How to Start a Startup - Lecture 5 Notes

October 7, 2014

I’m taking down notes for Sam Altman’s class, How to Start a Startup and I figured I’d start sharing them. This is for the fifth lecture with the following notes:

Strategy & Competition

Capturing value

What makes a business valuable?

Perfect competition


The lies that people tell

(diagram of two firms to show perception)


Example 2:

Example 3 (Startup version):

Example 4 (Large enterprise):

How to build a monopoly

Example: Amazon started as just a bookstore. It has all the books in the world. It’s online, there are things on it that you couldn’t do before. Gradually expand to different kinds of ecommerce.

Example: Ebay started selling pez dispensers -> beanie babies -> auctioning of different kinds of goods.

Example: PayPal started with power sellers on ebay. It was a small market of people selling junk on the internet.

Start big and then shrink

Last mover advantage

Characteristics of monopolies:

History of innovation

Psychology of competition

Q. Monopolies and perfect competition often look similar. How do you easily tell the difference?

Ans. Ask the question: what is the actual market?

Q. Which of the aspects of monopolies that you mentioned, do you think software companies like Google excel at?

Ans. They have network effects, they had the proprietary technology that gave them the initial page rank algorithm which was an order of magnitude better. They had the economies of scale with the need to store all these sites. So Google basically had all four.

Q. undistinguishable question

Ans. Palantir started with the intelligence market that used a very different approach.

Q. What do you think about lean startups, iterative thinking where you get feedback from people vs complexity that may not work?

Ans. Personally sceptical of all lean startup methodologies. The really great companies did something more of a quantum improvement that really differentiated them from everybody else. Typically, did not do massive surveys.

If you take enough time to figure out what people want, you most often will have missed the boat by then.

Q. If you were someone who worked at a high paying company for 6 months, left to Stanford in comp sci. How would you recommend rethinking .. (undistinguishable bit)

Ans. There isn’t any easy psychological formula to avoid it. Never under-estimate how big the problem really is.