

As a result, only two categories make it through your obstacle course: desperate beginners with nothing to lose, and professional interview-passers. The latter have learned to lie perfectly in questionnaires and bypass ATS systems, but cannot write a line of maintainable code. The paradox: you wanted to filter out spam — and filtered out the talent.

The hidden cost of "free" filtering

Companies often look at the funnel additions as cost-free — "the bot runs by itself, the test is automated, the take-home is the candidate's time." That accounting is wrong on three counts.

First, the take-home costs your senior engineers 4–8 hours per candidate to review. Multiply by the number of candidates who do reach that stage. That's senior time at fully-loaded cost, not free.

Second, every irritating stage extends the funnel. Your time-to-close goes up. Each extra week the seat is open costs the team a measurable share of output (we wrote about this in our piece on what a hire actually costs).

Third, the candidates who declined never tell you why. The signal disappears. You don't learn that the bot at the entry, the long test, the unpaid four-hour assignment — that's where the strong applicants drop. You only see that "few candidates make it through."

What works instead

Quality hiring is always precise work tailored to the role — not an impersonal conveyor. The components are simpler than the bot-and-test architecture:

• A reasoned reach-out with a specific reason this person, not a templated "are you open."
• A 30-minute calibration call with the hiring manager early — not after three filters. Both sides need to know if there's mutual interest before either invests more.
• Real work as the technical screen. Discuss architecture with backend candidates. Walk through component performance with frontend. Pair-debug a real issue if you can. None of that takes longer than your "leetcode round" did — and it tells you whether you actually want to work with the person.
• Decisive offer. After the final interview, decide within 1–2 days (see our piece on why slow offers lose finalists).

That's what we do as an agency: take the dirty work off your plate, scan the market deeply, and put 3–5 vetted candidates on the table. No bots, no stress-resistance tests — only substantive conversation about architecture, processes and the real tasks of your product.

If you remember nothing else

Every stage you add to your funnel costs you the candidates with the best alternatives first. The ones who stay are the ones with the fewest options. You can either accept that — and own it as a deliberate hiring strategy for the lower bands — or you can stop building barriers that filter out exactly the people you said you wanted.