CS183C Session 1: Introduction to Blitzscaling

This was the introductory lecture for the class. You can watch a recording here.

Logistical Notes

• As part of the class, we will be inviting you to an official, private LinkedIn Group for the class, and to a Medium collection for student contributions. Participation in the LinkedIn Group is mandatory, because it will be our primary communications channel outside the classroom. Participation in the Medium collection is optional but highly recommended. To add you to the Medium collection, we will be collecting your Medium usernames.
• Assignments will be posted to the LinkedIn Group as a new discussion.
• Turn in your assignments by posting to the appropriate discussion in LinkedIn, and/or posting to the student collection on Medium. The Medium collection is accessible to the public, so if you do not wish to make your assignments public, only post them to LinkedIn.
• The instructors and TAs will post official information to the class collection on Medium (including lecture notes), and will select the most notable student posts from the student collection for inclusion in the official class collection.
• Before each guest speaker, we will set up a LinkedIn discussion where you can post the questions that you would like us to ask the speaker. While we won’t be able to get to every question, we’ll do our best to get to them.

Content Notes

• The instructors were all Stanford undergrads…though ironically enough, none of them knew each other while they were at Stanford.
• Stanford has incredible resources for entrepreneurship, of which this is a small part
• CS183C is a spiritual successor to CS183A (“Zero to One”) and CS183B (“How To Start A Startup”)
• Your mileage may vary!
• Building a great company is not a simple process; there are heuristics not rules. This class focuses on the big concepts.
• We need a specialized term for the process of scaling up startups into massive businesses. Our current term is “blitzscaling,” and we are open to suggestions.
• The mythos of Silicon Valley is that it is a land of startups
• This is partially true. Startups are critical.
• However, you can assemble 10–20 smart people to start a company in many places around the world. Silicon Valley is no longer special in this regard.
• Yet, Silicon Valley still produces a disproportionate number of very important companies. The secret sauce is the ability to scale quickly. Not just startups, but scaleups.
• When you look at $10 Billion publicly traded companies, they are all from either Silicon Valley or China.
• Even when you look at unicorns (privately held companies valued at over $1 billion), Silicon Valley has a disproportionate share (followed by China)
• The Bay Area (7 million people) generates half of the Unicorns in the world. There’s something special happening here. And it’s accelerating.
• We’re going to bring in guest speakers who have actually done this to illustrate how this works in practice in Silicon Valley.
• So why the intense concentration? And why does it persist? Networks.
• Scaleups depend on networks — of talent, capital, users, etc — that exist outside the organization itself. Networks in Silicon Valley work very well, and amplify what startups are trying to do. We have a 60-year lead on places like New York — just look at the names on all the buildings on campus.
• The reason we see so much blitzscaling is that we live in the Networked Age.
• The key themes we’ll talk about over and over:
• Networks
  o Positive feedback loops
  o Distribution advantage
  o Organizational, Revenue, and Customer scale
• The types of scale interrelate; while you may want to minimize the organizational scale, as you increase revenue and customer scale, you will be forced to increase the size of the organization.
• As you scale an organization, the answers to the key questions keep changing with the orders of magnitude of scale.
  o For example, competition doesn’t matter much when you’re starting out, but when you’re an established business it can be critical.
• If you don’t win the current game, you won’t get a chance to win the next game. But you can look ahead….
  o Aneel Bhusri and Dave Duffield interviewed every Workday employee until 500 — this wasn’t scalable, but acted to preserve the culture
• The Five Stages of Scale: Family, Tribe, Village, City, Nation (see slides for complete definition)
  o 3 founders is totally different from 15 people, which is profoundly different from a 100-person company
  o Some companies break the mold — WhatsApp had 19 people and 600,000,000 users. But usually, they correlate
  o We’ll spend roughly two weeks on each stage, with a focus on Tribe, Village, and City (1 lecture, 3 guest speakers per section)
• The Key Functions of the Organization
  o People
  o Product
  o GTM
  o Tech
  o Strategy
  o Operations
  o Finance
  o Management
  o Hiring
  o New Functions (e.g. Corp Dev, which is useless for OS1 and critical for OS5)
• “If you’re not embarrassed by your first product release, you’re probably going too slow.” (Reid Hoffman)
  o This is all about the importance of speed
  o For young consumer internet companies, you need to get things out and fix them along the way…but if you keep doing that at Stage 4 or 5, you’ll die.
• Most startup content is focused on OS1; we’re trying to help you think about how to fill out the entire board so that you can successfully grow
• There is a key inflection point between OS2 and OS3, and again between OS3 and OS4
• General observations
  o When to blitzscale?
   You generally don’t try to blitzscale right away. You need to apply judgment and intelligence to win a 10-year game. If you try to hit the afterburners before your business and company is ready, you’ll miss a turn and die.
  o Generalists to specialists.
   When you have a couple of founders, the founders do everything.
   Non-founders need to be flexible as well; people who can learn quickly, deal with uncertainty
   As you scale, you hire specialists with specific functions
  o Doers to Doer-Managers to Managers to Execs
   The primary focus of the Executive is the organization — how it works, how it’s structured
  o Scale and innovation
   Innovation isn’t just for startups. You have to keep innovating as you scale — both in terms of product and organization
  o The choice between adaptation vs. operational excellence
   Sometimes, you need to focus on adaptability rather than efficiency
   PayPal: We were growing between 2–5% per day. By the second week, we were 20,000 customer service emails behind. 24 hours a day, you could pick up the phone and talk with an angry customer. The objective was to solve the problem right away, rather than to find the most efficient solution. We set up a customer service center in Omaha, and hired 200 people in a month. Within three months, we had churned through 70% of those employees. Not efficient, but necessary.
  o Global reach
   When we launched LinkedIn, there were 15 companies in the “Country” dropdown. We would add countries as people complained. “I had no idea the Faroe Islands were a country!”
  o Capital requirements
   You almost always have to have access to a lot of capital to blitzscale
   You can blitzscale even in down markets; blitzscale is a relative rate
   Currently, there is a major tension between unit economics (how much do you make or lose on each transaction) and winning the market (what proportion of the market do you control)
• “There is no *one* story.” Every company passes through these phases in a different way.
• The LinkedIn Story
  o We did not blitzscale at the beginning. We spent two years in Stage 1.
  o Family: We had a single idea — if you could build a professional network with reputable relationships inside it, it would have a thousand useful uses. Everyone we hired was someone we had worked with before. Every single moment of every day was focused on what was going to be valuable to the user. We released in May 2003, and immediately began learning what product-market fit meant for us. We had a theory, but what we found was that it was different. “It was immediately obvious that recruiters were going to love this sh*t.” However, we had to build a critical mass of people for recruiters to search. 500,000 users
  o Tribe: We had to build out the operations — customer service, sales, and general & administrative functions. 2,000,000 users
  o Village: 13,000,000 users. We needed to build a great product for recruiters, but we also needed to find a fit for other professionals. We broke our organization up — from 1 R&D organization to 5. One focused on growth, one focused on revenue, three focused on new markets. That required new leadership. We brought in a CEO, Dan Nye, from Intuit. He built out our sales organization and processes. We also had to bring in product and engineering leadership.
  o City: 100,000,000 users. In 2009, we started blitzscaling. We had to bring all these people in to take advantage of the growth we had achieved to date, and now we needed to raise the bar on everything. A lot of customer growth was coming in from overseas. We brought in a new CEO, Jeff Weiner, and the very first thing he did was to prepare us to blitzscale. For example, we had never written down our corporate culture. We were too small to worry about it. We doubled in size every year from 2009 to 2014. We also had to change our technology strategy at this point — we now had hundreds of developers, and needed to focus more on productivity. We IPOd so that we could start making acquisitions.
  o Nation: 300,000,000 users. Now, we have as many employees working outside the US as inside the US. 8,500 employees, $2 billion in revenues.
• There’s no clean path that gets you from one thing to the next, but we want you to be aware of the need to change.
• We want you to see how companies change and what they need to do to succeed at those changes.
• The Family: Key Functions
  o Product
   “There are very few product geniuses who can think it, build it, release it, and it just works.”
  o Product-market fit
  o Role of founders
  o Hiring & company culture/values
   How do you get people to care?
  o Staying solvent/financed
• Other things may be important, but you don’t have time for them.
• “There are fires still burning when you go home. You have to know which ones you can leave burning, and which you can’t.” Know what’s important, and ignore the rest.
• “Letting fires burn is key, but they have to be the right fires.”
• “You (today’s students) are in a much better position to blitzscale. Many of the things you don’t want to focus on are now available off the shelf. The bad thing is, the same applies to all of your competitors.”
• Readings for Thursday
  o “Do Things That Don’t Scale”
   Airbnb found people renting rooms on Craigslist, and went door to door to convince those people to sign up for Airbnb. It also turned out that good photographs were important, so the founders paid photographers, and took photographs themselves.
  o “Startup Advice, Briefly”
  o “Why Silicon Valley Works”
• Mozilla
  o Founded 1998, minimal traction until 2004. Then in 2005, Firefox took off. Mozilla had built a fast, reliable browser right when IE was becoming unusable.
  o If we had tried to hit the accelerator earlier, it wouldn’t have worked.
  o “When you’re starting to blitzscale, it’s uncomfortable all of the time. You’re not doing anything well, just trying to keep the wings on the plane.”
• Coming on Thursday: Sam Altman
  o He see thousands of new ventures and helps hundreds
  o We’ll be asking him for patterns that work, patterns that fail
• Questions
  o Will you be covering B2B?
   Yes. The instructors have more consumer internet experience, but we’ll be covering “enterprise software” as well
  o Is blitzscaling relevant to hardware?
   “Bits” are becoming more important to “atoms.” Tesla isn’t an electric car, it’s a software care
   There are differences — you need to have a much lower error rate than for software
  o Do you need to move faster than other startups that aren’t competitors?
   You don’t focus that much on your competitors anyways; the focus is on owning the market opportunity. Competition matters if it gets in the way of that goal.
   The goal is scale; the slower you grow, the longer it will take to get to scale
   LinkedIn: We didn’t pay attention to the competition in the early days. We were focused on getting the largest collection of professionals in one place, because that is what would allow us to own the market, even when other companies were trying to monetize earlier. “Most of our competitors don’t exist anymore, and they were focused on selling to corporations.”