In spring of 2012, Peter Thiel, cofounder of Paypal and Palantir and early investor in Facebook, taught a course at Stanford in the Computer Science department called CS183: Startup. One of the students in the class, a law student named Blake Masters, took meticulous notes that were widely shared across the web (1M+ pg views).
The ideas were intriguing, and ran counter to much of the standard startup wisdom. I had read some of the notes when they came out but they’re pretty lengthy and I didn’t get through them all.
Luckily for all of us, the two have collaborated to write a book based on the class called Zero to One, which comes out September 18th.
Having snagged an advance reader’s copy, the book has definitely cut the word length down while retaining the ideas and most of the nuance. Instead of summarizing the book, I thought I’d share my favorite and least favorite idea.
Best Idea: Startups are based around secrets
Thiel has a lot to say about secrets, which he defines as truths that are unpopular or not widely known. Every great startup idea, he argues, is based around a thesis that most people don’t believe is true.
- Twitch: millions of people will tune in to watch other people play video games online
- SpaceX: it is possible for a founding team without significant space or rocket experience to build a company that makes manned space flights better than NASA
- Airbnb: hundreds of thousands of people would be willing host strangers in their homes, and millions of travelers would prefer those homes over a hotel
I think this is a really important point, and one that most casual observers of startups don’t get. If everyone thinks a startup idea is a good one, then it would have happened already, or there’s some non-obvious challenge that prevents it from happening. Only ideas that seem dumb or impossible to most people fall into the realm of ideas that could be potentially correct.
Paul Graham puts this another way: “the best startup ideas seem at first like bad ideas.”
But even beyond the startup idea itself, most of the time founders have a strategy that they can’t reveal because the element of surprise is part of what will make it work. That’s why reporters and the general public are often cofounded by the vague and imprecise plans that founders provide when asked things like “what markets will you expand into?” or “how will you monetize?”. 
Anyway, it’s an important idea and you can read the full lecture on Secrets here.
Worst: Every great startup has to have a grand plan upfront
Thiel believes that the United States has moved from a mindset of determinate optimism (the future will be good and we know what it will hold – space travel, flying cars, robotic assistants) to a mindset of indeterminate optimism (the future will be good but we don’t know exactly how it will play out). Thiel thinks this is a problem because without a grand vision, you only make little bets, and little bets don’t change the world. He argues that one of the big problems with Dot-Com bubble is that Silicon Valley stopped making grand plans and adopted the “lean startup” method of lots of testing and iterating. In the original notes, he says this:
[blockquote]”Airbnb, Pinterest, Dropbox, and Path all have a very anti-statistical feel. There’s a sense in which there’s some telepathic link between these products and what people want. “[/blockquote]
While he’s got a point that entrepreneurs must have a thesis about their product and where the market is going, but I think he overemphasizes this grand plan thing. Sure, you can’t A/B test your way to a vision, but even Airbnb definitely has had multiple re-envisionings of its own product. Before they got into Y Combinator, they literally just wanted to put air beds in people’s houses. It was only after the Sequoia funding, as they realized how big the home rental market was, that they got a bigger, grander plan. And of course Path’s apparent lock on their users preferences appears to have faded since 2012.
Ultimately, the world is more complex now than it has ever been, and even experts making the most basic of predictions can’t beat a “steady state” algorithm. Entrepreneurs need to have a vision, and we’re often told the best way to predict the future is to create it, but it’s also the case that no battle plan survives contact with the enemy. Let’s let founders make awesome things, and not ask them to be fortune-tellers.
Final Recommendation: Get the Book
There are plenty of other thought-provoking ideas that I haven’t touched on here:
- the best startups avoid competitive industries or otherwise become so good that they become defacto monopolies
- distribution for startups matter, and most channels follow a power curve
- we need founders who are a little extreme because that’s what it takes to build a successful company
Thiel is a really smart guy and incredibly successful as an entrepreneur and investor, so even my gripes about the book aren’t particularly frustrating. It’s well-written and definitely worth reading if you work in tech or have any interest in entrepreneurship.
 Of course it might also be because they have no fucking clue, but that’s also understandable and reasonable, since no one has ever done what they’re trying to do, how could anyone truly be sure their plan was the right one anyway?