The Margin Test.

Written and narrated by Serguei Poppeleer.

Every pitch deck has a beautiful margin. Costs on the optimistic line, volumes on the hopeful one, a tidy gap in between. It is the easiest slide to build and the easiest one to disbelieve.

A best-case margin isn't a margin. It's a wish with a percentage sign.

The third side of the triangle

The margin test asks whether the unit economics survive real costs — the bad day, not the good one. The supplier price that moves. The wastage nobody modelled. The discount the first big customer demands. A margin that only holds when everything goes right is not a margin you will get to keep, and an experienced investor knows it on sight.

Proof · Margin on real costs — a fuel venture

The margin was anchored in real feedstock economics, not spreadsheet optimism: vegetable-oil trading costs, bad-day supply assumptions, and modular units rather than one fragile mega-asset. The numbers were built from what things cost when they stop going perfectly — the only version of the numbers an investor trusts.

How to pass it

Build the unit economics from bad-day costs, then stress them harder. Find the break point on paper — the price, the volume, the cost at which the margin disappears — before the market finds it for you. A model that knows its own break point is far more backable than one that pretends it doesn't have one.

"We'll fix the margin at scale."

Scale doesn't fix a broken margin. It multiplies it. If you lose a little on every unit, the cure for ten units is not ten thousand.

Scale doesn't fix a broken margin. It multiplies it.

Next in the series

Next: “Numbers in the Head.” If your model lives only in your head, it isn't a model — it's a memory only you can access.

Does your margin survive the bad day?

Get your Delivery Score. Fifteen Yes/No questions, about four minutes. It shows which of the five controls to put in place first, before you build.

An idea is worth nothing. A model is worth backing.