Every founder walks in with a forecast, and every experienced investor applies the same silent correction to it. This chapter is about what happens after that correction: the numbers investors actually run in their heads while you talk, the unit economics they trust instead of your projections, how to price a company that has no revenue, and the multiple logic that decides whether your raise is a business case or an expensive job. The frame is public below. The detailed unit economics, the bet-math a portfolio investor runs, and the reconciled reference of ranges every raise is measured against sit in the free member layer: your email opens them.
Your forecast gets divided by ten
"If I see projections that go up like that, I just divide them by ten."
That is not cynicism, it is pattern recognition. Every deck shows a curve bending up and to the right, so the curve carries no information. And multi-year forecasts fail on principle: nobody knows what any market looks like in five years, least of all a market being reshaped by AI in real time. A precise revenue number for year five does not signal planning. It signals that you think the listener cannot do math.
What convinces instead is command of the machine that produces the numbers: revenue per user, conversion rate, acquisition cost against lifetime value, and above all the next growth lever, the one thing that visibly changes the trajectory once you turn it on.
Investors do not believe your forecast. They believe, or refuse to believe, your mechanics. One investor put it to a group of our founders plainly: he prices the round while you talk, and so does everyone else at the table. Give them clean inputs and they will build a better forecast than your slide, and this time they will believe it, because they built it themselves.