A guided walk through the machine investors expect you to understand: marketing dollars → customers → revenue → burn → runway — how much more you spend than you take in each month, and how many months of cash that leaves you. This is the classroom; the Excel starter workbook is the homework — same steps, same order, same names.
Before any spreadsheet, answer the first question every investor asks: what is a customer worth (LTV), and what does one cost (CAC)? LTV is the hard ceiling on what you may ever pay for a customer. If this ratio doesn't work, more marketing just loses money faster.
The LTV in Step 1 is a claim — a formula built on one average churn number. Investors don't underwrite claims; they ask for cohorts: group customers by the month they arrived and watch what each group actually does. Real cohorts leak hardest in the first 90 days, then settle — which is why a flattering average can hide a broken product. The table is a triangle because young cohorts have fewer observed months.
The one rule that separates credible models from fantasy: you never type revenue in. Revenue only emerges: marketing budget ÷ CPA = new customers; customers × price = revenue. Your budget is the throttle — and churn is the leak. Try moving the sliders and watch the machine respond.
Good unit economics at the customer level does not make the company profitable — you need enough volume of customers to cover the fixed costs of payroll and overhead. Payroll is almost always the biggest line. (In the workbook this is a headcount grid × salaries with a +25% load for taxes and benefits; here it's simplified to totals.)
Profit is an opinion; cash is a fact. A money-losing startup lives and dies by its cash position. The deepest point of this curve — the trough — is the number that sizes your raise: your round must cover it, with a buffer. Express that ask in months to the next fundable milestone, not merely months of survival. Budget six months of runway to run the raise — to money in the bank, not to term sheet — and remember investors judge you on your trailing 12 months. That is why 12–18 months is the planning range: twelve to build the record they read, six to close on it. Raise when it's easy, not when you need it.
Every metric chips away at five questions: what's a customer worth, what does one cost, the ratio between those, where the business is now, and how fast it's growing. Here is your model's answer sheet — and the red flags an investor would raise in the first ten minutes.