What would have to be true.
"Invert, always invert."
The most common move on a hot deal is to argue why it works. The Butterfly Memo panel runs the opposite. For each thesis the deal depends on, the panel asks: what would have to be true for this to fail?
It's the same structural move as a pre-mortem (Klein, 2007) — imagine the failure, work backward to the cause. The difference is that twelve lenses run it in parallel, each from a different vantage. Marks looks for the second-order failure. Taleb looks for the fat-tail failure. Bessemer looks for the failure mode you'd publish on the anti-portfolio in five years.
For each thesis you bring, the panel produces a structured chain: the thesis, the inversion, the evidence you'd need to disprove the inversion, and the trip-wire you'd watch for over the life of the investment.
You see all twelve inversions in the memo, side by side. The IC sees the deal's real surface area — not just the founder's slide.