Raising capital
A seed plan that survives the second meeting
The first meeting is about the story. The second is about the model — and that is the one where most decks quietly fall apart, because the numbers on the slides were typed rather than computed.
The figure on slide nine is the figure in the model, because it is a reference into the model rather than text.
Bottom-up sizing, because share-of-a-large-market is a tell
A market section that starts from a published industry total and claims a percentage of it is the classic unfundable one. We build from the buyers up — how many there are, what they spend, what share your capacity can actually serve — and print the arithmetic line by line, then cross-check it against a published figure and against your own model's year-three revenue.
- TAM, SAM and SOM with every step's working shown
- An uncited published figure is inadmissible, not quietly accepted
- A model that outruns the market it claims is caught before you send it
Unit economics an investor can interrogate
Lifetime value computed on gross margin rather than revenue, which is the most common way the figure gets inflated threefold. Payback in months of gross profit. Acquisition cost with the monthly spend the growth rate actually implies, which is the line most plans leave out entirely.
- LTV, CAC, the ratio and payback, from the same metrics module the product uses
- Contribution margin and break-even on both profit and cash
- Cohort-style churn treated as an estimate and labelled as one
A cap table that survives a SAFE
Pre-money and post-money, SAFE conversion with caps and discounts, priced rounds, the option pool and where it is taken from, and the dilution waterfall that follows. The founder ownership figure in the narrative is read from that table, not typed beside it.
The deck is generated from the plan
Problem, solution, why now, market, product, model, traction, competition, team, financials, ask — in the sequence investors expect, exported to PowerPoint or PDF. Because it is built from the same model as the document, there is no second copy of the truth to fall out of date between the raise starting and the term sheet arriving.
What they are actually testing
- Whether you know your own numbers
- An investor changes one assumption out loud and watches what you do. When the model is driver-built rather than growth-rate-built, you can answer; when it is not, you cannot.
- Whether the market claim is arithmetic or vibes
- Nobody believes a TAM. What they are checking is whether you can derive one — and whether the customer count that falls out of your obtainable share is a number your own operations could serve.
- Whether the deck and the plan agree
- Diligence finds the discrepancy, and the discrepancy is what costs you credibility rather than the number itself. Ours cannot disagree: the consistency checker covers the deck exactly as it covers the plan, and blocks export until they reconcile.
What this does not do
Any page listing only strengths is marketing. These are the limits we would want to know about.
- We will not make a weak business look fundable. A model built on drivers is easier to interrogate, and that cuts both ways.
- Traction is traction. We can present it precisely; we cannot manufacture it.
Start with the numbers.
Free to generate and read. Pay once, when you are ready to send it.