🚀 How Startups Work

What you’ll learn

1. The Graveyard Has No SpeakersUnderstand survivorship bias through Wald's bombers and see why it makes almost all popular startup advice unreliable — establishing the course's mechanism-over-mythology approach.Abraham Wald realised the bullet holes on returning bombers marked the survivable places; the unhit engines were where a hit meant the plane never came back. Everything most people know about startups comes from survivors, who also narrate sincere after-the-fact explanations that exclude everyone who did the same and failed. This course therefore teaches mechanisms — equity, dilution, preferences, base rates — because they're arithmetic and don't lie.
2. What a Startup Actually IsDefine a startup as a temporary organisation searching for a repeatable, scalable business model — and see why that job requires equity rather than debt.A bakery executes a known model; a startup searches for an unknown one, which is why it looks from outside like not knowing what it's doing. Because the search produces knowledge rather than salable output, it has no revenue and no collateral, so banks won't fund it. Equity — no downside protection in exchange for uncapped upside — is the only instrument that fits, and that trade seeds nearly every other mechanism in the course.
3. Product-Market FitUnderstand product-market fit as the hinge of the whole enterprise, and why scaling before it converts remaining runway into a bigger version of the wrong thing.Marc Andreessen described fit as being in a good market with a product that satisfies it, and argued its presence or absence is unmistakable. The defining signal is pull: before fit you hand-carry every customer, after fit they arrive faster than you can serve them. Determined founders can hand-sell without fit, which disguises the problem — and spending on growth then buys rented customers while burning the runway needed to fix the real issue.
4. Why Equity ExistsUnderstand equity as a claim on the future and vesting as the mechanism that makes founding teams and companies workable.A startup needs expensive people and has no money, so it pays with ownership — a claim on a future that may be worth nothing. Founding splits are simple because the shares are worth almost nothing, and everything afterwards complicates them. Standard four-year vesting with a one-year cliff means a co-founder who leaves at two months takes nothing, which protects the founder who stays and makes the company fundable.
5. The Priced RoundCompute dilution correctly — post-money as the honest denominator — and understand the option pool shuffle hidden inside a headline valuation.$2M at an $8M pre-money means a $10M post-money and 20% ownership, because once the cheque clears the company genuinely contains that cash. Across a seed and a Series A with a 10% pool, founders fall from 100% to 56% — but their percentage fell while their value rose, since 100% of an unfunded hypothesis is worth nothing. The pool is almost always carved from the pre-money, diluting founders rather than the incoming investor, so the headline number is never the deal.
6. Two Classes of StockDistinguish common from preferred stock and understand why 'I own 1%' is an incomplete description of anyone's position.Founders and employees hold common stock; investors buy preferred, a different instrument whose rights only become visible when something happens to the company. The asymmetry it prices is real: a failed company costs the founder years and the investor cash. Preferred typically carries a liquidation preference, anti-dilution protection, pro-rata rights, and board seats with protective provisions — so the question is always 1% of what, and behind whom.
7. Liquidation PreferenceUnderstand the liquidation preference and the preference stack — the mechanism most likely to surprise employees — and the difference between non-participating, participating and multiples.A liquidation preference pays preferred holders back before common receives anything, so a company that raised $60M and sells for $50M leaves common holders with nothing at all. The headline is the sale price; your shares are a claim on the residual, and raising more money makes a mid-sized exit worse for employees. Terms have converged to the founder-friendly end — Cooley's Q4 2025 report found 98% of deals at 1x and 96% non-participating — but conventions tighten when money is scarce.
8. SAFEs and Deferred QuestionsUnderstand SAFEs — caps, discounts, and why deferring the valuation question creates a bill that arrives all at once.Pricing a company that is two people and a prototype is fiction, so Y Combinator's SAFE (2013) defers the question: money now, shares later at the next priced round. Investors are compensated for going first by a discount and, more consequentially, a valuation cap that lets them convert as if the company were worth far less. Because nothing appears on the cap table until conversion, founders can sign several without watching any percentage fall — then meet the total at their Series A.
9. OptionsUnderstand options as a bet that costs money to collect — strike, 409A, vesting — and the exercise-window trap that costs departing employees their vested equity.Equity grants are usually options: the right to buy shares at a strike price set by a 409A valuation, worth nothing unless the shares exceed it. On leaving, a classic 90-day window forces a departing employee to find a large sum for shares in a private company they cannot sell — potentially plus alternative minimum tax on paper gains — or forfeit four years of vested work. Some companies extend the window to five or ten years, a real and checkable difference that almost nobody asks about.
10. Burn, Runway, and the ClockLearn burn, runway and default alive/default dead — and why the raise must start long before the money runs out.Burn is net cash out per month, runway is cash divided by burn, and Paul Graham's default alive/default dead asks whether current growth reaches profitability before the money ends. Runway is the only number that's a genuine deadline, which is why the founder's job is often described as never being surprised by the date. A raise takes months, and a founder with two months left isn't negotiating but pleading.
11. The Base RatesConfront the honest base rates — business survival, seed-to-Series-A graduation, and the VC power law — and see timing's role made numerical.BLS Business Employment Dynamics data shows 77.9% of new US businesses survived one year, 51.4% five years and 34.7% ten years — and that's the gentle dataset, counting every corner shop. Carta data shows 30.6% of the Q1 2018 seed cohort reached a Series A within two years versus 15.4% of the early-2022 cohort: the same milestone, four years apart, with timing as the dominant variable. Correlation Ventures' study of 21,000+ financings found roughly 65% of deals returned less than invested and about 4% returned over 10x, which is why VC economics demand rare enormous outcomes.
12. The Mythology and the MechanismReturn to Wald and assemble the course's mechanisms into the engines — the fatal areas the survivor stories never show.The engines are the failures that never reach a conference stage: running out of runway before fit, scaling before fit, SAFEs converting all at once, the pool shuffle, a preference stack that zeroes a $50M exit, a 90-day window that costs four years of work, raising with no alternative, and timing nobody chose. The course deliberately supplies no formula for success, because confident formulas are holes read in returning bombers. It supplies the machine's workings instead, so that starting a company is a choice rather than a wish.

Questions this course answers

Grounded in trusted sources

• Abraham Wald — 'A Method of Estimating Plane Vulnerability Based on Damage of Survivors', Statistical Research Group, Columbia University (1943)
• Steve Blank — 'The Four Steps to the Epiphany' (2005); 'What's A Startup? First Principles' (2010)
• Marc Andreessen — 'The Pmarca Guide to Startups, part 4: The only thing that matters' (2007)
• Brad Feld & Jason Mendelson — 'Venture Deals: Be Smarter Than Your Lawyer and Venture Capitalist'
• Cooley LLP — 'Q4 2025 Venture Financing Report' (9 February 2026) — https://www.cooley.com/news/insight/2026/2026-02-09-q4-2025-venture-financing-report
• Y Combinator — 'Safe Financing Documents' — https://www.ycombinator.com/documents
• Paul Graham — 'Default Alive or Default Dead?' (2015) — https://www.paulgraham.com/aord.html
• U.S. Bureau of Labor Statistics — Business Employment Dynamics, business survival rates, as reported by LendingTree — https://www.lendingtree.com/business/small/failure-rate/

Every Wunder lesson is built from real, reputable sources — never invented.

Related Math courses