By When an investor opens your financial model, they are not really checking whether you can forecast the future. Everyone in the room knows the numbers three years out are fiction. What they are actually doing is reading the model like a personality test, looking for whether the person who built it thinks clearly, honestly and knows which levers move this particular business. The forecast is the excuse. Your judgement is the real thing on trial.
This is why a beautiful model full of hockey sticks can lose the room while a rougher, honest one wins it. Investors have seen a thousand spreadsheets bend gracefully up and to the right and they trust none of them. What they trust is a model where the assumptions are visible, sensible and tied to reality and where the founder can explain, without flinching, why each number is what it is.
- Assumptions you can defend, not just outputs
- A clear link between spend and the growth it buys
- Honest unit economics, not blended averages hiding the truth
- A downside case, not only the dream
- Numbers that reconcile with the story you just told
They test the assumptions, not the total
So the questions come at the assumptions, never the totals. Why does acquisition cost fall as you scale, walk me through it. What has to be true for this growth line to hold. What happens if it takes twice as long. A founder who can answer those calmly, this number comes from our actual last two quarters, this one assumes the channel keeps performing and here is why, sounds like an operator. One who says the model just outputs that has already lost.
What a model says about you
Show the downside on purpose
The move that builds the most trust is the one founders most fear, showing the downside yourself. Build the case where things go slower, where a channel underperforms and show you still know what you would do. It feels like handing the investor a reason to say no. It does the opposite. It tells them you have thought about how this could go wrong, which is exactly what they are quietly trying to work out and it reassures far more than a model that only knows how to go up.
Show the downside too
So do not polish your model into a fantasy. Make the assumptions visible, tie them to what has actually happened, include the ugly case and be ready to defend every lever in plain language. The number three years out will still be wrong. But you will have shown them the one thing they came to see, that the person building this can be trusted with the truth.