Zero to One: Difference between revisions
Content deleted Content added
No edit summary Tag: Reverted |
No edit summary Tag: Reverted |
||
Line 47:
🧱 '''9 – Foundations.''' Beginnings fix trajectories: in the universe’s earliest microseconds the cosmos expanded by a factor of 10^30, and in Philadelphia in 1787 the Framers made design choices so durable that only 17 amendments have followed the Bill of Rights since 1791. Founding a company works the same way—choose the wrong partners or early hires and the damage is hard to undo—so “founding matrimony” matters as much as technical skill. To keep people aligned, distinguish ownership (equity), possession (day‑to‑day operation), and control (the board), and watch for misalignment of incentives like the DMV’s gap between nominal ownership and real bureaucratic power. Most conflicts arise between founders (ownership) and investors (control), so keep the board small and careful: three is ideal and it should never exceed five unless the company is public, where the average is nine. Commitment must be binary: everyone involved should be full‑time, with narrow exceptions for outside counsel and accountants; broad part‑time arrangements corrode focus and culture. Structure compensation and roles to minimize drift—clear titles, vesting that rewards staying power, and decision rights that match responsibility. The chapter’s claim is that early design choices lock in path dependence; the wrong structure traps a company in conflicts that later heroics can’t fix. Alignment across ownership, possession, and control is the mechanism that lets a startup compound advantages instead of firefighting politics. ''I stress this so often that friends have teasingly nicknamed it “Thiel’s law”: a startup messed up at its foundation cannot be fixed.''
🤝 '''10 – Mechanics of mafia.''' The PayPal “mafia” is the concrete case: after selling PayPal to eBay for $1.5 billion in 2002, former colleagues went on to start or lead companies like SpaceX, Tesla, LinkedIn, YouTube, Yelp, Yammer, and Palantir—each ultimately worth more than $1 billion—showing how a tight early team compounds far beyond one company. The group wasn’t assembled by sorting résumés for generic talent; it was built by hiring people excited to work specifically with one another on a shared mission. Recruiting is treated as a core competency, sharpened by the litmus test of why a great engineer would become the twentieth hire rather than take a safer job elsewhere. Mission fit is concrete: at PayPal, enthusiasm for creating a new digital currency to replace the U.S. dollar separated true believers from tourists. Perk wars are rejected; instead, cover basics and offer the chance to do irreplaceable work with people who care. Outwardly, a strong culture makes people “different in the same way,” the hoodie-and-logo uniform of a small tribe devoted to the same aim. Internally, role clarity eliminates friction; making each person responsible for exactly one thing reduces conflict and speeds decision‑making. The result is a group that can keep trust intact under pressure, move faster than larger rivals, and extend its relationships long after the first company is sold. Culture is the operating system that enables a small team to pursue a secret together, not decoration added later. Alignment—mission clarity, personal fit, and single‑owner roles—turns trust into speed and sustained output that competitors can’t copy. ''“Company culture” doesn’t exist apart from the company itself: no company has a culture; every company is a culture.''
📣 '''11 – If you build it, will they come?.''' A science‑fiction parable frames the bias: in Douglas Adams’s tale, humanity flees on three ships—the thinkers on the A ship, the workers on the C ship, and the salespeople and consultants on the B ship—only for the B ship to land on Earth, a joke at sales’s expense that mirrors Silicon Valley’s blind spot. The real economy says otherwise: U.S. advertising brings in about $150 billion with more than 600,000 employees, while sales tops $450 billion with roughly 3.2 million people, evidence that persuasion drives outcomes at scale. Distribution is designed, not accidental, and two guardrails—customer lifetime value versus customer acquisition cost—govern which paths work. At the “complex sales” end, SpaceX won multibillion‑dollar NASA contracts within a few years; Palantir’s CEO spends most of each month with customers because seven‑figure deals demand it. In “personal sales,” Box grew by landing small departmental wins—like a 2009 account at the Stanford Sleep Clinic—then spreading across institutions until Stanford University and its hospital standardized on the product. Viral growth has its own math: PayPal paid users to join and to refer friends, buying 7% daily growth and near‑ten‑day doublings until hundreds of thousands of accounts made fees exceed acquisition costs. Distribution itself follows a power law, so one channel typically outperforms all others by orders of magnitude. Every stakeholder—customers, hires, investors, media—must be sold because products do not speak for themselves. Superior products win only when paired with a matching go‑to‑market; the mechanism is focus on a single channel that fits price, cycle, and audience until it compounds. ''If you can get just one distribution channel to work, you have a great business.''
🤖 '''12 – Man and machine.''' The fraud war at PayPal is the named study: in mid‑2000 the company was losing upwards of $10 million a month to credit‑card fraud, so Max Levchin’s team built a hybrid system—software to flag suspicious transactions and human analysts to decide—which they nicknamed “Igor”; by the first quarter of 2002 PayPal posted its first profit after a $29.3 million quarterly loss a year earlier, and the FBI asked to use the tool. That complementarity became a company template: after PayPal’s 2002 sale, Palantir launched in 2004 to pair human judgment with software across intelligence, finance, and public health. Meanwhile computers kept conquering bounded tasks—Deep Blue beat Kasparov in 1997; IBM’s Watson beat Ken Jennings in 2011; Google’s self‑driving cars appeared on California roads; a 2012 Google system learned to recognize cats after scanning 10 million YouTube thumbnails—but each milestone underscored how different machine strengths are from human ones. At scale, globalization is substitution—people competing with people—whereas technology is complementarity—tools that multiply human capability without demanding human resources in return. The most valuable businesses won’t replace professionals; they’ll amplify them, as LinkedIn did by giving recruiters a universal search-and-filter layer used by the vast majority of the field. The practical lens is to seek problems where computers process oceans of data and people make final judgments—fraud, diagnosis, compliance, intelligence, and education. The chapter’s claim is that progress lies between Luddite fear and strong‑AI fantasies, in systems that pair silicon speed with human sense. The mechanism is hybrid design: encode what machines do uniquely well, then build interfaces that surface the right decisions to people. ''computers are complements for humans, not substitutes.''
🌱 '''13 – Seeing green.'''
| |||