Advice for ambitious Stanford students interested in startups
June 2018: If you're ambitious, sign up for this: https://airtable.com/appMvL5omTtoQe8LB/shr1OSItdgX4TIhuY.
What's most valuable — problem selection, people selection, product building, or engineering skill?
"Engineering skill is the most important thing — I better get really good at engineering, and get to know all the best engineers in the school."
No.
Exception: If you can't build a small web app, make it a high priority to build one (suggestion: pick a problem you have, and hack together a solution!)
- The two most powerful inputs on a company's long term value are: 1) problem selection and 2) people selection. Engineering skill, company building skill/operations, and all other skills are far less important. #3 would probably be 3) building a product that people love (includes marketing, seduction, and customer development). Problem selection is more important than building something that people love, because if you select a problem that is improvable and people care deeply enough about, they will force you to build a product that they love ("the market pulls the product out of the startup").
- Clarification: If your goal is to make $5 million, then you can quite possibly do that solely by being excellent at engineering. If your goal is to maximize the odds that you start a company on the scale of Facebook/Apple/Amazon/Google, then you should develop your problem selection and people selection skills.
- Engineering skill for most companies (w exceptions where it is completely core to the product, like Docker, DeepMind, etc) is 0.5x–2x.
- You can have everything to the right of problem selection and people selection perfect, and you will still not create a large amount of value. Optimizing $0 of value is still $0, or perhaps at best $1. This is why it didn't matter that Zuckerberg didn't know better than to incorporate in Florida. Zuckerberg was great at problem selection, people selection, and building a product people loved.
- Engineering skill is important to the extent that it is required to build the first version of your problem. Zuckerberg needed to be able to build the v1 of Facebook, so that it was good enough to get strong product market fit amongst the students of Harvard. Steve Jobs needed enough engineering competence to be able to recruit Wozniak.
- This means that unless the problems you are most interested in solving are hard-core engineering problems, then you should focus on knowing enough to build problems — don't over value engineering skill in the early days — you just need enough to be able to solve the problem for a first group of users. What does this look like in practice? Probably good if you've built 2–3 small web applications, and if you have done OK in one Kaggle competition (machine learning).
Should I work at a tech company?
No, don't work at Google/$techco.
- Working at Google (assuming Goals == create large company) is not going to be value adding, but also not going to be value detracting. Net zero. (So opportunity cost, but not strictly negative by itself.)
Unless:
- It's the most interesting thing to you.
- You want to solve the sort of problems you'll experience working at Google. Things like: business communication for the largest enterprises.
- Or you want to bring problems Google has solved elsewhere. Things like: development infrastructure tools.
Instead, the following can be extremely valuable:
- Seek higher breadth early unless you've found some areas of intense interest/curiosity.
- Even something like the APM program where you work on rotation for 2 years is suboptimal for most. Career wise, great move. But not for solving problems no-one else can. Litmus test: would Elon Musk or Steve Jobs have ever done something like the APM program?
- Does depend on goals. If your goal is to build a company and make tens of millions of dollars, and you're a) not attached to value creation and b) not yet convinced that this is the wrong thing to do from a personal satisfaction standpoint — then the APM program is ideal for you.
If you have an area of deep interest/curiosity
Seek exposure to problems in your area of deeper interest/curiosity.
Excellent. Care deeply / deeply interested in medical industry? Shadow a doctor for a month. Obsessed with robotics? Shadow the owner of a factory that uses robots.
Do things that give you proprietary information. How many people that can build are there that work at Google? Thousands. How many people are there that can build, and have shadowed a doctor or a factory owner for a month? Almost none.
"How uniquely are we developing a personal point of view — a personal approach — a personal set of 'eccentric hobbies, personal lifestyles, and motor behaviors' that will uniquely prepare us to create? This, in a nutshell, is why I believe that most creative people are better off with more life experience and journeys afield into seemingly unrelated areas, as opposed to more formal domain-specific education — at least if they want to create." — pmarchive.com/luck_and_the_entrepreneur.html
Do something that boosts your personal runway to >5 years. Create a product that brings in recurring income.
Good. Short stint at an investment firm.
- Requires: True access. You want to be able to sit in on the important meetings and listen to what the people allocating the capital are saying.
- Benefit: Broad exposure.
- Benefit: Experience towards the skill of having and recognizing good ideas (assuming you're at a good investment firm, or you're at a bad one but you're able to think independently of those around you).
Decent. High variability (i.e. doing a lot of things) role at:
a) Small startup, pre-traction, with excellent people. See Kevin Systrom becoming the first intern at Odeo, before it became Twitter: "I learned so much from Ev and Jack," he says. "The whole thing was really eye-opening for me."
b) Early, extremely high growth (pants on fire level growth, where they are in the stage where everything is breaking, more hands always needed to solve things — this is very, very rare, and don't settle for less!). Gives broad exposure to problems. Extremely high growth means young people usually given outsized responsibility. Not the case with reasonably high growth startups that have much more of their shit together. Unless you can swing some kind of broad exposure role. However the energy required to get a role like this would likely be so high as to be better spent building things.
Exceptions to "don't work at Google/$techco":
- A small team / high information access role that really matches your deep interests and curiosities.
- If your goals don't match the goals here.
Goals here, summarized: satisfaction in life → create value → work on hardest/most valuable things possible. Aristotle's view: "A good life is one where you develop your strengths, realize your potential and become what it is in your nature to become."
Thoughts on various clubs
BASES. Personally, I avoided. There are a lot of people in BASES because they want to be in the "business group" so they can get a good job after college. This is totally fine, but those aren't "my people", so I decided I didn't want to apply.
Reader comment: "I avoided BASES because I could tell Stanford frosh were going to BASES to be part of the 'ingroup' and avoid FOMO. FOMO is a REAL force at Stanford, and be very wary of making decisions based on it."
Should I do StartX/Lightspeed Summer Fellowship?
Be cautious.
It's relatively doable to start a "company" and get accepted into StartX/Lightspeed. This can make it feel like you're being successful (you have a company that's been accepted into some accelerator, possibly been given money), when you're really not maximizing your potential.
You need to have standards higher than those around you. Many people will applaud you in StartX/Lightspeed if you do moderately well. If you want to do more than just moderately well, you need to have standards that are higher than anyone around you, so that you keep going even when those around you are telling you "congrats on the company / getting into Lightspeed / raising $2mm" — this is meaningless validation.
Should I drop out?
The correct approach to deciding to drop out is the Mark Zuckerberg approach.
There are three possible ways to start a company:
- Through personal reflection and lots of thinking, figuring out what you 'have' to do and care about so much that you have to solve it (see 'Idea evaluation checklists'), and then making repeated attempts to solve this problem.
- Create tons of projects, none of which you commit to upfront, until one of them takes off to such a degree that you'd be insane not to work on it.
- Start a company where neither of the above two things are true (the worst option).
You should only drop out when either 1 or 2 is true, ideally both. Do not drop out unless either something has taken off to such a degree you'd be crazy not to do it (this does not mean validation from investors, it means extreme Snapchat/Facebook/etc level growth from users), or where you have been so obsessed with a certain problem for >6 months that you know you'll go insane if you're not working on it full time.
Do not drop out just because you could get investment money. Investment money is much easier to get than a product with true promise, that users really love, in a huge or small but growing market.
Should I build something massive? Should I just go for financial independence?
You might be thinking through your life goals.
"Should I build something massive and try and change the world, or should I build small businesses to give me financial independence?"
I was thinking through that question in freshman year.
Don't worry — the answers will come, and everything in this document applies regardless of which choice you make. (I do think one choice is more correct than the other.)
Suggestion: Read this — waitbutwhy.com/2014/10/religion-for-the-nonreligious.html
Things to know if you're an international student on an F-1 Visa
One of the best startup immigration lawyers in the US informed me of the following:
- Look into the O-1 visa. "Extraordinary alien." People into startups often get this visa type. You can do certain easy things to make getting an O-1 visa easier — for example, writing a book that's a collection of interviews and getting it published can count as one of the O-1 evaluation criteria, but obviously writing a book that's a collection of interviews is much easier than writing a full book by yourself.
- Know that OPT is great, and you get 12 months post graduation where you can work for yourself with complete ease. You don't have to pay yourself a salary.
- During summers, you'll do CPT. It's pretty easy, just get the help of someone in your department. You can apply to jobs before getting CPT, and then once you have the job, get an offer letter (this will be needed for CPT).
Instructions for Summer CPT at Stanford
- Get an offer letter.
- Send the offer letter and a note about how it relates to your major to the person responsible in your department.
- You'll get a permission code from them.
- Enroll in 1 unit CPT for the summer session in Axess.
- Fill out the online CPT form, which routes to your advisor.
- Confirm it with your advisor.
Startup ecosystem primer
VC firms are ranked by 'tiers.' You have tier 1, 2 and tier 3 firms.
Top three firms by perception: Benchmark, Sequoia, Andreessen Horowitz.
Other tier one: Accel, Greylock, Kleiner Perkins.
Other firms to pay attention to:
- 8VC
- Social Capital
- Founders Fund
- DCVC
- Khosla Ventures
- General Catalyst
- Spark Capital
Skills that will be useful
Being able to build the things that you want to build. i.e. basics of webdev, ML.
Networking / meeting people. Summary: Take a genuine interest in people. Read a summary of 'how to win friends and influence people': hubspot.com/sales/how-to-win-friends-and-influence-people-summary.
Opportunities come through people.
Best excuse to meet someone? You have a project you're doing (that you'll do regardless of their help — people like helping where their help isn't required, it's just a benefit; people don't like helping as much if it seems like you'll only do the project if they help) where their advice would be helpful, or when you have a question / thing you've been researching that you really want to know more about where their advice would be genuinely helpful.
Knowing when people are lying and when they are telling the truth. slideshare.net/xamde/summary-of-the-mom-test.
Startup related readings
Also, go on Twitter, and follow @paulg, @balajis, @pmarca, @rabois, @sama, @naval, @EricRWeinstein, @msuster.
CS183 Lectures are useful for startup philosophy: blakemasters.com/peter-thiels-cs183-startup
CS183B Lectures are useful for specific startup skills: startupclass.samaltman.com
CS184 Lectures are useful for the tech component of building a startup.
Background readings. First, for the big picture see Why Software is Eating the World, The Rise and Fall of Personal Computing, and Internet Trends.
Next, look at these articles on Stanford's Facebook class, The Social Network, and Massively Collaborative Mathematics.
Finally, read Startup = Growth and as much of the CS183 notes as you can.
Next step
If you're at Stanford, Cal, or Brown, email anon1@alumni.stanford.edu — I'll happily connect you with other great students. I can also answer further short specific non-easily Googleable questions over email. Please include some data points that make it easy for me to see your ambition / smarts / curiosity / thoughtfulness at a quick glance.