Unit Economics: The Numbers Investors Check First (And Founders Learn Last)

Every startup pitch eventually reaches the same moment: an investor asks about unit economics, and the founder either lights up or changes the subject. It's the most reliable tell in the room — because unit economics is just the question "do you make money on each customer, or lose it?" and not knowing means you're scaling blind.

This is the founder's working guide: the four numbers, the honest way to calculate them, and what good looks like.

The Four Numbers

1. CAC — cost to acquire a customer

Everything you spent on sales and marketing in a period, divided by customers won in it. Everything — salaries and founder time included, not just the ad bill. Founders who exclude their own 30 hours a week of selling are calculating fiction. (Price it at what a salesperson would cost.)

2. Contribution margin — what each sale actually leaves behind

Revenue per customer minus the direct costs of serving them: delivery, support time, payment fees, infrastructure that scales with usage. What's left "contributes" to fixed costs and, eventually, profit. If this number is negative, growth is a doom loop — every new customer digs the hole deeper, and the fix is pricing or cost, never volume. Our pricing guide is usually the medicine.

3. LTV — lifetime value

Contribution margin per period × how long customers stay. The trap: young companies don't know how long customers stay, so they guess generously. Discipline: with under a year of data, cap your assumed lifetime at what you've observed, and label every LTV figure with its assumption. "LTV is X if customers stay 24 months — we've verified 7 so far" is a sentence investors trust precisely because it's modest.

4. Payback period — the one cash actually feels

How many months of contribution margin to earn back CAC. This number decides whether growth consumes or generates cash — a 6-month payback means you can reinvest twice a year; 24 months means every burst of growth is a financing event. For most startups, payback matters more than the LTV:CAC ratio everyone quotes.

The Afternoon Calculation, With AI

This used to be an analyst's week. Now:

1. Export — customer list with dates and revenue; expense export tagged sales/marketing vs delivery. (If your books aren't clean enough for this, fix the P&L first — same afternoon, honestly.)
2. Compute — hand both files to an AI analysis tool (Julius AI, or ChatGPT/Claude with the data): "Calculate monthly CAC, contribution margin per customer, observed retention by cohort, and payback. State every assumption you make." The assumption list it returns is half the value.
3. Interrogate — "Which cohort has the best economics? What changed?" "If CAC rises 30% as we scale channels, where do we break?" Ten minutes of questions you'd never have run in a spreadsheet.

The Data & Analytics and FinTech shelves of the directory have the current toolset.

What Good Looks Like (Rules of Thumb, Not Laws)

• LTV:CAC above 3 — the classic bar; below 1 means every sale loses money on purpose
• Payback under 12 months — under 6 and you can grow on your own cash
• Contribution margin positive from early on — negative is only defensible with a stated, dated plan for why scale fixes it
• Economics improving cohort over cohort — the trend persuades more than the level

The Startup-Specific Warning

AI has made it cheap to look like you have great unit economics — pretty dashboards, confident decks. It has not changed what the numbers are. Use the tools to find the truth faster, not to decorate it; diligence in 2026 involves the investor's own AI re-deriving your numbers from raw data, and the gap gets found.

Founders building this muscle properly — models, metrics, investor materials, with AI at every step — will find it's a core week of AI for Startups. Pair this piece with the runway guide for the full financial picture.