This governs one decision sitting underneath every other decision in a business: whether you are copying a model that already works or building something that does not yet exist. Thiel's core claim is that competition and profit are opposites, so a business that competes well still ends up with nothing. The work is to find the one thing you can do that nobody can substitute for, then start in a market small enough to actually own it.

Do

Write down your contrarian answer before you write a plan. Force the sentence into the form "most people believe X, but the opposite is true." If you cannot complete it, you are building on knowledge that is already agreed upon, which by definition gives you no edge. [Source: "Chapter 1: The Challenge of the Future"]

Start in a market small enough that you can take all of it. PayPal spent its early effort chasing millions of scattered PalmPilot users and got nowhere, then took roughly a quarter of eBay's few thousand highest-volume sellers in three months. Pick the smallest group you can dominate completely, then expand outward from a position you already hold. [Source: "Chapter 5: Last Mover Advantage"]

Require a tenfold improvement from yourself, not an incremental one. Anything short of 10x reads to a buyer as marginal, and buyers arrive with years of exaggerated claims making them skeptical. Solyndra's cylindrical cells were physically less efficient per unit area than flat ones, a deficit no amount of added mirrors could close. [Source: "Chapter 13, Part 1: Seeing Green and the Seven Questions"]

Commit to a definite multi-year plan and hold it when the market disagrees. Apple's sequence of iPod, then iPhone, then iPad was dismissed by analysts at each launch and worked anyway, because it was planned rather than reacted to. Zuckerberg turning down Yahoo's billion-dollar offer in 2006 is what a founder with a real plan looks like from the outside. [Source: "Chapter 6: You Are Not a Lottery Ticket"]

Pick one distribution channel and get it fully working before adding another. Distribution obeys a power law: most companies get zero channels working, and one working channel is enough for a great business. The kitchen-sink approach (a little advertising, a little sales, a little virality bolted on afterward) is a common way to die. [Source: "Chapter 11: If You Build It, Will They Come?"]

Settle ownership, control, and daily possession before you begin, not after. Choosing a co-founder is closer to marriage than to hiring, and a startup broken at its foundation usually stays broken. Keep the board at three people, five at the outside, and keep everyone involved full time and holding equity. [Source: "Chapter 9: Foundations"]

Recruit by answering why this specific person should join you and not somebody bigger. Generic answers about smart colleagues, valuable stock, and important problems fail because every company says them. The only durable answer names what makes your mission uniquely important and why this person specifically fits it. [Source: "Chapter 10: The Mechanics of Mafia"]

Concentrate on the one bet that could carry everything else. In a good venture fund the single best investment beats every other investment combined, and a person cannot diversify a single career any more than a founder can run several companies at once. The advice to spread your eggs across baskets is folk wisdom, not strategy. [Source: "Chapter 7: Follow the Money"]

Hunt for your secret in a field that matters but nobody has standardized. Thiel's example is nutrition: it affects everyone, is barely a real university major, has few rigorous recent studies, and produced official guidance shaped more by lobbying than by evidence. Look where importance and institutional attention have come apart. [Source: "Chapter 8: Secrets"]

Don't

Don't define yourself by the company you are trying to beat. Microsoft and Google started in genuinely different businesses and then spent years mirroring each other feature for feature, while Apple passed both and exceeded their combined market value by January 2013. Rivalry makes combatants more alike, not more differentiated. [Source: "Chapter 4: The Ideology of Competition"]

Don't redraw your market boundaries until the numbers flatter you. Companies without a monopoly define their market absurdly narrowly to appear dominant, the way a British restaurant in Palo Alto can claim to own the market for British food specifically. If you have to shrink the definition to look like a leader, you have already answered the question. [Source: "Chapter 3: All Happy Companies Are Different"]

Don't treat flexibility as a substitute for having a plan. The four dogmas that came out of the dot-com crash (be incremental, stay unplanned, improve on competitors, ignore sales) were reactions to one specific historical mistake, not timeless wisdom. A bad plan beats no plan. [Source: "Chapter 2: Party Like It's 1999"]

Don't justify the business with a reason everyone already accepts. Cleantech pitches nearly all rested on the same socially approved motive, that the world needs cleaner energy, which is exactly what made them interchangeable. A widely shared justification is evidence you have no secret, not evidence you are right. [Source: "Chapter 13, Part 1: Seeing Green and the Seven Questions"]

Don't try to automate the human judgment out of your product. PayPal was losing over $10 million a month to card fraud and pure automation kept failing, because human fraudsters adapted faster than static rules could. The system that worked had software flag suspicious transactions and human analysts make the final call, and it carried the company to its first profitable quarter. [Source: "Chapter 12: Man and Machine"]

Don't pay yourself a large cash salary early. Thiel's rule is that CEO pay at an early-stage venture-backed startup should not exceed $150,000, because high cash compensation turns a founder into someone defending the status quo rather than building past it. Equity, not salary, is what ties anyone to the long term. [Source: "Chapter 9: Foundations"]

Don't read current growth as durability. Zynga could not reliably repeat a hit after Farmville, and Groupon struggled to turn one-time merchant trials into repeat business. Weekly actives and quarterly revenue can conceal the absence of anything that holds for a decade. [Source: "Chapter 5: Last Mover Advantage"]

Don't start believing either the praise or the jeering about you. The founder archetype attracts fame and infamy at once, and the single greatest danger Thiel names is a founder growing so certain of his own myth that he loses his mind. The equal and opposite failure is a company that loses all sense of mission and calls the disenchantment wisdom. [Source: "Chapter 14: The Founder's Paradox"]

The one line

Stop trying to win the market everyone is already fighting over, find the small one you can own outright, and build the thing that has no close substitute.