Teach For All Talks With Reed Hastings

Well, I was fortunate to run a company before Netflix, Pure Software, which grew to about 700 people and went public in 1995. As an engineer, every time a problem came up, I would hate that and try to put a process in place to make sure it didn't occur again. When you build an elaborate set of these processes, it's pretty impressive what can happen — you attract people who can follow the rules, you get very specialized, you don't make many mistakes, and you're very consistent. But what we didn't appreciate is that we were also very rigid. We were very specialized at doing one thing incredibly well, and then inevitably the market shifted. In that case it was from C++ to Java, but all markets shift over time. The company had very little creative, inventive DNA for figuring out the new world because we were so highly optimized at this other thing, and we ended up having to sell to our largest competitor. In hindsight, I realized it's kind of like agricultural monocropping — if you plant 100,000 hectares of the same crop, in some narrow sense you gain a yield advantage because of consistency, but you're extremely vulnerable to a single bug or disease because of this hyper-specialization. I realized there's this tension between getting better at something and staying creative, and I didn't know how to resolve that tension. In some fields, like safety-critical fields — airlines and to a degree schools — you do need incredible consistency because of the stakes involved, and eliminating error is a really good thing. Much of our economy is influenced by this manufacturing mentality where you want lots of process, great scale, reliability, and little variation. And then there's always been pockets of our economy — like advertising agencies or political campaigns — where it's the opposite, where you really want to increase variation because you're trying to learn quickly and dynamically. The fundamental model of the industrial model is to reduce variation, and in a learning model you're trying to increase variation. So in starting Netflix, we vowed to see how flexible, how few rules we could operate with. That could lead to chaos, and in many organizations it does — they start with a dozen people and everybody gets the mission, then they get to a hundred people and it gets chaotic, so they start putting in rules. Our realization was that if we have the right people and we do a really good job of setting context — constantly talking about what our values are, what the mission is — we needed very few rules. Think of it as inspiring people rather than managing people, and in the worst case, supervising people. If you're running a creative organization and you're trying to come up with new ideas, you really want to push hard on looseness, flexibility, and good judgment. You make up for it by increasing people quality as you grow. Netflix is 6,000 people now, spread around the world, with extremely few rules. Simple examples: anybody can sign contracts — for most people that's crazy. We have unlimited vacation, very flexible. Hiring and budgeting is very loose and approximate. We're so self-disciplined, and we give people incredible information internally so they can tell what's going on. We're trying to feed people rather than manage them — inspire over management. That's the big picture of this social experiment called Netflix.

Yeah, what we have to do is create room for error. Let's go back to the monoculture — the 100,000 hectares of one crop. It is short-term more efficient. If your orientation is "I've got to have the best results this year," then you probably do want to eliminate error. But if your orientation is let's tolerate lots of small errors along the way — none of them horrible, not safety-critical, but like something got spent that shouldn't have been, something got hired that shouldn't have been, something got tweeted that shouldn't have been — you're building up the leadership, the judgment muscle, so that when the big changes come, you've got people who are first-principle thinkers and are used to taking independent action. You're definitely less optimal in the short term, and we think much more optimal in the long term. Netflix started on DVD by post in the US, then expanded to streaming in the US, then streaming internationally, and then into original content. Most companies aren't able to make those kinds of adjustments because they're so highly optimized for the single business. It is a tension — if you allow catastrophic error, the organization can die. But think about it in terms of your own kids: you want your kids to have adventures but not die. A broken arm is okay, a broken back is not. We would try to have our kids have adventures — climb trees and such. It isn't optimal in the short term, but it is in terms of developing confidence and strength. You're trying to figure out where can you run a little looser and see if you only get small issues, and the benefit is developing the independent thinking muscle.

I think you just want to split some things. The safety-critical stuff — you're going to have some process around that. But then the pedagogical work and the inspiring of students — because you're certainly right that preparing students for this creative world, they have to learn and break an arm, to use that metaphor. They have to try things that don't quite work and then have that zest for going out and trying to figure out hard problems.

In the Netflix context, it's about providing entertainment that people want instead of other entertainment — we compete with sports, HBO, YouTube, lots of things. We want to do entertainment that people choose instead of those other options, and we can measure success in terms of how much viewing we win and the trend lines. That sets an overall context — we're trying to please our members, provide them joy, measured in hours of viewing. For you guys, at the highest level you're trying to transform societies by providing a change in education that results in kids being better leaders, better thinkers, more connected. Then you have to think about what are the intermediate points of how you measure that, because without any of that, it's just chaos and no one can tell if anything is better. It takes a lot of thought for you guys in terms of how lives are changing. If you focus on your ideas and bringing about change because you want to develop young minds in creative ways, and you want to do this a little bit differently in each country with unique context — there's flexibility at that level — but it's setting how you would recognize, other than smiling children, what's a great program versus merely a good one.

It's tricky, and I don't know that I would switch to a completely open model all at once. It's building a muscle around good judgment throughout the company, such that people not only are well-intentioned but have the relevant facts about what's acceptable or not. We do have a set of non-negotiables — integrity, anti-harassment, following the laws — that are not squishy, and everyone knows about those. Inevitably, a couple times a year somebody crosses that line, and then we publicize it pretty broadly within the company as a teaching moment — not trying to humiliate the person but making sure everyone understands we're all playing on a high-consequence field. We have productions all over the world, and often in some locations we'll get extorted — you have to pay a special payment to get a road open to do this movie shot. That potentially violates US and British law, so we have to be very careful. We try to develop a high understanding in our people of the rules, and then we don't paper over issues internally — we're pretty open about them. That helps the whole learning base. There are occasional violations or contracts we wish we didn't have, but we compare that to having no innovation, and so we're willing to take these small errors.

We invest a lot in it, but we haven't found much formal coursework that helps. What we tend to do is teach by example, and then we write things up — it's more like a business school where you study cases. Things happen, we write them up using Google Docs. Almost everything can be edited by anyone, so it's an evolving corpus of things you can read about what's happened and the lessons. It tends to emphasize people who are self-motivated to learn if you give them access to information. We haven't done essentially any of the formal coursework — like "Judgment 101" — or where we've tried some, they lasted in one department for two or four years but never ended up being all that helpful. You're looking at hiring people who want not to be held back, who are naturally seeking to learn and grow. Your job as a leader is to provide resources but not spoon-feed. I think of formal classes as probably too much spoon-feeding. We try to turn it around and say self-development is the fundamental skill, and we'll support your efforts, but it's fundamentally an employee responsibility — or a human responsibility — to develop themselves.

We are doing some of it. For our top 20%, we now do a version of a program that Wendy went through — one week, and that's very powerful but very general. You read Lean In and Martin Luther King and Nelson Mandela and talk about personal impacts. It's not specific skills — it's around cooperatively learning from each other. Our basic mantra is: to be a better leader, you have to be a better person. It's not around the two-minute manager or little tips and tricks to motivate people. It's about being an authentic, seeking, curious person about how to make the world a better place, staying humble in front of that challenge, and constantly stimulating that conversation. When something goes wrong, what we try to say is: okay, let's not — assuming that was a smart person who did something that really doesn't make sense — what we say is not "what did that person do wrong?" but "what context did I fail to set? How could a good person have gone off and done this?" So anytime a mistake happens, we look at it as a mistake in the context that we haven't set well. Internally, we have a document called our strategy bets — it's a Google Doc with about 50 different bets that are ambiguous about what we're doing. If something is 99% likely, we don't put it in there. What we put is all the ambiguous stuff. It makes us articulate what are all the judgment calls, and then we can collectively debate them. That's a good example of openness and being super explicit about context.

I would say we're mostly pull — there's lots of resources available. We tend to avoid every group broadcasting an email to a lot of people that has tons of information of uneven necessity, because the challenge as the organization grows is what do I need to know. We tend to put things in a Google Doc and provide summaries and links so people will go on and read more and follow the whole network of links if they're trying to understand more about an issue. On a quarterly basis, we get the top 20% together in person — that's a big investment in travel and time for two days — and we do a bunch of memos, maybe 50 pages worth, that everybody reads, summarizing their area for the last 90 days. Then it's a day of discussions on what the issues are. That's part of our meeting rhythms. But I would say it's almost all pull, and we're getting better and better about that compared to using email so much.

It's just being an incredible observer of the world. If you look at what works in literature, in cooking, politics, education — reading a lot, watching a lot, interacting, and just trying to understand why do some things work and why do they not. They're not easy to understand — they're puzzles. But through that reflection and then talking with people about why is this successful or not, I think you can develop a really strong intuitive sense of the kinds of things that make an impact on the world and the kinds of things that sounded good but kind of came and went.

Has Wendy had you read The Culture Map? That book is a great insight for people who've done a lot of international work — these different biases, like Europeans talking about the theory and then getting to specific examples, where Americans are socialized to do specific examples and then abstract to the theory. Until you see these different dimensions exposed, you don't realize how culture-specific we all are. Particularly in terms of feedback, direct feedback is hard in many cultures. What we try to do is make up for that with relationships — spending time getting to know each other, trying not to be all business, opening meetings with everyone talking about what's going on in their life a little, so there's enough relationship that people are willing to be direct. Second, particularly with our Asian employees, critiquing more senior and older people empowers — and critiquing in meetings, which we like to do for efficiency — "how could this meeting have gone better" — to our easterners that's quite shocking in a public room. I don't know that we've solved it. We try to help with relationships, and because we're so clear about what we are when people join, we get a self-selection — the Koreans that join us know what they're getting into. What we try to do is help our westerners understand that to critique someone in public in a way that's comfortable in the west would be the equivalent of yelling at them if you've grown up in an Eastern, more deferential environment. So you dial it down a little bit. What I do is spend each time in meetings talking about the feedback I'm getting, which is sometimes pretty direct and harsh. That helps set an example — Reed can talk about what he's struggling with, and that makes it more okay. The more I embody a strong desire for improvement that I'm willing to be critiqued, the better the atmosphere is. If you think about it, it's a little like athletics — if you can take the pain of exercise, you can get stronger. That's true about character and about performance too. It's girding yourself for that pain — can you do that extra mile, that extra push-up — and mapping that into the legitimate emotional difficulty of giving each other feedback.

It's an excellent question. I don't feel like we've cracked that. There's another book I'd recommend called The Advantage by Patrick Lencioni — he's got some specific exercises about creating trust and understanding team vulnerability that's quite useful. That might be one place to start. Big picture, it's just recognizing that we're all struggling with many of the same things, and if you can accept fault in each other and that we're all working to be better and make a difference, I think that helps. You certainly have to set it in a leadership example. Of course, there's a Netflix special just about this very topic — Brené Brown's Call to Courage. It's a one-hour special, significantly better than her TED Talks, and you might all start with a team watching of that.

The easy one: the top 20% is just the 20 most senior people by title — VPs and directors — which is about 900 now out of 6,000. In terms of tradeoffs — as you grow, you have to be explicit that we're going to give up some things. We're going to give up intimacy. When we were small, I knew everyone at the office. As we grow, we make more impact — that's the positive — and the negative is loss of intimacy. You walk around the buildings and you hardly know anyone. I try to be super explicit about this is what's coming: as we make a bigger impact, it's going to feel less intimate. That's a sad thing, but it's worth it to change the world. If you want intimacy without impact, you can go open a restaurant — you know everybody, your clients, it's incredibly intimate, but you're not changing the world. If you try to make the tradeoffs explicit for people, then when they actually feel that loss of intimacy that comes with growth, it's not as anxiety-provoking because they recognize, "Oh, that's right, that's the price to be able to make a big impact, and I know I want to make a big impact in the world."

Lots of people have been working on education reform for a couple hundred years. There was a famous commission in the US — the Committee of Ten with the president of Harvard and all these people — in 1892, trying to work on education. You have to think about it as: education is how society sees itself, and it's always a spectacular focus for people. Each country has a different system. Most countries in Europe, it's a state system; in the US, it's each local city and town — very fragmented. For education in the US, I think the main trouble is we have these local elected school boards. If you think of Teach For All and imagine the board of directors were publicly elected around the world — it would be very uneven who would run for it, and most of the time they win election to change the organization. You would have chaos, just as you would have chaos if the board of Netflix were publicly elected. I've come to believe that's the fundamental issue in American public education. The charter schools — which have nonprofits a lot like Teach For All — are the basic organizational structure, and if we have the organizational structure better, then the pedagogy would flow out of that. In New Orleans, one of our major cities — it was the lowest-performing city in America for 40 years, 1960 to 2000. Then the charter sector started growing there and now it's 100% charter or nonprofit public schools, and now it's solidly in the middle. It's one of the most incredible renaissance stories ever in American public education. But it's very abstract — most people want to hear about how computers are going to change education or project-based learning, as opposed to an organizational approach that unleashes all of those things. We've got a lot of political resistance. Things take a long time to change. All the problems of the world today are the problems that our parents and grandparents couldn't fix — they're really hard problems. Even simple things like democracy and one person one vote take hundreds of years to really take hold. We're on a great long-term mission, but it's not going to be an overnight fix.

It's decentralized. I talked about those top strategy bets that are throughout the company and everyone thinks about. We also have top threats — what are the threats, like Disney launching a global service to compete with us. They're about four times larger than us in revenue — 80 billion versus 20. They're incredibly good at developing creative properties, they have a lot of strengths. So we spend a lot of time thinking about that, or about what are the new ways entertainment's going to happen — maybe with augmented reality or virtual reality, or imagine a pharmaceutical product where you took a blue pill and got to peacefully hallucinate for several hours and then took a white pill and it popped you right back. That might be hard to compete with for a series or a movie. Those substitution threats are what get you. There used to be a thing in the US called the Yellow Pages — this giant phone book, a great business directory. They had a lot of competition from other so-called yellow books, but ultimately it was the internet that replaced the whole field. That would be an example of a big dislocation.

Before we got to streaming, there was a lot of anxiety because the internet's probably going to be around for 50, 100 years, but we knew DVDs were pretty temporary. We were definitely in a panic to free ourselves from dependence on DVD. We set that context: most companies who do one type of business are not going to get to the new world, and they're run by good people, so we're going to have to take more risk than most people want to, than normal human beings take. We had set the context that we have to take huge risks, and we did in terms of a number of commercial deals, most of which worked out. We had one famous miss in Qwikster with our separation of DVD, but we look at it in hindsight and say it's part and parcel of moving so quickly. I just gave a talk for our lawyers — we now have 500 lawyers in the company. I talked about needing to make more mistakes because they're moving so carefully that we really only had one bad contract with a company called Relativity in 10 years, and that cost us about $50 million. One $50 million mistake isn't enough — we should be moving quicker, faster. I don't want us to make billion-dollar mistakes, that would be quite expensive. But if we had one $50 million mistake a year — afterwards, of course, they were all talking about it, because as lawyers they want to make sure there's no mistakes. I was like, no, pick up the pace — the speed and the flexibility — until you can see you're just over the line, or an occasional toe across the line. Not ethically, but in terms of the contract — who's going to pay who if a certain situation comes about. Throughout the company, that's an example of evangelizing the benefit of even frequent small errors so that we're fast and flexible.

That's a great point. It definitely would be different. When people join Netflix temporarily — to work on a certain title because they're producing the title — we actually don't make them employees and don't try to do all this investment. They're here to make an amazing show, and then they're off to make another show, usually for someone else. That would be our closest equivalent, and we don't include them in that inner circle. That's a unique challenge that we haven't faced, and it's something you'll have to think through about how you balance that compared to someone who might spend five to ten years with you.

OK, well let's say at a high level you have a one-in-three chance of succeeding in your mission. What you want to do to maximize the odds is have about the same amount of risk from making too many catastrophic errors on one side and too little innovation or variation in approaches between countries on the other. If you think of those as the extreme ends, then you're trying to figure out if you're in the middle — you want to say, yeah, if we don't succeed one in three, it's because we were running so loose that these catastrophic things happened that killed the brand and we never recovered, or on the other hand, we were so tightly managed that we executed the play really well but it didn't turn out to be the play in many countries — we were close but we never discovered the winning solution. If your two edge-case failure things are about equally balanced, that might be a good exercise — because most people lean to the tight, because people are naturally conservative. What you're doing is incredibly hard, no one's ever done it, and so you've got to be willing to say most likely we don't succeed, and given that, we need to take some real risk to be able to invent what kind of models make a sustained impact in countries. That's okay — we're innovative, we're doing the thing that very few people are willing to do, which is to be creative and aggressive enough that we may fail, and we're willing to acknowledge that. But if we succeed, that is such a big impact in the world that it's worth it.

I would say there are several organizational moves. One is about being inclusive in its more traditional sense — race, gender, class — which is a global movement but probably strongest in the US. And then second is around being less American-centric while still keeping English as the core language of the organization. We are trying to do a lot more to facilitate people all over the world feeling they can thrive in this firm and that it feels less and less American. If you think of Nike and Adidas — for most consumers around the world, they don't know where Adidas is from, it's just got the three stripes, it's generally good. Nike is profoundly American. We really strive over time to become more like Adidas, which happens to be German. That's an effort underway — both being more effectively multinational and being inclusive are works in progress, and most of our employees would probably say we've got a long way to go on both. But it's a big effort. Well, thanks, you guys — it's a great treat to visit with you all. I'm so profoundly proud to know your organization a little bit, and the work you're doing around the world is just incredible. As Wendy may have told you, my first job out of college was being a Peace Corps math teacher in Swaziland, and so I got my taste of different types of education systems. I've never forgotten the amazing students I had in rural Swaziland. So the fact that you guys are trying to do this at scale all around the world is incredibly impressive. Thanks to all of you.