The Problem With “Culture Fit”
Every engineering leader I know has sat in a debrief where someone said “they just don’t seem like a good culture fit” — and then the room nodded and moved on. I’ve been in those rooms. I’ve nodded. And I’ve spent the last several years trying to understand why that phrase is one of the most dangerous things you can say in a hiring process.
Let’s start with what “culture fit” was supposed to mean.
The Original Intent
When organizational psychologists first introduced culture fit as a hiring criterion, the idea was legitimate: does this person share the company’s core values? Do they approach work in ways that are compatible with the team’s working style? Can they thrive in this environment? These are real questions. A company that values deep asynchronous thinking will genuinely struggle to integrate someone who needs constant real-time collaboration. A startup that moves at breakneck speed will exhaust someone who does their best work through deliberate, methodical process.
Used correctly, culture fit was about compatibility — not conformity.
What It Became
Here’s what research has consistently shown: in practice, “culture fit” is a proxy for demographic similarity.
The MIT and University of Chicago audit studies are the most damning. In callback-rate studies where identical resumes were submitted with stereotypically Black-sounding vs. white-sounding names, callback rates differed by 50% or more. But the effect is not just about race — it’s about every dimension of perceived similarity. Interviewers rate candidates higher on “culture fit” when they share hobbies, attended similar schools, have similar communication styles, and come from similar socioeconomic backgrounds.
A 2012 study by Lauren Rivera (“Hiring as Cultural Matching”) documented elite professional services firms explicitly using culture fit to select candidates who shared interviewers’ leisure activities — squash, sailing, certain kinds of travel. The work itself was barely discussed.
This is what culture fit screening has become: an unstructured, unaudited, bias-amplifying mechanism that lets interviewers act on gut feeling while believing they’re making principled decisions.
The Business Irony
Here’s what makes this particularly frustrating from a strategic standpoint: the business case for cognitive diversity is well-established. Teams with diverse perspectives identify more potential failure modes, generate more creative solutions, and make fewer correlated errors. McKinsey’s diversity research consistently shows that companies in the top quartile for ethnic and gender diversity outperform on profitability.
So we have a hiring criterion — culture fit — that was designed to build cohesive teams, but instead screens out the cognitive diversity that makes teams actually perform better. We’re optimizing for comfort and producing underperformance.
Culture Add: A Better Frame
“Culture add” does not mean abandoning the goal of cultural cohesion. It means asking a different question: not “does this person fit what we already are” but “does this person share our core values AND bring something we’re missing?”
This requires two things that culture fit never required: (1) being explicit about what your core values actually are, and (2) being honest about what gaps exist in your current team.
If your core values are things like “ships high-quality work,” “gives and receives feedback directly,” and “takes ownership through ambiguity” — those are evaluable criteria. You can build interview questions around them. You can score candidates against them. You can audit your pass rates across demographic groups.
Structural Changes That Actually Work
Here are the practices that have evidence behind them:
Structured interviews with shared rubrics. Every interviewer scores the same dimensions using the same criteria defined before the interview. This dramatically reduces the surface area for culture-fit rationalization.
Diverse interview panels. Not tokenism — genuine representation across demographic dimensions, seniority, and working style. Research shows diverse panels produce more consistent, less biased evaluations.
Work-sample tests. A take-home problem, a pair-programming session, or a system design exercise evaluated against explicit criteria tells you far more about job performance than how comfortable you feel in a 45-minute conversation.
Blind resume screening. Remove names, graduation years, and school names from initial screening. Several studies show this alone significantly increases demographic diversity in interview pools.
Calibration sessions. After each hiring cycle, review your pass rates by demographic group. If you cannot see disparities, you cannot address them.
What Cohesion Without Conformity Looks Like
The best engineering organizations I’ve seen achieve something harder than either culture fit OR culture add in isolation: they have very explicit, well-understood values that people are genuinely selected for — and they hire people who are genuinely different from each other within that values framework.
That requires doing the harder work of defining what you actually care about, building evaluation systems that measure it, and having the discipline to override gut feeling when the process contradicts it.
It’s more work than nodding in a debrief. It’s also how you build a team that’s actually capable of the work ahead.
