REID HOFFMAN: Hi listeners, Reid here. We’re going to start today’s episode with a visualization exercise. So sit back. Relax. And open your mind to the sound I’m about to play for you. Ready?
[DRAMATIC CHIPMUNK SOUND]
HOFFMAN: If those three notes made you picture the grainy image of a prairie dog turning suddenly to eyeball the camera, then you are probably one of the millions of people to have seen the video “Dramatic Chipmunk.”
That video was first uploaded to YouTube in 2007. YouTube was only two years old at the time, but it was already averaging over 100 million views a day.
It gave us new celebrities…
JUSTIN BEIBER: What’s up guys? This is Justin Beiber…
HOFFMAN: New catchphrases …
RYAN HIGA: How to be ninja is a DVD for you.
DAVID AFTER DENTIST: I feel funny. Why is this happening?
ANTOINE DODSON: Hide ya kids, hide ya wife.
HOFFMAN: New genres…
GAMER: Welcome to a Minecraft Lets Play video.
RHETT McLAUGHLIN AND LINK NEAL: Good Mythical Morning. Now you’ve heard of the combination restaurant? Sure, let’s talk about that.
HOFFMAN: And a new job description: YouTuber.
SEAN EVANS: Hey what’s going on everybody? For First We Feast, I’m Sean Evans and you’re watching Hot Ones.
YOUTUBER: Hey guys what’s up? And welcome back to my YouTube channel.
LILLY SINGH: What up everyone? It’s your girl Lilly.
YOUTUBER: Smash that subscribe button and we’ll see you next time. Bye everybody.
HOFFMAN: The magical thing about YouTube was that anyone could post a video of basically anything, anytime.
YOSEMITEBEAR62: It’s a double rainbow all the way. Wow, that’s so intense.
HOFFMAN: So as the platform scaled…
[NUMA NUMA MUSIC]
HARRY DAVIES-CARR: Charlie bit me. Charlie, that really hurt!
HOFFMAN: And scaled….
REBECCA BLACK: It’s Friday. It’s Friday. Gotta get down on Friday.
HOFFMAN: And scaled…
[BABY SHARK SONG]
HOFFMAN: One thing became clear:
KIMBERLY WILKINS: Ain’t nobody got time for that.
HOFFMAN: Well yes, that. But also, YouTube would need some guidance to support this rich, strange, diverse ecosystem.
The bigger you get, the harder that becomes … and the more critical it becomes, too. That’s why I believe that to navigate massive scale, you have to find – and keep – “true north.” Establish clear landmarks that point to your founding principles. They’ll guide you when you’re at risk of losing your way.
HOFFMAN: I’m Reid Hoffman, co-founder of LinkedIn, partner at Greylock, and your host. And I believe to navigate massive scale, you have to find – and keep – “true north.”
When it comes to navigation, it’s hard to find a more classic phrase than “Follow the North Star.” It sits very close to the North Pole. The southern hemisphere has its own pole star, Sigma Octantis. But “Follow Sigma Octantis” hasn’t quite caught on as a saying.
The North Star never appears to go anywhere, because it’s so near the axis around which the world turns. That makes it perfect for navigation. Cowhands would use it on cattle drives. Explorers would use it on expeditions. And if you were in the Scouts, your Troop Leader might have taught you how to spot it if you ever get lost.
No matter where you are in the northern hemisphere: if you just find the North Star, you’ll know how to get your bearings.
Finding and keeping true north is just as important when you’re navigating a company through massive scale. This kind of growth happens fast. If you and your team haven’t established founding principles as clear and bright as the North Star, it’s all too easy to become hopelessly lost.
I wanted to talk to Susan Wojcicki about this because as the CEO of YouTube, she’s led the organization through dizzying scale, and the wrenching challenges that come with it.
In 15 years, YouTube has become the world’s largest video platform. It’s also the world’s second-largest search engine, next to its parent, Google – where Susan was one of the chief architects of Google’s advertising and analytics model.
There’s maybe no one in Silicon Valley better qualified to speak on the experience of navigating a business through rapid growth – at Google and at YouTube – all the while trying to keep one eye on the horizon. Especially because she remembers when that horizon – and the entire landscape around Silicon Valley – looked very different.
SUSAN WOJCICKI: When I grew up, nobody knew where Palo Alto was. In fact, I always used to have to say “outside of San Francisco.” People knew either Stanford or the bike shop.
HOFFMAN: Susan’s dad was a physics professor, and she grew up on the Stanford campus with her sisters Janet and Anne. (You may remember Anne as the founder of 23andMe, and a previous guest on this show.) Growing up in Silicon Valley would give Susan a unique vantage point, from which she could actually watch the landscape change.
WOJCICKI: I would go to this place where you get temp work – it was called Kelly Girl at the time, it’s been renamed Kelly Services. I was actually sent to the Palo Alto Sanitation Company for three weeks where I answered the phone and did filing for them. But then I was sent to a startup. And that’s when I realized: “There’s something happening here that’s really interesting.”
HOFFMAN: That “something” was the start of the dot-com era. It would transform Silicon Valley’s trajectory – and Susan’s as well. Growing up among academics, she’d always assumed she’d become a professor.
WOJCICKI: I was doing a master’s in economics and I was about to apply for a PhD and I just couldn’t bring myself to apply. I was so close. I had already completed this master’s program, but it just wasn’t in me.
HOFFMAN: Susan had been on a lifelong path toward academia when the siren call of Silicon Valley’s growing potential demanded that she change course. Her sense of true north had shifted.
She got her MBA instead. And that brings us to 1998. Susan and her husband have just bought a house in Menlo Park – a Menlo Park now at the epicenter of the tech boom.
WOJCICKI: We didn’t know if we could afford the mortgage, so we decided we were going to rent part of our house. Then it turned out that Sergey and Larry were starting their company and they needed office space.
HOFFMAN: She’s referring, of course, to Sergey Brin and Larry Page, co-founders of Google.
WOJCICKI: It was actually really hard to find office space. Our joint friend suggested, “Why don’t you rent Susan’s and Dennis’s house?” So they showed up and they said, “Oh, this looks great.” They were living in their dorm.
HOFFMAN: So Google and its seven employees started out in Susan’s literal garage. Susan laid out the map to orient us – for an origin story from that time period that I’d never heard.
WOJCICKI: They entered through the garage and they actually had a hallway with a few bedrooms off of it. And it sounds like it’s really big, but the whole space was probably less than a thousand square feet.
I did have this small issue because I had ordered a new refrigerator for my kitchen and I was super excited about it as a newly married, new homeowner and the delivery time was something like 8:00 to 5:00. And so I took a shower at 8:00 and at 8:05 when I came down, the refrigerator had already been delivered, and Sergey and Larry had put it in their space. They didn’t know. They just thought there’s a refrigerator. This is where it went.
I had to call Sears at the time and say, “Can you reinstall my refrigerator?” And they were confused like, “Why did you install your refrigerator in the wrong place? Nobody has that issue.”
HOFFMAN: I find this story delightful. A lesson in how quickly directions can go astray if you lose vigilance in the wrong five minutes.
But refrigerator mishaps aside, Susan’s early arrangement with Google would end up pointing her toward her own true north.
Google soon outgrew Susan’s Menlo Park garage. They were focused on one thing: Make Search Better. That was their clear true north. And all of their resources went into making sure they did it better than anyone else.
But they needed guidance on how to grow their users. So they tapped their former landlord to be Employee #16, Google’s first director of marketing.
WOJCICKI: At the time, it was unclear that Google was going to be successful because there were so many other search engines, none of which really did a good job, but it was a very crowded space. Nobody thought there was actually any money in search.
HOFFMAN: Even if you know Google’s origin story – even if you lived through it – it’s truly challenging to recall what that mindset was like. Search engines at the time were choked with paid results and banner ads. Most described themselves as “portals” and crammed their home pages with headlines and widgets designed to keep you tethered to ad-bearing content as long as possible.
So when Google launched its plain white home page, it was a revelation.
And a result of following their true north. Google kept their compass trained on those three words. Not their famous slogan “Don’t be evil.” But their internal mantra: “Make search better.” This would be the key to their scale.
But as Susan said, no one at the time could see how there’d be much money in search – they didn’t have a map. And let’s be clear: at that time, neither Susan nor Google had a map either.
WOJCICKI: I didn’t really know what I was doing. In fact, I was a little scared because I thought I knew about marketing, but I’d never really been a marketer. And if I really had been a marketer, everything they would’ve said to me would’ve scared me away. In business school, you learn who’s your target audience and you build a targeted campaign and so I would say, “Who do you want me to market to?” And they said, “Everybody. We just want everybody to use Google. It can be useful to everybody.” Which is true, but how do you do that with no budget?
HOFFMAN: With no clear directions to guide her, and no map to follow, Susan used Google’s core product – its excellence in search – as her North Star.
WOJCICKI: I quickly learned, because we had no budget, that building product and using the product and the property that we had was the highest-leverage way.
HOFFMAN: What Google needed most was more users. So Susan looked for groups who really needed search.
WOJCICKI: We created this way for all universities to have Google search on their properties, and that was a program I worked on. Everybody in college at the time had internet and they had lots of searching requirements. And it was just a really quick way to spread and to get the word out.
HOFFMAN: How did you make the universities aware? Was it just simply a “make a search bar inclusible and sent out an email”? Was there a PR strategy? Was there targeted outreach?
WOJCICKI: So we created a program that was free where you just copy and paste some HTML and put it on your site, and then it created an automatic site search of your university and whatever domains you chose, as well as Google.
HOFFMAN: Yep. I remember that bit of HTML, very clever.
HOFFMAN: For those who don’t remember, this was just a simple embed code that would let universities adopt Google’s web crawler as their own. It solved a problem for universities, saving them untold hours of programming their own search tool. And it saved Susan untold hours of relationship-building.
WOJCICKI: I could call a few new universities and I would have a long conversation – or I could just put something on the website and make it free where you copy and paste. I would come in every day and I would see thousands of universities who had downloaded it and had used it. So I very quickly learned that this was a much more scalable way to be able to get the word out.
HOFFMAN: Google’s user numbers took off. By the year 2000, they had captured a quarter of the search market. But headline after headline quipped, “How Will Google Ever Make Money?”
This definitely seems quaint now. But at the time it was a serious question. They’d given out their search bar for free. They’d gotten rid of the traditional ways that search engines made money in the past. How could Google earn revenue, without ruining the guiding principles that drove their success?
Go back to what your Scoutmaster might have told you – the key to navigating uncharted territory is: if you don’t have a map, use a compass. For companies, it’s the same. When there is no map, use your compass. The one that points true north to your founding principles.
For Google, quality of search results was the magical metric they used to define success. And they followed this compass as they designed their own ad model – a transformation from industry norms.
WOJCICKI: One of Google’s biggest innovations was focusing on user quality. And saying “ads is going to have its own quality, but it’s going to be in a separate section and we’ll never mix the two” – because the early search engines had this whole paid inclusion model where you could pay to be in the search results, and Google from the very beginning rejected that.
HOFFMAN: So Google made the decision – radical at the time – not to mix paid ads with organic search results. Their next move was equally contrarian.
WOJCICKI: Google probably had about 30 people and we decided to build our own ad system. And that was just really, in retrospect, an insane decision that we have no people that understand advertising. No one has ever sold advertising. No one has ever built an advertising system, but yet we’re going to build one when there already are all these existing systems.
HOFFMAN: It may have seemed like hubris – or as Susan called it, “insane” – to decide you’re better off inventing a brand-new model than using what’s there. And in most cases, I would not recommend it.
But Google’s founders understood their own true north. Their prime directive was to improve search for individual users. And that had to be true for their revenue model too. Whatever they built could not compromise their product. The ads would have to be additive. And this was no small feat.
WOJCICKI: Our first ad models never really worked that well, and the first one completely failed where we just took Amazon affiliate links and served them and nobody clicked on them. There was a moment where everyone thought maybe we were wrong and we should just go back to banners. But we innovated a number of times on the ad model.
The vision was – and I don’t want to say this was mine – but they said we should really make the ads be targeted. So if you type in “Picasso,” we should show you ads related to Picasso; or you type in, I don’t know, Golden State Warriors, there should be Golden State Warriors merchandise. And we should do that in every single language of the world and it should be sub-second delivery time because we would never want to slow down Google for ads. And it should be super targeted and relevant.
We had been working on it for over two years that we actually came on a model that’s pretty close to what we have today with the existing AdWords.
COMPUTER VOICE: AdWords: Google’s advertising platform. Today, called Google Ads.
HOFFMAN: Google’s targeted ad model has worked so effectively, it can feel as though your browser is spying on you.
But that startling accuracy is by design. Remember: Google’s true north is the quality of search result. Did the ad turn up useful results for the user? And when they clicked, did the ad take them somewhere they wanted to go? These metrics are consolidated into a unified Quality Score. And at this calculation, Google has gotten very, very good.
Before we get to the next part of Susan’s story, I thought it would be useful to understand what “true north” actually means, scientifically. Even if what we learn is that it doesn’t really exist.
BECKY SMETHURST: North is not really a natural concept. It’s a human definition. So that means there’s no overriding sort of “true north” in the universe.
HOFFMAN: That’s Dr. Becky Smethurst, astrophysicist at the University of Oxford. And her insights on true north build toward a great metaphor for entrepreneurs…
SMETHURST: So Earth, for example, spins on its axis with north tilted at 23 degrees compared to the sun’s north. And that’s what gives us our seasons, right? We’re pointing towards the sun in summer and we’re pointing away from it in winter.
Venus though, it spins in completely the opposite direction. So its north is flipped compared to ours. Uranus spins on its side 98 degrees compared to the sun. So every single object in the solar system has its own individual unique north.
HOFFMAN: And if this is already making it feel like you don’t know which end is up, don’t worry: Dr. Becky says, physics has your back.
SMETHURST: For something that rotates, you can always define a “north.” The definition that physicists have agreed on was that if you make a fist with your right hand and let your fingers curl in the direction that the object rotates in, and then you pop your thumb out, then the direction your thumb is pointing is north. And that’s something we call the right-hand rule.
HOFFMAN: The right-hand rule is a useful way to remember that north is relative to each rotating body – each planet, each star, or even each galaxy.
SMETHURST: The Milky Way is made up of over a hundred billion stars, and all those stars are orbiting around the black hole at the very center, just like the planets orbit around the sun in our solar system. So there’s a whole system of stars. It also rotates.
HOFFMAN: So since our galaxy rotates, it has its own north. But hang on – finding it gets a little more complicated. For years, astronomers assumed that the galaxy’s north and the Earth’s north were the same.
SMETHURST: Before physicists really agreed on what the definition of “north” was, people referred to anything in the Earth’s northern hemisphere as “north.” So for example, traditionally we have called the Milky Way’s North Pole as the pole of the Milky Way that is in the Earth northern hemisphere.
HOFFMAN: But then, as our tools for observing the night sky improved, we learned which direction the Milky Way was actually turning.
SMETHURST: And what we find is that it rotates in the opposite way to what the solar system and the sun does. So if we use the right-hand rule from before, essentially, we’d sort of be pointing thumbs down. So the Milky Way’s North Pole is in the direction of the Earth’s South Pole.
HOFFMAN: So what did scientists do when faced with this case of mistaken direction? Did they go back and scrub all the textbooks, and change all the planetarium show scripts? Nope. They kept the old, incorrect definition.
SMETHURST: Technically what we call the galaxy’s North Pole is actually the galaxy’s South Pole. You’re sort of stuck with this traditional definition, that if we change now, we’ll make communication very difficult.
HOFFMAN: Even though it sounds dreadfully unscientific to keep describing the wrong north for the entire galaxy, it would have caused even more chaos to rename everything. Because after all, the only reason we bother to determine north in the first place is to establish a common language.
SMETHURST: The main point of defining what to call north is to agree on a coordinate system. So if I say that a star is visible from the northern hemisphere, you immediately know where on Earth I mean. If I say that a star is located in the galaxy’s northern hemisphere, astronomers and astrophysicists all agree on where that is.
Gone are the days of lone hero science, right? We have these huge international teams and collaborations working all on the same problem. So we need these naming and language conventions just to be able to communicate and work effectively together. So I think that’s a nice analogy for a business too.
HOFFMAN: It is a nice analogy. Defining true north for your organization, especially as it scales, gives everyone on your extended team a way to communicate – and to agree on what’s most important. This impacts not just grand corporate strategy, but the teeny day-to-day decisions made by everyone on your team.
Thank you, Becky, for extending our metaphor. Again, that was Dr. Becky Smethurst, astrophysicist at Oxford… and YouTuber. Her channel is called Dr. Becky.
And that brings us back to Susan’s story, which is about to merge with the story of YouTube.
When we last left her, Susan had helped Google scale its mission of making the Internet navigable. What is the next undiscovered country?
WOJCICKI: I started working in video with Google’s product called Google Video. And we actually started, believe it or not, just with a simple web page where we said, “Send us your video, upload your video to Google.” We didn’t even tell people what we were going to do with it. And surprisingly, people all over the world uploaded video. Why they did that? They just wanted to share it.
I remember the first video I saw, which were these puppets that were just singing in a Nordic language. It was really fascinating.
HOFFMAN: Google Video was going to fit squarely into Google’s mission of Making Search Better. At essence, better search means better access. For everyone. And as Google grew, its true north evolved – to democratizing the Internet.
WOJCICKI: Google early on had a lot of philosophy about democratizing, like: how do you make advertising accessible to everyone? How do you make publishing accessible to everyone? And so this was very aligned with making video production.
HOFFMAN: But Google wasn’t the only entrant into the race toward a video platform to democratize broadcasting. There was … another.
JAWED KARIM: All right, so here we are, in front of the elephants. Cool thing about these guys is that they have really, really long trunks, and that’s cool. And that’s pretty much all there is to say.
HOFFMAN: That is the very first video uploaded to YouTube, posted by – and starring – co-founder Jawed Karim on April 23, 2005. You may well have seen it – since then, it’s been viewed 106 million times. And if you watch this video, it says so much about what YouTube’s appeal has always been. The host is standing at the San Diego Zoo on an overcast day. He’s wearing a casual windbreaker; his head is actually obscuring part of the elephants he’s describing. The focus isn’t even on the elephants. It’s on him.
WOJCICKI: So YouTube, like … “broadcast yourself” was the original tagline. There was this appeal of the democratization of broadcasting what you had to say.
HOFFMAN: Capturing this “democratized video” market was something Google Video and YouTube were racing to do. And YouTube was winning.
WOJCICKI: YouTube launched a few months after us and were very soon bigger than us. It had to do with some of the ways their product worked and the fast turnaround in terms of upload to having that be available for all their users. And we realized that we were losing. It was both a moment of incredible highs while we found this great area that we could develop and build product and then soon afterwards realizing, “wow, as soon as we thought we had found something, we were failing.”
HOFFMAN: Failing at scale was not a familiar experience for Google back in 2005. The company had gone public the year before. And they had just reported a 700 percent increase in third-quarter profit, largely thanks to the success of AdWords.
Should they go toe-to-toe with YouTube, and try to kill them? Or was it more aligned with their true north to make them part of the mission, and buy them?
WOJCICKI: It was clear to me this was just a huge opportunity in terms of future video and I was a big advocate, along with Salar Kamangar who was also the CEO of YouTube. We got together and we had a good conversation with Sergey and Larry.
I produced a model. I did a model in like 15 minutes to show that this actually had huge potential in the future, not just in views, but in revenue too. We didn’t have a lot of time. We had to make the decision really quickly because it was for sale and we decided “yes” and the rest is really history.
HOFFMAN: As Susan just mentioned, this move happened fast. Their ability to act in the moment was thanks in part to a culture of decisiveness championed by then-CEO Eric Schmidt. For more thoughts on that, you can find his episode of Masters of Scale in our show feed.
Google was also able to move quickly because their compass was clearly pointing the way. They knew why they were buying YouTube, and where they believed it would take them. But the stakes were high.
WOJCICKI: We purchased it for $1.65 billion. And so the first direction that we got was, “Don’t screw it up.”
HOFFMAN: We’re back with Susan Wojcicki. When we left her, she had just helped broker the sale of YouTube to Google. She’d been given a single guide star: “Don’t screw it up.”
Those words would ring true a few years later, when Susan was given another sudden decision to make.
WOJCICKI: What happened was Larry just asked me. I remember he said, “What do you think of YouTube?” He didn’t say that “Oh, I’m offering you the job.” And by the way, it wasn’t called CEO at the time. I was just like SVP of YouTube.
WOJCICKI: Of YouTube, yeah.
HOFFMAN: If it’s not already clear: becoming CEO is a huge shift in responsibility, even if you’ve been working for more than a decade at the parent company. Suddenly the one whose job it is to keep true north … is you.
WOJCICKI: I hadn’t prepared anything. I didn’t know that’s what we were going to talk about that day. But I just gave him my off-the-cuff thoughts about YouTube and said I was really interested. A couple weeks later I became CEO of YouTube.
HOFFMAN: This moment – the moment of instantaneous decision – is a perfect example of why it’s important to keep true north, not just for your business but for yourself. When you’re caught off-guard, when you’ve prepared nothing, and are asked to make a gut call, you want your gut to be right.
WOJCICKI: YouTube was a lot smaller when I first joined. I think the challenges were people thought that the video was really low quality. It’s just people filming stuff in their house. Why would you want to watch that? And that was part of the magic, is that people did want to watch other people like them, and that not everything had to be produced in a studio. So a lot of traditional media people miss that, because they had to spend so much time on perfecting what you see on TV that then suddenly when you see someone filming something and they’re in their bedroom and their room is a mess and the drawer is open and stuff is falling out of it and the bed is unmade, they were like, “Why would you want to watch that?” But people do.
HOFFMAN: As it turned out, the missions of YouTube and Google naturally aligned – because both were about increasing access to the world’s information. Searchable video was a huge component of that.
WOJCICKI: We’d always use the example of how to tie a tie. That’s always going to be much better on video than in a text document.
[“How to Tie a Tie” video]
HOFFMAN: The ‘How-to’ category of YouTube has grown so big it still catches Susan off-guard.
WOJCICKI: I’m always surprised how much video we have. No matter what I want to fix, there’s a video for it. I just wanted to learn how to install the ice maker in my refrigerator. There’s a video for that. So just the depth and the long tail of videos is pretty amazing to me. But also the fact that there’s a lot of genres that we didn’t think would happen.
TV never had gaming. Gaming is a huge area for us on YouTube….
One channel, 3Blue 1Brown, has almost 3 million subscribers, and their goal is to make hard concepts really interesting and entertaining. How do you explain neural networks in a really compelling way?
HOFFMAN: The multifarious, multifaceted content on YouTube might be hard to keep track of – except that their single guiding principle is right in the name. YOU have the potential to be an expert. YOU have a platform to share that expertise with the world.
WOJCICKI: You look at something like hair and makeup and you see people of all different backgrounds, all different ethnicities, all different ages talking about hair and beauty.
And that could never have happened on traditional TV. But there’s just a multitude of perspectives here that thanks to YouTube, you can explore regardless of what your background is.
HOFFMAN: This is a key point, and central to YouTube’s true north: not everyone is an expert, but anyone can be. And if your expertise is valued, the public – and your viewer count – will let you know.
One way that YouTube put this into practice was introducing revenue-sharing with its content creators. Or as YouTubers know it: monetization.
Not every content creator on YouTube is monetized, of course. But the quest to become so is what has driven the frenzy for traffic among YouTubers. And it only functions because of an innovation Susan and her team took to video ads.
WOJCICKI: We knew we had to be different because we didn’t want to just take text ads. We wanted to have video ads and have the video ads be relevant. And that at the time seemed really unique. And the idea that you only pay if somebody actually watches those ads.
How do we optimize between revenue and our users, and how do we measure quality? If it says “free iPod” and there’s no free iPod at the end, that would not be a good experience.
HOFFMAN: Just as they did with AdWords, the YouTube team measured “quality” by relevance.
One too many “free iPod” ads, your click rate plummets. And then your audience. And then your brand partners. They needed to innovate a video ad that wouldn’t pull the viewer off-course.
WOJCICKI: We have a 6-second ad. At first, people always said, “How could you build a 6-second ad? TV has always had 15-second ads or 30-second ads. What can you do in 6 seconds? It was pretty impressive just to see that you really can tell a brand story in 6 seconds.
HOFFMAN: Just like with Google, getting ads right gave YouTube the runway to grow.
WOJCICKI: I like high-growth environments. In many ways it was like going back in time. I already know how you scale a company and how you grow it. I saw that with Google. Now, I can do it here with YouTube, and we can implement all the good things that I learned and skip the bad parts.
HOFFMAN: Implementing the good things we’ve learned, and skipping the bad parts. That’s the dream. And it would be so tempting to leave the story here. But I’m sure you know, that’s not where the story ends.
There’s a reason I wanted to talk to Susan of all people about how to keep true north as your company scales. Because few platforms have had such a public struggle with this exact topic.
WOJCICKI: You’re alluding to some of the challenges around content moderation and what we’ve termed as responsibility and that work there.
I’ll say when I accepted the job, I didn’t even think about that at all, because YouTube was really much more of an entertainment platform. It was very focused on music and gaming. They weren’t seen to have a lot of responsibility issues. They did have the “Innocence of Muslim.”
HOFFMAN: We’re not going to play that clip.
The Innocence of Muslims was a baldly anti-Islamic short film, with anti-Muslim dialogue that had been dubbed over the actors’ lines without their knowledge. It was posted to YouTube in 2012, in the midst of the Arab Spring. Susan wasn’t yet CEO. But it left an impact.
WOJCICKI: I remember sort of asking like, “Wow, like wait. Who makes all those decisions?” when I first joined. And realizing, “Wow, this could get really tough.”
HOFFMAN: The job has only gotten tougher now that the CEO chair is hers. Our politics have only grown more polarized, and disinformation more virulent. Meanwhile, YouTube has only gotten bigger and harder to wrangle. Right now, YouTube sees two billion active users per month. That represents about a third of users across the entire Internet.
Over 50 million of those users are contributors to the platform – not all of them with intentions as innocent as showing how to tie a tie.
WOJCICKI: These are incredibly hard issues, and they’re not just technical issues. It’s a new era of how are platforms going to be managed to both be responsible but enable free speech? It’s been trying to strike a balance of making sure that whenever we take a step in responsibility that it’s really well thought out, that we have spoken with the experts. Because you can always have these unintended consequences of removing content that you didn’t mean to.
So having it very cleanly defined so that thousands of people around the world who are reviewers can all make the same decisions consistently. And that users and our uploaders have an understanding of what those policies are. Realizing that some people will say, “You didn’t go far enough,” and others will say, “You went too far.” So it is a very tough job in the sense that no one is ever happy about it. But you do your best to do what you think is the right long-term benefit for your community.
HOFFMAN: What is best for the community? A little while ago, I had Shishir Mehrotra on the show and we touched on this. In his six years at YouTube, Shishir was responsible for much of their scaling success. He also led the team that shifted YouTube’s metrics focus from number of videos clicked to hours of YouTube watched.
SHISHIR MEHROTRA: So we set this goal, “We’re going to get to a billion hours a day.” And had this very big positive rallying factor. And everybody feels like that’s a big goal. And if you ask anybody at Google in that period and said, “Hey, what are the YouTube guys working on?” They’d probably say, “Oh, they’re working on this crazy billion-hour thing.” And it was good: It was well branded, it was understandable, it was specific. You could make decisions based on it, and so on. But it had an obvious flaw. Is it actually a good thing for people to watch a billion hours a day of YouTube?
HOFFMAN: Notice that the North Star guiding Shishir and his team hadn’t changed. They were fulfilling Google and YouTube’s constant promise of quality search results. Focusing on hours watched instead of videos clicked meant devaluing clickbait, in favor of videos that match the viewer’s interests. The more hours a viewer watches, the better YouTube is at it’s #1 job: Make Video Search Better. Right?
Well, YouTube hit their billion-hour goal. But this led to the mother of all unintended consequences. A rise in hate speech, conspiracy theories, and violence. YouTube’s recommendation algorithm even came under fire in Congress for sending viewers down rabbit holes of more and more fringe voices.
So did that mean YouTube lost their way? Or did their own guiding principles need to change? I asked Susan about it.
HOFFMAN: What were some of the things you had, “Okay. Here’s some of the early lessons about how to maintain a true north within the melee of a lot of different people with a lot of different opinions?
WOJCICKI: I’d say on true north, certainly one of them has just been about trying to, on one hand, enable freedom of speech and as many voices as possible whether we agree with them or not, but really having the boundary be around human rights. So if you have different cultures who may have different perspectives on gender, sexual orientation, we want to make sure that no one is discriminated against. So regardless of what the culture is that we’re not doing anything that would in the end lead to harm.
HOFFMAN: Pursuing this goal meant a programmatic shift in how YouTube defines “quality.” Previously, “quality” meant search results and recommendations that were the most accurate to search queries. Now, YouTube was shifting toward a new definition: away from pure accuracy, and toward responsibility.
They addressed this change in their community standards as well. In December of last year, YouTube rolled out an updated Terms of Service that caused a stir in the YouTuber community. It stated, in no uncertain terms, that YouTube is under no obligation to host or serve content. This change was made to expand YouTube’s ability to take down offensive or abusive videos.
WOJCICKI: We’ve had to make a number of changes in terms of how we handle monetization because brands required it. Brands said, “We don’t want to be on content that has profanity or violence or certain types of topics,” right? And I can assure you that pretty much like every kid in the world knows about demonetization and the policies because all creators talk about it.
HOFFMAN: That doesn’t mean that the changes have solved everything.
WOJCICKI: What we struggle probably the most with is content that we’d say is really borderline to our policies, but is not necessarily violative. And that might be content like, “Aliens landed in my backyard and they told me something about COVID.” So that is not necessarily a violation of our policy, but it’s not necessarily the content that is useful for our users.
HOFFMAN: It’s a massive undertaking to support endless user-created content in a way that stays diverse while filtering out violent extremists, conspiracy theorists, or those who think COVID came from aliens or 5G.
But navigating through this process has allowed Susan to reckon with the tradeoffs and challenges in navigating YouTube’s true north.
WOJCICKI: Our goal is to be the best home for creators, and so we want to build the best services and make sure that they can achieve their goals of both revenue as well as traffic. But at the same time know that they’re watching you carefully, because every single thing you do has some kind of impact for them. And they’re going to let you know really quickly if you made a mistake.
HOFFMAN: Keeping true north isn’t one single decision, one set of navigations. It’s many, many micro-decisions along the way. The direction may be clear, but you’ll still need to course-correct often, especially when you make a wrong turn.
And to be clear: YouTube hasn’t figured it all out yet. Growing a great, open place for creators that also keeps corrosive content at bay will remain a constant challenge. And it may actually get more complex as time goes on.
But here’s the secret: it’s not supposed to be easy. What separates a true north from just a direction you happen to be going is this: Will you follow it even if it hurts? Even if it means you fail? If you’d rather go under than violate your first principles, you know they’re guiding you true.
I’m Reid Hoffman. Thanks for listening.