Notes on Stanford’s “How to start a startup class”

[update: I’ve added lecture 4’s notes, and compiled all notes in a single page here.]

I have been watching Sam Alman (Y Combinator president) brilliant attempt at explaining the unexplainable: the art of starting a business. It is great to listen to people able to put in words what one has done instinctively, sometimes right, often wrong. Lots of great lessons for entrepreneurs and wanna be startupers. Here are my notes & videos from the first three lectures.

Lecture 1 – Sam Altman, Dustin Moskovitz (video)

– startups = idea + product + team + execution + luck (from 0 to 10’000)
– idea not much importance, but still. Bad idea can derail you.
– having a mission is better, cause it will be easier to federate people around it
– copycat companies don’t excite the public or the team
– good ideas can look terrible at the beginning (ex: Facebook: social network for moneyless students)

– find a small market inside which you can have a monopoly and quickly expand
– first version of your product doesn’t have to sound very big. Product can be imperfect but has to improve rapidly
– investors make the mistake of only thinking of the growth of the startup, while they should look at the growth of the market
– you can change everything in a startup but the market
– ask “why now” (Sequoia question), why not two years ago, why not in two years?
– two things you should be doing: build a great product, talk to your customers
– most startups don’t die of competition, but because they don’t do something that people love
– do sales at the beginning cause you need to understand client needs. Great founders don’t put anyone between themselves and their users. Don’t hire sales and customer support people at the beginning
– start with a simple product
– build something a few people love, and expand to a lot of people (orange in the graph below), rather than build a thing a lot of people like, and expand to love (grey in graph)

– iterate around the same loop:
1) user feedback
2) product decision
3) show and test
– startups do what founders measure: measure growth indicators like total registrations, active users, activity level, cohort retention, revenue, net promoter score (

– you’re responsible for opportunity cost of the time of people who follow you
– you’re always on call
– fundraising
– unwanted media attention
– you’re more committed, i.e. less option value
– number one thing is to manage your own psychology

Phil Libin (Evernote CEO): “People have this vision of being the CEO of a company they started and being on top of the pyramid. Some people are motivated by that, but that’s not at all what it’s like. What it’s really like: Everyone else is your boss–all of your employees, customers, partners, users, media are your boss. I’ve never had more bosses and needed to account for more people today. If you want to exercise power and authority over people, join the military or go into politics. Don’t be an entrepreneur.”

Lecture 2 – Sam Altman (video)

– Use your intuition
– watch what younger generations are using

– co-founders are the most important part of startup teams. Co-founder fights one of the main reason startups fail
– choosing someone that you’re not friends with usually ends up in disaster
– be a student or work in a cool company like google or Facebook cause it has lots of co-founders
– much better to have co-founders than solo founder
– qualities of co-founder: unflappable, tough, calm, creative, decisive, ready for anything: James Bond 😉
– If you’re not technical get a technical co-founder
– know your cofounders or early hires for years
– 2-3 co-founders good, 4 or 1 usually does not work, 5 really bad
– skeptical of remote teams: for early days, communication and speed are key. Don’t work remotely, really really tough.

– it sucks to have a lot of employees (high burn rates, slow decisions, complex management). Try to have very few employees in the first year. Should definitely hire (in the early days) when dying need. Then when growth comes you’ll need to hire fast. Very bad first hires can kill the company.

– best source of people: people you already know or your staff already knows
– at Facebook or Google, HR sits down with new hires and goes through their network for good people
– look outside your geographical boundaries, bring people from the outside

– for early hires: go for attitude over experience. Most good hires are people who have never done it.
– attitude: good communication skills, manically determined, like a little bit of risk,  people who you would feel confortable reporting to.
– mediocre people in a small company will kill a startup
– lose a mediocre potential hire (who can kill the company) over losing a client (you can recover from that).

– took 5 months for airbnb to hire first employee
– 5 questions:
1) are they smart?
2) do they get things done?
3) do I want to spend a lot of time with them?
4) what are concrete projects the person have worked on?
5) call references, people who worked with them in the past. Really dig in: are they in the top 5% you worked with? would you hire them back? what are they good at?
– many people are bad at hiring, but good at assessing a person after they have worked with them. Try to work with people on a project before you hire them
– have a list of values and make sure people adhere to them
– have an extremely high bar, hire slowly
– founders underestimate how hard it is to get the very best people
– founders should spend 0 to 25% time hiring people.

– employee equity: you should aim to give 10% of the company to the first 10 employees. Founders are generous with equity to investors, less with employees. Totally wrong. Employees have more value over time, while investors write a check and tend to vanish
– At YC, companies who have been more generous with equities to employees are the most successful

– set cofounders equity very soon after you start working together. If you don’t want to give people the same equity than you, you should think hard about whether you want them as a cofounder
– pre-negotiate what happens if one of the founders leaves. In Silicon Valley, it takes 4 years to own all your equity. So if you leave after one year, you get 25% of your equity. If you don’t do that, deadweight in equity table will make it hard to find investors.

– keeping people: founder thinks he/she’s doing things the best but need to learn to delegate and trust people
– praise your team for things that work.
– Get people new areas of responsibilities
– be aware that as a first time founder you’re likely to be a bad manager

– fire fast: firing people is one of the worst part of running a company. Every first time founder waits too long.
– fire fast: better for company, better for employee
– fire people who are bad are their jobs, create politics, are negative. Completely toxic to the company. Might work in a large company, will kill a startup
– one or two fuck ups -> ok, team spirit. If somebody consistently fucks up -> fire them. Fire those who are consistently doing bad things over weeks, and people doing the opposite of what you would do.

1) set the vision
– focus
– say no a lot
– have a few clear goals
– communicate on goals so whole company is aligned
– always focus on growth and momentum, see how you’re doing against metrics
– don’t get caught up in the PR, can make you feel like you’ve made it while you’re not
– be in the same space, remote slows down the cycle time
– indecisiveness is a startup killer
2) raise money
3) evangelize
4) hire and manage
5) make sure the entire company executes (intensity)

– being a founder is signing up for years of grind around execution
– work hard, pay attention to details, care for customers
– company needs to see CEO as maniacal execution machine
– ideas aren’t worth anything, execution is where the value is
– most good funders have a small number of goals for the company. Everybody in the company should be able to say each week what the goals are.
– always keep momentum: always keep growing, don’t let your ship date slip
– relentless operating rhythm (Facebook: “move fast and break things”)
– obsession with execution quality
– every time you talk to a good team, they have gotten new things done
– when momentum is down, get small wins. “Sales fix everything”

– ship products, launch new features
– review/report metrics and milestones
– repeat

– don’t worry about a competitor at all until they really beat you with their product
– don’t let your company get down because a competitor is in the press
– Henry Ford: “the competitor to be feared is one who never bothers about you at all, but goes on making his own business better all the time”.

Lecture 3 – Paul Graham (video)

– startups are very counter intuitive, it’s an area where you can’t only rely on your intuition
– many people ignore YC advice because they seem wrong
– you can however trust your instinct about people. Big mistake is to ignore these intuitions about people, because they seem impressive, etc
– pick people as if you were picking people to be friends with

– what you need to succeed in startups is not an expertise in startups
– what you need is expertise in your own users
– one of the risk of young founder is they will go through the motions of starting a startup, and neglect the one thing: making something people want
– universities can teach you about startups, but it’s not what you need to know. What you need to learn is the needs of users, and you can’t learn that before starting the company. You can only learn about startups by doing it.

– starting a startup is where gaming the system stops. Can work in a large company (like sending late emails to pretend you’re working evenings). In startups that doesn’t work. There is no boss to trick. All users care is that your product works, you can not trick them.
– faking does work with investors. You can foul them for one or two rounds, but you’re wasting your own time cause all you’ll do is lose your time sending your company down in more time than it would have taken.

– if it succeeds it will take over a long time, 10 years or your own life
– being the boss of a very successful company has drawbacks: there are many things only you as the emperor can deal with. As the company daddy you can never show fear or weakness. As a billionaire you get no sympathy when complaining about stuff.
– as startup grows, it never gets easier. New kind of problems, but total volume of worry never decreases, if anything it increases. It’s similar to having kids, it never stops.

– don’t start at 20, travel and enjoy life first because you can’t do it later, and you’ll learn a lot
– success takes a lot of serendipity out of your life. You run your company as much as your company runs you. Serendipity gives you more options.

– you probably haven’t done a startup when you start one
– as an investor, easy to predict if someone is smart. Hard to predict how tough and ambitious they are
– in the army, you can’t tell who is going to be successful between the arrogant and the quiet recruits. Same for entrepreneurs.
– if you’re terrified of starting a startup, don’t do it, unless you’re someone who is fuelled by fear and excels under pressure

– the way to come up with good ideas is not to make a conscious effort, and take a step back
– twitter, google, yahoo, apple, facebook were all side projects, rejected by the conscious mind as ideas for companies
– how to turn your brain into the kind that has startup ideas unconsciously:?
1) learn a lot about things that matter
2) think of problems that interest you
3) with people you like and respect (how you get cofounders at the same time than the idea)
– get yourself on the edge of a technology, “live in the future”
– example: student who built voice over IP because he wanted to talk for free with his girlfriend

– role of non technical founder? run the business side of things. For ex: Uber, non technical founder will bring domain expertise, and run sales.
– any value in business school for entrepreneurs? Not really. What business school was designed for was to teach people management, a problem you only have in startups if you’re sufficiently successful. You might be better off going to design school. Best way to learn is to start something. You might fail but you will learn.
– ideally, you’re successful before you hire your first employees. So you don’t need to manage before you have traction.
– are we in a bubble? There is a difference between high prices and bubble prices. In bubble years, VCs knowingly invested in bullshit trying to unload it on retail investors before it would blow up. That is now what’s happening today. Prices are high, it won’t be as easy as today. Bubble right now? no.

I am an entrepreneur and researcher passionate about understanding the social implications of digital technologies.
1 comment
  1. […] class at Stanford, featuring Adora Cheung, founder of Homejoy. Notes from class 1-3 are here. I also consolidate all notes into a single page […]

Leave a Reply

Your email address will not be published. Required fields are marked *