"Hire for cultural fit."
A polite phrase for hiring people who already agree with you.
Lie No. 02 — Hire for cultural fit.
The Myth
Hire people who mesh with the team and share the company's values, and everything runs smoother. "Fit" keeps the culture strong.
The Cold Truth
"Cultural fit" is usually a polite name for homophily — favouring candidates who mirror us, because sameness feels like competence (an effect known as "looking-glass merit"). Build a whole company that way and you get a monoculture: tidy, harmonious, and fragile, with shared blind spots and no one to see what everyone missed. The difficult, brilliant people who'd actually push you forward rarely survive a fit screen. And it's worse now that it's automated: AI hiring tools learn from your past "good fits" and scale that sameness instantly — AI doesn't remove the bias, it systematises it. That's why employment AI is now regulated as high-risk.
The Reframe
Don't abandon culture — stop confusing it with conformity.
Five moves:
Change the question. Ask "what do they add that we don't already have?" instead of "do they fit?" Culture add, not culture fit.
Separate values from sameness. Keep shared standards for how people treat each other; drop "is like us / agrees with us." Hire for the first, never the second.
Build the process against the mirror. Same job-relevant questions for all, scorecards tied to the role, a panel that isn't all one kind of person.
Audit your AI, don't trust it. It was trained on your past sameness. Demand bias testing and human oversight — now legally required in many places — and never call a vendor's black box "objective."
Make trust the foundation. Difference without trust is just friction. Build a place where it's safe to disagree, or your diverse hires assimilate or leave.
The goal was never people who fit. It was people who make the place better than it was before they arrived.
References
"Looking-glass merit" / similarity bias (homophily) in hiring, and the "culture add vs culture fit" idea.
EU AI Act (Regulation 2024/1689): recruitment AI is classed "high-risk." Compliance obligations were due 2 August 2026 but, under the 2026 AI Omnibus agreement, are now expected to apply from 2 December 2027 (deferral provisionally agreed May 2026, pending formal adoption). The high-risk classification itself is unchanged.
AI hiring-bias rules (trend, 2025–26): NYC Local Law 144 bias audits; California ADS regulations; EEOC algorithmic-fairness priority. Case to watch: Mobley v. Workday.