Thiel's Law

1. Foundings Are Forever

Thiel opens with constitutions: the United States still lives under compromises struck at its founding — tiny Alaska gets as many senators as giant California. Companies work the same way. The founding moment is when the rules, formal and cultural, get written, and they keep governing long after everyone forgets writing them. At Google, internal debates years later were settled by appealing to what the founders decided at the start. The founding era arguably lasts as long as a company keeps creating genuinely new things — which is also why founders should stay in charge while it does.

A startup messed up at its foundation cannot be fixed.

The single most important founding decision is who you found with. Co-founders need a real shared history and complementary strengths; grabbing one in a hurry because you feel you need one is how companies get pre-broken. A brilliant team that cannot get along fails no matter how good everything else is.

Getting married to the first person you meet at the slot machines in Vegas probably doesn't [make sense].

2. Structure: Delaware C-corp and the Three Kinds of Power

The boring answer is the right one: be a Delaware C corporation. S-corps allow only one class of stock and cannot grant options; LLCs look tax-efficient but make preferred stock and option grants awkward — and acquirers price every target as if it were double-taxed anyway, so the LLC advantage evaporates at exit. Delaware wins because its corporate law is clear and its Chancery courts are fast and predictable; more than half of large U.S. corporations incorporate there.

• Ownership: who legally holds the equity — founders, employees, investors.
• Possession: who actually runs the company day to day.
• Control: who formally governs — in practice, the board of directors.
• When the three drift apart you get DMV-style dysfunction: citizens nominally own it, window clerks possess it, bureaucrats control it — and nobody is aligned with anybody.

A solo founder is perfectly aligned with himself. Every person added — co-founder, employee, investor — splits the three kinds of power further: employees have possession but little ownership or control; investors have ownership and control but no possession. Each split is a new place where misalignment can take root and compound as the company grows.

3. Alignment: Equity, Cheap CEOs, and Vesting

Cash pay decouples people from outcomes; equity ties everyone's payoff to the same number — the value of the company. Thiel's most concrete rule comes from Founders Fund diligence: CEO salary turned out to be the single most predictive variable they found. A CEO earning under about $150k is betting on the equity and will attack problems; one earning $300k risks becoming a politician whose real job is defending the salary. And because the CEO's pay caps the whole pay scale, a cheap CEO keeps the entire company hungry.

• Standard vesting: four years with a one-year cliff — 25% vests at the one-year mark, the rest monthly over the next 36 months.
• Founders should vest too: it protects everyone left behind if a founder walks away early.
• Fully-vested grants and cash-paid consultants create bad incentives; equity should always be earned over time.

Everyone in a substantive role must be full-time and paid mainly in equity. Part-timers, advisors, and consultants are misaligned by construction: they capture upside without sharing the risk. In Thiel's framing, you are either on the bus or off the bus — there is no half-seat.

4. Equity: Forms, and the Only Number That Matters

Startup equity confuses smart people because absolute numbers feel meaningful and are not. Share count, share price, even the nominal size of an option grant are all noise; the signal is the percentage of the fully-diluted company you would own. Most founders — and most employees — never do this arithmetic.

1. 1

   Common stock

   The baseline ownership unit for founders and employees. 200k shares out of 10m is exactly the same as 20m out of 2bn: 2%.

2. 2

   Stock options (ISOs / NSOs)

   The right to buy shares at a strike price set at fair market value to avoid immediate tax. ISOs get friendlier tax treatment and expire in ten years, which quietly locks employees in; NSO gains are taxed as ordinary income at exercise.

3. 3

   Restricted stock

   Stock bought cheaply up front, with the company's right to repurchase it lapsing over time — vesting run in reverse.

Because risk falls as a company matures, earlier equity is worth far more per unit of work: at eBay, secretaries who joined three years early ended up making about 100x what their later-hired Stanford MBA bosses did. That is rational risk-reward — but explosive if visible, which is why companies keep individual grants confidential.

5. Raising Money: Angels, Convertible Notes, Series A

The essay works the arithmetic of a first round: two founders each buy 1m shares at $0.001; an angel puts in $200k at $1 per share for 200k new shares; early hires and consultants get grants of 100k shares each. Afterward roughly 3m shares are outstanding — the angel owns 6.7%, each founder 33.3%, and the notional valuation is $3m. Note the mechanism: dilution happens because the company issues new shares, not because anyone sells their own.

• A priced equity round in Silicon Valley runs roughly $30–40k in transaction costs; a convertible note is far cheaper and faster.
• A note with a valuation cap (say $4m) and a discount (say 20%) defers the valuation question until professional VCs price the Series A.
• No price today means no reference point — so before the Series A, a down round is mathematically impossible.

At the Series A, VCs spend about a month on diligence — people, financials, technology — and then negotiate the option pool before their money goes in. A 15% pool leaves room to hire stars but dilutes founders up front; a 5% pool protects ownership but may cost you the one person you need. VCs push for a large pre-money pool precisely because that way the dilution lands on the founders, not on them. It is a pure fear-versus-greed tradeoff.

6. Protections, Boards, and the Long Dilution Game

A 1x non-participating liquidation preference — investors get their money back first, then everyone shares pro rata — protects against founder self-dealing without warping incentives. A 2x participating preference breaks alignment in the middle: in a $100m exit the investors double their money while the founders may see little, so the two sides start wanting different outcomes. Anti-dilution provisions — full ratchet at the harsh end, weighted average more commonly — retroactively reprice earlier investments whenever a round is done at a lower valuation, and it is founders and employees who absorb the hit.

Companies are essentially broken the day they have a down round.

• In a down round, anti-dilution triggers gut founder and employee equity, and owners, controllers, and operators start blaming one another. If one is truly unavoidable, make it catastrophic enough to wipe out and silence the angriest parties.
• Boards: less is more. Three people — two founders plus one VC — is ideal; five is the workable norm; every single member must be excellent, because each one shapes the company.
• Dilution benchmarks at IPO: Google's founders were at about 15.6%, Steve Jobs at 13.5% of Apple, Mark Pincus at 16% of Zynga. Staying above 10% through many rounds is a very good outcome.

The overlooked alternative is not raising at all. Craigslist would plausibly be worth around $5bn if run as a normal company; GoDaddy and Trilogy took no outside investors; Microsoft took a single small venture check just before its IPO — which is why Bill Gates still owned 49.2% when it went public. And whoever's money you do take, the closing question is the same one you ask about co-founders and employees: are these the people you want permanently tied to your company?

Then vs. now (2026)

Self-check quiz

Pick an answer to reveal the explanation.