Back people, not forecasts

Every investment arrives with a forecast: revenue in year five, margins at scale, the size of a market that doesn't exist yet. The numbers are precise to the decimal point, and almost all of them will be wrong. That isn't a criticism of the people who build them. The future simply doesn't report to a spreadsheet.

What a forecast can't tell you is the thing that decides most outcomes: what the people will do when the plan stops working. And the plan always stops working. A market shifts, a product misses, a key hire leaves. At that point the model is irrelevant, and the only question that matters is who is in the room, how they think, and whether they bend or break.

So we spend most of our diligence on people. How they talk about their mistakes. Whether they know the unglamorous details of their own business. How they treat those who can't help them. Whether they would keep going if nobody were watching. None of this fits in a cell, and all of it compounds.

We still do the numbers. Discipline matters, and a model is a useful way to find out what you have to believe. But we treat it as a question, not an answer. When the spreadsheet and our read of the people disagree, we have learned which one to lean on.