The Hiring Leak | Article 4: The Interview Performance Bias
When résumés stop giving us certainty, we turn to interviews.
Interviews feel fair. We get to ask questions, see how someone thinks, and watch them work in real time. It feels closer to reality than a list of past jobs.
But interviews don’t measure how someone works.
They measure how someone performs.
That difference is where the leak lives.
Interviews Are Artificial Environments
An interview is a strange situation. A person is:
- Being observed
- Being judged
- Operating with incomplete context
- Under time pressure
- Trying not to fail
This is not how real work happens.
In real work, people can:
- Pause
- Look things up
- Ask clarifying questions
- Revise their thinking
- Make mistakes without an audience
The interview removes all of that — and then treats the result as predictive.
We Reward Speed Over Judgment
Many interview formats quietly optimize for speed. Quick answers feel impressive. Silence feels uncomfortable. Candidates learn to keep talking.
But in real systems, speed without judgment is dangerous.
Good engineers often:
- Slow down when things are unclear
- Ask questions before committing
- Think through second-order effects
In interviews, those behaviors can look like hesitation. So we filter them out.
Confidence Becomes a Stand-In for Competence
Confidence is easy to detect. Correctness is not.
When someone answers decisively, we relax. When someone qualifies their answer, we worry.
But real work is full of:
- Uncertain requirements
- Conflicting constraints
- Partial information
Confidence in those environments isn’t a personality trait — it’s a risk. Interviews reward people who can sound sure without being sure.
Whiteboards Are Not Production
We ask people to solve problems in isolation, out loud, on the spot.
They can’t:
- Check assumptions
- Consult documentation
- Explore alternatives quietly
- Iterate without social pressure
So candidates optimize for:
- Memorized patterns
- Recognizable solutions
- Performing thoughtfulness instead of practicing it
What we see is fluency — not reliability.
The Silent Filter
Over time, a pattern emerges.
We hire people who:
- Think fast
- Speak smoothly
- Perform under observation
- Appear decisive
We reject people who:
- Think carefully
- Ask for context
- Pause before answering
- Treat uncertainty seriously
Then later, we ask why:
- Decisions feel rushed
- Assumptions go unchallenged
- Mistakes repeat
- Systems feel brittle
The system is consistent. The surprise is ours.
The Real Mismatch
When an interview hire struggles later, we say:
- “They interviewed so well.”
- “They seemed sharp.”
- “We checked all the boxes.”
And that’s true. They performed well in the simulation we designed. The problem is that the simulation rewarded the wrong traits.
The Core Insight
Interviews don’t reveal how someone works in a system.
They reveal how someone performs in a spotlight.
If your organization needs careful judgment, collaboration, and adaptability, then selecting for performance under pressure will quietly work against you.
The next leak appears when we celebrate people who succeed alone — and mistake that for strength in a distributed system.
That’s the Lone Wolf Signal.