Class 2 Notes Essay — Reid Hoffman, John Lilly, Chris Yeh, and Allen Blue’s CS183C Technology-enabled Blitzscaling course.

Here is an essay version of my class notes from Class 2 of Stanford University’s CS183C — Technology-enabled Blitzscaling — taught by Reid Hoffman, John Lilly, Chris Yeh, and Allen Blue. Errors and omissions are my own. Credit for good stuff is Reid, John, Chris Yeh, and Allen’s entirely.

This class was an interview by John Lilly of Sam Altman about the first stage of growth — the family stage. The below is my paraphrasing of their conversation. To recap from last class, the family stage — is the stage where small teams are building a product to get product market fit.

The video of the class, has just been published. (My notes are below)

I. Interview with Sam Altman of YCombinator

John Lilly: What is YCombinator?

Sam Altman: YCombinator is an organization that funds startups, which is a great way to enable innovation around the world. Right now we invest in 250 companies a year through our main program, and we have started a YCombinator Fellowship program which we hope to scale far beyond this amount.

Our goal at YCombinator is to create a company which helps startups get a 100x leg up of where they would be without any help. We do this through advice that matters, connections, and our alumni community. Since starting YCombinator we have invested in over 2,000 companies and our alumni have a strong affinity towards helping one another.

John Lilly: How did you first get involved with YCombinator?

Sam Altman: I dropped out of Stanford University to start my first company, Loopt. I started working on the company around the day the first YCombinator batch was announced and applied the day before the applications were due.

At that point in time (2005) the world of startups seemed mystical. To get investment it seemed like you needed to know the right people, know the right words to say, be a business person, and have a business plan.

When I first talked to YCombinator I knew that these were our people — Paul Graham was very much an engineer himself. They understood the value of product and understood what we were doing. It took YCombinator a couple years to realize how special this was, but from the perspective of the startups in the first batch we knew that this was special.

John Lilly: There was a post written about YCombinator called: YC Is A Network Effect Business, what do you think about this?

Sam Altman: I get asked this question a lot. Currently there are around 2,500 accelerators in the world. In all of the accelerators — 8 — $1B companies have been created and 100% of those have come from YCombinator.

Investment is a network effects business and this fact is even more true in the early stage side. The value our founders get from YCombinator is very much the network you get from our alumni — not strictly our advice.

John Lilly: What makes up the YCombinator alumni network and how do you think about the network?

Sam Altman: The first thing to understand is we are the first investor in many of the companies we fund. There is a strong affinity towards the first investors who believe in an idea first before anyone else. This loyalty decreases over time — strong affinity towards seed investors, good affinity towards Series-A investors, less affinity towards Series B/C investors, and no affinity at all to public investors post-IPO. Once you IPO’ed people will sell your stock at a dip in your quarter.

John Lilly: How was it taking over YCombinator as the managing partner vs. starting something from scratch (ex. his experience at Loopt)?

Sam Altman: The one thing to realize is I was helping out YCombinator from the beginning. I was there during the the start of YCombinator, I was there when the first check was given, I was advising companies for fun on my own time, I became a part time partner in 2011, and then in 2012 my company was acquired.

At that point in time I actually went out on my own and raised my own venture fund. This was a pivotal moment for me because I realized I didn’t want to be a traditional venture capital investor — I wasn’t having fun just allocating capital. I had the most enjoyment when I was investing $15K checks, and the companies I was investing in wouldn’t have been around if it wasn’t for me.

I wanted to enable new companies and looked for an opportunity to do that. I wanted to run a company and pick the company that gave me the most impact in doing so, and YCombinator seemed like a great fit for me.

It was really hard to take over from someone like Paul Graham but another thing to realize was YCombinator was still relatively small when I took over. I already knew the organization so well and was able to make a relatively smooth transition.

John Lilly: One of the things Reid Hoffman talks about is when Jeff Weiner took over LinkedIn in 2008 — it was considered the second founding of LinkedIn, do you feel similarly for YCombinator?

Sam Altman: With a new CEO/leader I believe it’s always a re-founding of the company. We’ve done things that looked dramatically different from the outside (greatly increase the number of startups YC is investing in / fellowship program / etc) but from the inside all of this has stayed true to our mission. What we are trying to do is find the best founders and even though we are doing this in different ways, internally it’s still the same mission.

II. Questions Around Founding Teams

John Lilly: Switching gears to talk about what you think about startups — given you see thousands of startups a year — what do you think about the initial team of startups? What consistent qualities do you see in the best founders?

Sam Altman: A few things in no particular order:

  • Clarity of vision — can the founder explain what they do and why they do it. If they can’t explain this clearly to us — it would be very difficult to recruit the initial team, get the press excited, and it also means they aren't a very clear thinker.
  • Determined and passionate about what they are doing — we look for founders who don’t take no for an answer and bend the world to their will. There are so many problems you face when starting a company, if they can’t solve them there is no way their startup would work. In terms of being passionate the best companies are always mission oriented.
  • Raw intelligence
  • Ability to get things done quickly — this isn’t the same as determination. We have seen speed is a good approximation for early startup success. We have run into many founders who look great on paper but lack this quality.

John Lilly: Going back to your first point how do you have clarity of vision if your initial product hasn’t been figured out yet?

Sam Altman: I’ll give you an example. With Airbnb they changed a few times but they always had clarity of what they were trying to accomplish. The founders shared a story about how their grandparents used to travel and stay on locals airbeds and this method was so much better than staying at a hotel. All of the hotels look the same — even the high end ones — but locals could give you a unique experience. While their idea took a long time to evolve their idea was very clear on why they were working on this and what they needed to work on in the short term.

John Lilly: How do you judge for these qualities in new founders?

Sam Altman: One of the benefits of looking at 20K applications is we get a lot of scale and data to be able to understand the right kinds of questions to test for things.

John Lilly: What do you guys think of solo founders vs. 2–3 person founding teams?

Sam Altman: YCombinator prefers 2–3 person founding teams but isn’t against solo founders. There are two main issues with solo founders 1) startups are a lot of work to do for one person 2) psychologically doing a startup is very tough for any one person to handle. It’s possible but very difficult.

At the same time having a bad co-founder is significantly worse than having no co-founder. We see this a lot where a solo founder doesn’t have a co-founder and just pulls a random person onto their team. This never works out.

In general the best companies tend to have 2–3 co-founders on the initial team. If you look at the companies worth $1B, I don’t think any of them have been started by solo founders or at least ones who haven’t added on a co-founder within 6 months of starting the company.

One other thing we have seen is 3 co-founders is doable, 4 co-founders is tough. In these cases we normally see at least one co-founder leave within the initial startup period.

John Lilly: How do you tell when co-founders aren’t getting along?

Sam Altman: All co-founders get into big fights at some point, sometimes they tell us and sometimes they don’t. We’ll try to help them work through it, but there are some times when we collectively say that the relationship is done, and both sides need to move on.

John Lilly: How do you think about the next couple of team members? And also how do you think about diversity of the team?

Sam Altman: Diversity of backgrounds is good, but diversity of vision is really bad. This isn’t about what product you are building but the type of company you are building — how big do you want the company to be, what are your working styles, what to do about disagreements, how do you think about hiring and culture of your company.

The biggest kinds of conflicts we see is not when people want different things but when they want the same things for themselves. this includes when both people want to be the CEO, when both people want their face to be on the cover of the magazine, etc.

Complimentary teams is the goal when people have the same vision for the company but bring very different skill-sets to the table. The classic example of this was one strong business person and one strong tech person (Jobs and Woz). However this have been shifting towards two strong technical people — which can work but also brings a higher chance of conflict.

John Lilly: How do you help people hire?

Sam Altman: Hiring has always been fraught with peril —and it is even harder now in this environment.

What we have seen is the best startups hire the least and the slowest. The worst startups say look we have 10 people working for us — the downside of this is very high burn, more organizational inertia, people thinking about career trajectories, etc. These things are important but only necessary when you are ready to scale — you don’t need them when you aren’t ready to scale.

For example

  • AirBnb took 9 months to hire their first person
  • Dropbox took “a long time” to hire their first person
  • Stripe took 10 months to hire their first person

Hiring more people gives your company more status but also makes it harder to do things.

John Lilly: What do you think about starting a company in the SF Bay Area vs. other places now?

Sam Altman: The SF Bay Area is the best place in the world to start a startup. If I was going to start a startup, I would start it here.

The one thing that makes this tough though is the high cost of living. Early stage startups have a huge benefit towards keeping their burn rate low because each month you can survive means each month you have a higher chance of succeeding. The high cost of living makes this difficult.

Another thing I have noticed is the SF Bay Area has gotten a lot less diverse over time. We’ve looked at a lot of companies outside of the SF Bay Area and they tend to be much more diverse than here. We have a big nationality spread but not a good gender spread. In terms of YCombinator we have 50% of our partners who are women and we try to reach out to as broad of a set of founders we possibly can.

III. Questions Around the Idea

John Lilly: Paul Graham said it’s best to find ideas that other people think are bad, what do you think of this?

Sam Altman: The goal should be to find the really good ideas that look bad — aka the undervalue assets — and work on those. These types of ideas have led to outsized returns for venture capital investing and for YCombinator.

This is much harder to do in practice than in theory, because we are all much more affected by what others think than we think. It’s hard to have conviction with an unpopular idea. The temptation to like what other people like is very strong.

The number one thing is we don’t want to invest in a derivative idea of a company which is big already. For example:

  • In 2006 everyone was trying to invest in the next Facebook
  • In 2009–2011 everyone was still trying to invest in the next Facebook
  • The companies to look at during the 2009–2011 time period were Uber and Airbnb — which did not look like Facebook at all.
  • The next $100B company will likely not look like Uber at all.
  • The goal should be to think about, find, and start the next $100B company which won’t look like the current companies at all.
  • However this idea is going to look bad now so you need the internal conviction to pursue that idea because everyone is going to tell you to start an on-demand company right now.

John Lilly: How have you seen founders keep their internal conviction?

Sam Altman: The quicker you are to launch and get users who love your product — the easier it is to ignore the haters.

When Uber launched, it was weird but they had a good install base, they had a core group of users who loved the product, so they could ignore the rest.

If the reason you have conviction is because you believe a fundamental shift has occurred in the world (aka for Uber — everyone had an internet device in their pocket) this helps keep your conviction in the early days.

You can ignore everyone else except users. If you can’t build a product which people care about, love, and use on a constant basis — then you are done.

John Lilly: Since you have taken over YCombinator, you guys have made a lot more investments in hardware. Do you think this is the next big area?

Sam Altman: People thought we would never understand hardware startups because they don’t look like internet startups.

Our belief is that hardware startups should look more like internet startups. We brought the hardware startup founders into the same YCombinator program as the internet startups and we have seen that the hardware founders compete just as strong. No one wants to see their peers grow while they are falling behind.

The trick for hardware startups is to figure out exactly what they want to measure — and then grow that number by 10% each week. If you can compound your growth of 10% each week for 3–4x years then you become a really big company. The key lesson is compound interest is a very strong force.

It’s still too early to tell but hardware feels like it’s working for us. Once a hardware startup we invested in becomes a large company, no one will criticize us for not knowing what we are doing.

IV. Questions About Pivots

Sam Altman: I feel Silicon Valley gets the whole pivot thing wrong.

We should be tolerant of failure but realize that failure sucks and we shouldn't do it again. Now at startup parties failure is being over glorified and people are flaunting the failures they have achieved. Failure shouldn’t be held against you but it shouldn’t be glorified.

There are two types of pivots which I have seen work:

  1. A pivot is the product the founder wanted to build the whole time.
    For example the Instagram founders originally built an app called Burbn but when it wasn’t working out Kevin (one of the original founders) really wanted to build a photography app because he loved photography. This became Instagram.
  2. A pivot is a result of discovering the right thing to build, from the first version.
    For example the Airbnb founders were building another startup which didn’t work and were using credit cards to finance their startup. Their rent was due and they didn’t have enough money to pay their rent, but they had an extra bed in their house. So they rented out the spare room in their house in order to pay the rent.

The pivots I have never seen work is when the startup isn’t working and the team decides to sit in front of a white board and figure something else out. The problem with this is the team is in a rush to find a new idea and it won’t be good.

I suggest that if your startup isn’t working you should:

  • Shut the company down
  • Return the money you have left to investors
  • Travel the world
  • Come back when you have an idea
  • Work on that

V. Career Advice

Sam Altman: If you are thinking about joining a startup I would suggest:

  • Go join the most successful startup which is growing the fastest which you can get into.
  • The lessons you learn from working in a great company that is scaling will be invaluable.
  • This is the best training for learning to do your own startup, better than anything else.

VI. Product

John Lilly: What does focusing on product mean to you?

Sam Altman: — If you are really good you can work 70–80 hours a week — the question is where should you allocate those hours too.

You should be allocating them 100% to talking to users, creating product, and making sure people love your product.

At every startup’s trajectory there will be a point in time where you have to make a choice to build something which lots of users use but only like the product, or build a product which a small group of users use the product but they LOVE the product.

You should always focus on the users who love your product, not just like it. All of the great companies I have seen so far have followed this path.

John Lilly: What is the YCslump?

Sam Altman: —After demo day, things change. Startups work really well when the founders have the gas pedal down to the floor but this also creates a lot of debt which needs to be cleaned up.

Many founders think they can focus on growth, stop and clean up the debt, then go back to focusing on growth. This never works.

There is a famous quote which says “growth solves all problems.” This isn’t entirely true but when you stop focusing on growth that causes many more problems.

The trick is to take some time off to clean up the debt, but don’t take all of your time off growth. This is another reason why having employees makes this transition harder because employees can sense a change in direction faster than the founders.

The reality is there will be crises at every stage, this only gets worse over time. The trade off is to get the big things right — and let all of the other things be a little wrong.

There is another concept which I call fake work. A good way to demonstrate this is to:

  • Look at the current (“Unnamed”) TV show about Silicon Valley
  • Think about what work would make for good TV
  • Do none of those things.

Sit in your desk, work on product, talk to users, etc. These things would all make for boring TV but are the most important things to do. Things that would be on a TV show are fun but don’t get you anywhere.

VI. Lessons Learned

John Lilly: What have you noticed the best founders screw up on?

Sam Altman: One red flag is when founders fall in love with their public image. They worry about being on the 30 under 30 list, they speak at conferences, and go to networking events.

Gradually they start spending more time outside the office than inside the office and fall away from the things which matter.

Secondly, all great founders screw up on this second point.

Waiting too long to fire people.

I believe ever founder does this because it’s not a teachable lesson. You have to go through it yourself and learn from it yourself. It sucks that you made a mistake and you have to go through this terrible thing of firing your employee.

John Lilly: This could be because most founders think they can fix things and think they can fix the situation when an employee isn’t working out.

Sam Altman: The best way I have found to think about this by knowing that you are in fact hurting the employee more by keeping them in the position. People know when things aren't working out and it’s better to end it sooner and help them find something else they are a better fit for. It’s much better to have a 1 month gap in a resume than a 1 year screwup.

This was the end of the interview for the first class. I’ll post the third lecture notes here after our next class.