The Hogwash Question (Breaking the Interview Mindset)
Ask what conventional wisdom is inaccurate — it's unanswerable without genuine opinion, so it exposes authenticity.
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 4
- Confidence
- 92%
Interviews systematically reward candidates who correctly guess what the interviewer wants to hear, which means strong but non-performative people get dismissed for saying nothing new. This question is engineered to be unanswerable by mirroring: there is no way to name conventional wisdom you bet against without revealing an actual opinion. It is deployed specifically as a 'save question' — a last check before rejecting someone whose credentials look strong but whose answers were bland.
Origin
Nikhil Singhal, VP of Product at Facebook (Meta), on Lenny Rachitsky's podcast. He describes it explicitly as his 'save question' — a tool to rescue high-quality candidates from being penalised by the interview format itself.
Core principles
- 01Prepared candidates run a prediction loop: they answer what they think you want to hear, not what they think.
- 02That loop causes false negatives — you dismiss genuinely good people because they told you nothing new.
- 03A question with no socially safe answer breaks the loop, because mirroring requires knowing the expected answer.
- 04Blandness is often an artefact of the interview format, not a property of the candidate — test for that before rejecting.
- 05Tailor the domain of the question to the candidate's actual expertise so they have real conviction to draw on.
How to run it
- 1
Notice the false-negative risk
Flag the candidate who is credible on paper and competent in conversation but has told you nothing you did not already know. Before rejecting, consider that the interview mindset — not their thinking — may be the cause.
Pro tip Keep this question in reserve rather than in your standard script; it works best as a deliberate intervention.
- 2
Ask the hogwash question
Ask: what's something everyone takes for granted that you think is essentially hogwash or inaccurate? There is no way to answer without being genuinely opinionated, because the candidate cannot infer which orthodoxy you want attacked.
Pro tip The two-part structure matters — first name the thing you believe I expect you to say, then tell me why it's wrong.
- 3
Anchor it to their real domain
Scope the question to where they have earned conviction. For a manager: you've managed hundreds of people — what conventional wisdom did you bet against and find inaccurate? For an AI specialist: what do people believe about AI that's wrong?
Pro tip Domain-anchoring lets you separate a contrarian pose from a contrarian position backed by scar tissue.
Watch out A generic version invites edgy, unearned takes. Anchor it or you will grade rhetoric.
- 4
Grade for authenticity, not for agreement
You are testing whether the person is authentic and genuinely opinionated — not whether you agree with the contrarian claim. Look for a position they have actually lived and can defend from experience.
Pro tip Ask what evidence changed their mind. A real bet against conventional wisdom has a cost story attached.
Watch out Do not penalise a candidate for holding a view you disagree with — that reintroduces the mirroring problem you were trying to break.
In the wild
Singhal describes the failure mode directly: everyone prepares for interviews, and their entire conversation becomes a prediction of what the interviewer wants to hear. The result is that high-quality people get dismissed for not being genuine — even though the format caused the problem. He uses the hogwash question as his save question, deployed precisely when he is inclined to reject someone for telling him nothing new.
→ By breaking the wall of the interview mindset, he gets a genuine, opinionated answer and avoids letting the interview process itself penalise a strong candidate.
For an experienced people leader he reframes it: you've managed hundreds of people in your career — what conventional wisdom did you bet against that you found was actually inaccurate? The candidate must draw on real management decisions rather than abstract opinion.
→ The domain anchor forces a specific, experience-backed claim, which is far harder to fake than a general contrarian statement.
Common mistakes
Treating the question as a culture-fit filter
The point is to detect authenticity, not agreement. If you reject people whose contrarian view clashes with your own, you are running the same mirroring test you claimed to be escaping, just with extra steps.
Using it on every candidate as a warm-up
Its power comes from surprise and from being asked late, when the candidate has already settled into performance mode. Front-loading it into your standard script makes it just another question people prep for.
Rejecting bland candidates without a save question
Blandness is frequently a format artefact. If you have no mechanism to break the interview mindset, your process will systematically reject genuine, non-performative people in favour of skilled interviewers.
Is it for you?
Best for
Interviewers who suspect their process is rejecting substantive-but-unpolished candidates, and who need a final authenticity check before a no-hire decision
Not ideal for
Junior candidates with little domain experience to hold real convictions about, or highly structured, legally-constrained interview loops with fixed scripts
From the transcript
“what's conventional wisdom that you bet against that you have found is actually actually inaccurate”
“everyone always prepares for interviews and then their entire conversation is predicting what you think you want me to say and as a result you…”
“when I break that wall I'm testing is this person authentic”
From the episode
My favorite interview questions from 100+ guests