You know the feeling. Someone asks a question about your decision — not a clarifying question, not a curiosity question — and your stomach tightens. Your first thought is not huh, interesting but that's not the point or you're missing context or that's a loaded question.
That question is the audit. The defensiveness is the signal.
The question that doesn't feel like a question
Last week a founder told me about their Series A term sheet. Clean terms, reputable firm, the valuation they wanted. They were ready to sign. Their advisor asked one question: If this firm pulls their support in eighteen months, what does your cap table look like?
The founder's face changed. That's not going to happen. They've committed.
But if it did?
Why would you ask that? They're not the kind of firm that—
Indulge me.
The founder walked through it. The follow-on depended on this firm leading. The other investors were along for the ride. If the lead walked, the round collapsed. The founder would be back at zero with six months of runway.
The question wasn't an attack. It was the only question that mattered. And it felt like betrayal because it made the founder's certainty — this deal is solid — look like the risk it actually was.
Why the betrayal feeling is the signal
Your brain builds certainty as a compression algorithm. It takes a messy, uncertain reality and compresses it into this is the right call so you can act. That compression is necessary. But it discards data — specifically, the data that would change your mind.
An audit question decompresses. It forces the discarded data back into view. Of course it feels like betrayal — it's literally undoing the work your brain did to let you sleep at night.
The questions that don't feel like betrayal — have you considered X? what about Y? — are not audit questions. They're decoration. They let you say yes, I considered it without changing anything.
The audit question is the one you would never ask yourself. Because asking it yourself would mean admitting you might be wrong. And you can't admit that while you're still deciding.
The three shapes of the betrayal question
The counterparty question. Who benefits if I'm wrong about this? (This is the one from last week's post — it's the backbone.) It forces you to model the other side's incentive structure instead of your own wishful thinking.
The reversal question. What would have to be true for me to change my mind? Not what would change my mind — that invites a list you'll never act on. What would have to be true forces a concrete scenario. If you can't name one, you're not deciding. You're believing.
The future-regret question. In two years, if this goes badly, what will I wish I had asked today? This one bypasses your current framing entirely. It speaks to the version of you who lives with the consequences, not the version who manages the narrative.
What to do when you feel it
Don't suppress the defensiveness. Don't explain it away. Name it.
This question makes me defensive. That means it's the audit.
Then answer it. Not with a rebuttal. With the decomposition the question demands. Lay out the scenario where you're wrong. Name the counterparty's incentive. Specify the reversal condition. Write the future regret.
You don't have to change your decision. But you do have to see what your certainty was hiding.
DAUDIT exists because the betrayal question is the one almost nobody asks out loud. Bring one decision. See what you're protecting.