The media mogul (twice over) on being both contagious and sticky
By Felix Salmon
Illustrations by Young Professionals
I first got to know Jonah Peretti in February 2011, very shortly after the Huffington Post was sold to AOL. At the time, he explained to me that BuzzFeed, a company he had built to seed viral content, had become astonishingly popular in its own right, which didn’t surprise me in the slightest, because virtually everything that Jonah Peretti touches seems to become contagiously popular.
Jonah had made a healthy amount of money from the Huffington Post sale — he was a co-founder, after all. But he wasn’t yet a media mogul in his own right. Now, in 2014, he is.
BuzzFeed, Jonah’s erstwhile viral experiment, has become a white-hot property which is almost certainly worth more than the $315 million AOL paid for the Huffington Post. (While the Huffington Post had just $30 million of revenue when it sold, BuzzFeed is expecting to bring in four times that amount in 2014.)
Jonah is not your typical media mogul, however. He’s smarter than most, and more accessible, and also much happier than many to share his thoughts. Which is why I asked Jonah if he’d be interested in talking to me over an extended period. To my delight, he said yes, and we ended up having four interviews spanning more than six hours.
The resulting Q&A is long, for which I make no apologies. You’ll learn a lot about Jonah Peretti and how he thinks — but you’ll also learn a great deal about the modern media world, the way the Internet has evolved, and the way that Jonah has evolved with it.
If you want to learn the secret of how Jonah managed to build two of the world’s most important online media properties, you’ll find that here, too. Which brings me to another way in which Jonah differs from most other moguls. If you succeed in building something similarly successful as a result, he will be cheering you all the way.
I ❤ Print Media
“I think The New York Times does a lot of tremendous journalism. It’s just sometimes, the stories are boring.” Jump
The Story of Jonah
“What are we doing here?”
“There’s sometimes moments where networks are so amenable to spread.” Jump
Too Soon in Tech
“The thing is, if you care about having an impact on the world, the too-early mode is the highest leverage point.” Jump
“What Ken Lerer said to me was,
‘I know business, you know the Internet, let’s build something together. Let’s start something.’” Jump
BuzzFeed as Willy Wonka’s Lab
“It wasn’t easy raising money for BuzzFeed. It was always, is there any way you can do this without having any writers or content creators or journalists?” Jump
How to Win the
“What would this be if the readers and the publishers were not focused on making something similar to print?” Jump
“Life is tricky because it happens once and there’s no opportunity for A/B testing.” Jump
I ❤ Print Media
“I think that The New York Times does a lot of tremendous journalism. It’s just sometimes, the stories are boring.”
Back to Top
Felix: Let’s talk about the only subject of conversation among media types for the past couple of weeks.
Jonah: NBA playoffs.
FS: You guys had a little scoop there with The New York Times Innovation Report.
JP: Yeah, we did.
FS: Did you read it? I thought it was pretty good, pretty smart, intelligent stuff.
JP: I did read it. There were a lot of interesting things in it. I think in some places, they were a little bit overly critical of their tech and product team. When you look around the industry, The New York Times has a really great website. They’re building lots of things themselves and integrating them. It doesn’t feel like a Frankenstein website with things bolted on from millions of other places. I was a little surprised at the tone, how critical they were of their web products.
FS: I think so too.
JP: And they were a little too nice to some of the competitors they mentioned. They were like, this site has this feature, and this site has that feature. But, a lot of the sites they were mentioning, the Times is actually ahead of on tech and products. And you can find one thing that another site has that you don’t have, but you shouldn’t lacerate yourself for that.
Then the other thing was this bedrock assumption that The New York Times’s content is just the best content. I think that they do a lot of tremendous journalism. They have a lot of great stuff. It’s just sometimes, the stories are boring, or, by certain metrics of excellence, they’re not excellent content. The report was a little too harsh on their tech and product team, and not critical enough on their editorial team.
I was a little bit surprised that the report didn’t spend much time tackling the hardest issue, which is why do they need to have so much revenue? It’s because their cost structure is made for print. When you look at how much revenue comes from print and the scale of their operation because of print, the challenge that they’re facing moving forward is how do they move into a post-print world.
FS: A post-print world, by its nature, is a lower revenue world than a print world.
JP: It might not be eventually, but it is right now, and how do you switch over? That seems to be the actual tricky issue.
It just seems like if you’re reading a secret internal report for The New York Times, the things that people would be stressed about, isn’t that, oh, the website’s not good enough, or they haven’t moved fast enough with this feature or that feature, but more like how do we deal with this very different cost structure of our future business, compared to our past business.
FS: There’s real value to print, still. And covers. There’s an amazing value to the New York Post front page. There’s amazing value to the front cover of The Economist. There’s an amazing value to the front cover of Vogue. And it seems to be that all of that value has either disappeared or is disappearing in the digital world, or else it’s being captured by the platforms rather than the publications.
JP: There’s something about covers. For a while we were making covers. If a story was big or important, I’d make a cover. Maybe the fact that you only have one cover each week means that you’re taking the force of an entire brand and making one big bet. Every week. Or every month.
FS: If you had something you wanted to change in the world, how much marginal extra impact could you have by that act of pushing it? Do you have the equivalent of a front page or cover on BuzzFeed?
JP: With gay rights and Sochi, we pushed really hard along with a few other groups, and it resulted in statements by Russian authorities and Obama.
Six months before the Olympic games we were pushing on this. It was a combination of what we call Buzz Team stuff, you know, the lists and quizzes and entertaining stuff, and it was reporting and scoops and more classically journalistic work. The combination of those things did draw a lot of attention early on. But it’s not like one cover and one story.
FS: It shouldn’t be. This is the great advantage of digital. Things don’t end once they’ve gone to print.
JP: We see with our longform stories that, in some cases, the sheer length and rigor of a piece will make the piece have a bigger impact. Just the fact that it’s 6,000 words or 12,000 words.
FS: Can I just say, for the record, that your piece about gays in figure skating was way too long?
JP: Was that one too long? Well, they shouldn’t be long for the sake of being long. I didn’t read that one.
FS: No one read that one. They got about a third of the way through and then they’re like, “Oh my god, there’s like 8,000 more words. I just can’t bear it.”
The Story of Jonah
“What are we doing here?”
Back to Top
Jonah: What are we doing here?
Felix: What are we doing here? That’s a good question. What we’re doing here is something that I don’t think I’ve ever done before, which is a longform Q&A.
JP: This will be longer than that figure skating article, and even more boring.
FS: But you know what? It’s got a rhythm to it, which makes it so much easier to read a long Q&A, if it’s well edited.
JP: It’s like a list.
FS: It’s a little bit like a list. It’s a list of questions. I guess you can skip. But you never do, though. When you’re reading those things, you don’t skip questions, do you? It just makes me feel like I’m making progress.
So, let’s get to it. You have been, over the course of your career, pretty astonishingly successful at capturing these moods of the Internet, for lack of a better word, and knowing what’s going to work, or seemingly knowing what’s going to work, or at least being able to find out what’s going to work and doing it. And you’ve done that in a bunch of different areas, in a bunch of different companies. I guess that one of the things I want to do is go back to the beginning and some of the earlier stuff you did, and say, back when you forwarded an email from Nike to a few friends of yours or whatever, was that all part of some grand career plan you had going on?
JP: Every step is just perfectly planned out.
FS: It’s so easy to create a narrative with hindsight, to put this beautiful narrative career progression onwards and upwards.
JP: One theory about early successes is that they often are fairly random. So there’s lots of smart people, there’s lots of pretty talented people and then, in some cases, they end up being a talented computer science grad at Stanford. Some of them go on to be professors and teach or work at good startups—and then Larry and Sergey are running this giant empire. Part of that is chance, or luck, or happenstance, the initial project being right for the moment and taking off and getting attention and it snowballs from there. There’s some theory that a lot of this is the result of preferential attachment or cumulative advantage. Preferential attachment is a node in a network that you connect to early and then it’s more connected so it’s more likely to attract other connections.
The Nike email was this accidental, random thing that resulted in me being able to find Ken and Arianna Huffington to be partners with because I had made this big thing that was big on the Internet. And then because of that it was easier to do BuzzFeed. You have more opportunity when you have an early success, then there’s more opportunity and then that opportunity allows you to do other things and it snowballs. So small, lucky things that happen early on can sometimes have a ripple effect and then people look back on it and say, oh wow, that person is smart or talented or something because it seems like it’s a string of these things, but actually they’re not independent variables.
FS: So where would you begin in the story of Jonah Peretti.
JP: I probably wouldn’t.
FS: This is where you have to talk about yourself. This is the uncomfortable bit.
JP: The Nike email was the thing that opened my eyes to the kind of stuff I’m doing now and it really was an accident. It was, “I should be writing my master’s thesis. I’m procrastinating.”
FS: You’re where at this point?
JP: I’m at the MIT Media Lab and, before that, I was a schoolteacher in New Orleans. I could start back there if you want. I was just in New Orleans this weekend. I saw the guy who gave me my first job, Dale Smith, who was the head of the computer department at Isidore Newman School in New Orleans. He hired me when I was twenty-two years old and I had some teaching experience in summers and some technology experience. I moved to New Orleans and became a schoolteacher. I was twenty-two years old, wearing a tie, being called Mr. Peretti by your students.
FS: And you’re teaching them computers?
JP: I taught sixth graders how to make websites on GeoCities. I taught sixth graders how to make games like Myst, using Hyperstudio. We’d make a giant map and then everyone in the class would make different rooms and they’d all connect together. I taught Logo programming, things like “design a flower that is different every single time because there’s randomness in it, but the petals have to connect.” Teaching things like order and randomness. We did social protest letters, using the Microsoft Office Letter Wizard.
These sixth graders would send a letter to Mayor Morial or send a letter to The Limited complaining that the prices were too high and their parents wouldn’t let them buy outfits there. Dale, or Mr. Smith, hired me and then let me just come up with creative lessons and experiment with ideas. Just so long as it involved technology, I could do it. Before there was the Second Life craze—which has now passed—ten years before that there were these virtual worlds, and we used those to do historical role playing between kids in the South and kids in the North.
I did all these projects with students and then I would write them up and submit them to educational technology conferences. I ended up going to Calgary or Seattle or Orlando and presenting the work that I was doing with these kids. That was really only possible because Dale was supportive, and the school even paid for me to go on a couple of these trips. It was really fun and actually Peggy Wang, the first editor of BuzzFeed, was my student at Newman and is still here.
That is how I was able to get into MIT. I had published these papers and had done all these interesting lessons and there was a professor there who was interested in cognitive science and behavioral psychology and learning, so I joined that group to do educational technology work. And then I got to take classes from Marvin Minsky, the artificial intelligence pioneer, and Seymour Papert, the guy who invented the Logo programming language, and a bunch of people at MIT who were thinking about learning and education.
That stuff is relevant to the media business. You’re explaining things and communicating.
FS: And being creative.
JP: And you have lots of practice communicating.
I was trying to teach twelfth graders. In the school handbook, there were three grading scales, one was a hundred-point-scale, and one was a four-point scale and one was a letter grade scale. I noticed that they couldn’t just be converted between each other. Like an F and an A would average to a C, but a hundred and a zero would average to an F, because it would be fifty. So I noticed that there were these weird things and I was like, “Oh, that seems like an interesting thing to teach these students, especially because they think about their grades and they want to get good grades and the way that grades are calculated could have an effect on them.”
In one system, if you have a lot of A’s you can be like, “I’m not even going to take this exam,” and then in another one, you’re like “Oh, I’ve got to go take the exam and get an F.” We had Excel and so I was trying to use Excel to model some of it. It was so hard. The first time I taught it in a class of twenty, maybe five students really got the nuances and could use it in different ways, could understand the idea of these different systems. The next time I taught it, maybe I got half the class to get it. And then I kind of got stuck there.
I got maybe 60 percent of the class to understand this. The default is always, and I saw this with all the good teachers, that it’s your fault, not the students’ fault. You should be able to figure out how to teach students something, and if there’s a student who has a learning disability, you should figure out how to route around it and explain it in another way. The worst teachers, in my view, were the teachers who said, “Well, that kid’s just lazy” or “This kid’s just stupid,” and they would only teach for the 80 percent of students who thought in a normal way.
When you think about the media industry, it’s also, “How do you reach people and how do you get people to understand?” If you write something and nobody understands it, it’s easy to be, like, “Oh those are all the dumb people.” Sometimes writing something that’s very sophisticated and difficult and technical for a particular audience is totally fine, but you should be able to communicate in simple language.
FS: Just about anything, right?
JP: Yeah. The thing is, there are dangers in this, because you can also explain something in a way that makes people feel like they understand it when they actually don’t.
FS: Fox News is very good at that.
JP: Right. You can figure out a way to frame something and explain it so that it feels like it confirms what people already believe, including incorrect things they believe. You learn a lot from teaching and you learn all about how to communicate. I came out of a university system that was, at that particular moment in the ’90s, glorifying postmodern critical theory. There was a sense that the best way to show you understand something is to write something incomprehensible. So all of a sudden, sitting in front of a bunch of twelve-year-olds or a bunch of seventeen-year-olds and having to try to explain something to them, it was a good reminder that it’s possible to communicate.
FS: Was that you? Were you sitting there, an undergrad, reading French philosophers?
JP: I was reading Foucault and Roland Barthes and Kant and Marx and Freud.
FS: And also doing the tech stuff.
FS: This was where and when?
JP: This was UC Santa Cruz, 1996. I grew up in Oakland, California, on the border with Berkeley. My mom taught at Berkeley, and it was like right there.
So I only applied to UC Santa Cruz. My college guidance counselor told me that they don’t even read the essay if your scores are at a certain level. So I was just like, “Okay, that means I can just apply to one school,” and they were like, “Yeah you can.” Some friends were like, “you’re throwing your education away.” There were no grades at UC Santa Cruz. We’re really going deep. If I didn’t know you, I wouldn’t talk to you about any of this stuff.
FS: This is what [writer] Jeff Bercovici said about [Gawker Media founder] Nick Denton, as well [during a recent Playboy interview]: “I could do this only because we’ve known each other for years.”
JP: There’s something about talking about yourself to someone you know, as opposed to… Anyway, I just made every class I was in an independent study. I’d go to the professor and be like, “Hey, can I do this extra thing, or do that?” As a freshman, I’d go to the upper division classes and say, “Hey, can I take this class?” And by sophomore year, I would go take graduate classes. A lot of the best professors were in this History of Consciousness program, which was essentially each professor competing to be the most radical, whether it was French feminism, postmodernism, critical theory, post-colonial studies, Lacanian psychoanalysis.
Then, each summer, I worked as a teacher: sometimes as a technology teacher and sometimes just at educational summer programs.
“There’s sometimes moments where networks are so amenable to spread.”
Back to Top
Felix: From here, you then get this awesome-sounding job in New Orleans. You had that for how long?
Jonah: For three years. Then off to MIT where I just wander around with a bunch of nerds and eat tacos and talk about ideas and work on projects and have no real responsibility at all.
The Nike thing happened in the end of my time there and was accidental and got me thinking about networks and scale and network science and how do ideas spread and things like that. But most of my time when I was there, I was thinking about teaching and learning and technology. I built a prototype of a Lego robotics thing that, while you’re watching educational television, it beams code to the thing you’re building so then it animates it. These are the kind of things you do there. You’d hack something together in a couple of weeks and you’d get to the point of it being a demo and some of those times you’d just stop there and then do something else.
The laboratory for computer science thought that maybe our people were a little bit flaky.
JP: Yeah. They’re working on one algorithm or whatever for like a year, and we would just kind of do projects and demo them. The demo was a huge part of the Media Lab. Sometimes corporate sponsors or just your peers would come by during demo week and you would show them stuff. If you had a cool demo, people would gather round and get interested and want to hear more and want to contribute to it. If you had a lame demo, people would shun you.
FS: And then, while you’re doing your master’s, you just decide to troll Nike by email.
JP: I wasn’t trying to troll them, I was just trying to check out their service. I was kind of like, “Oh, that’s cool.” The idea of product personalization is the kind of idea that could be a media demo. I wanted to see how it worked.
It was a little disappointing because you couldn’t really do that much. It was more just change the color or put a name on. And then I think I tried a bad word or something and it was banned. Then I put sweatshop and it wasn’t. Oh, interesting. So I ordered them. I was like, “They’re going to send me these shoes.” I just thought that was a funny idea. But then that led to this person writing back and I was surprised. I thought, “Oh, I’ll get a form letter back.” And it was clear that what they wrote back was part formula, part human. Like a human was editing a form letter.
I was like, “Oh, there’s someone on the other end of the line.” So I said, “No, sweatshop is in the dictionary. It’s a shop where factory workers toil under unhealthy conditions.” And then they wrote back again. It was in January of 2001, so there wasn’t YouTube, there wasn’t Facebook, there wasn’t Twitter. There wasn’t a concept of things going viral. It’s funny how that’s become so normalized where people say, “Oh, maybe that will go viral.” At that time, nobody ever said, “This thing could go viral.” That wasn’t part of anyone’s consciousness. There were email forwards, and blogging was an interesting subculture.
So it became one of the early email forwards, which you already know, and resulted in reporters calling me to ask about it. I ended up on the Today Show with Katie Couric and Nike’s head of PR, debating something I didn’t even know anything about, thinking, “Why am I here instead of people who’ve dedicated their lives fighting for human rights?” So to me it seemed like this interesting communications phenomenon.
The Media Lab had Lexis Nexis, so I found all these examples of press articles. In all this press, people were saying, “and he just told a few friends and then it went viral.” I found that there were just dozens and dozens and dozens of press articles. You could do a find and replace of the Nike emblem and put some other thing that had been written and it was always like, “This kid did this little thing and it was just kind of a joke and they were just trying to make a dancing baby for this one thing and then, all of a sudden…” There were dozens and dozens of these.
One school was trying to see if they could have an email that would reach all the way across the world to Australia. They sent out an email and said sign this thing and forward it on and try to reach someone in Australia. Their servers were taken down because so many inbound emails came and it crashed. So there are all these examples. From this totally naive perspective, they all said this accidental thing happened, and a cascade through a network emerged, although they didn’t use that language. And then, like, “Wow, isn’t this an interesting, weird phenomenon” without really any insight beyond that.
Then it turned out there were also all these people studying network science, that I didn’t know about at the time, including this guy Duncan Watts, who’s now BuzzFeed’s science advisor and a good friend of mine. He was working out the mathematics of six degrees of separation as a grad student at Cornell and studying the networks of how a cricket chirp spread. Should I stop?
FS: No, keep on going.
JP: A cricket would chirp, and then neighboring crickets would chirp, and that sound would cascade. So he was measuring that empirically. [Psychologist] Stanley Milgram, in the 1970s, had coined the phrase six degrees of separation. He had figured out experimentally that you could start with a chain letter in Chicago and reach a stockbroker in Boston and it took on average six steps, but he did it without any mathematical rigor and with all these broken chains, where it would go three hops and go dark. So Duncan essentially looked at that and said, “Well, what if we re-did this experiment? What if we mapped this out?” With his advisor, he created the Watts-Strogatz theorem, which described the conditions that a network needs to have for it to be a small world, where through a few short hops, you can get to any other node in the network.
Duncan was studying networks and then he starts looking at the study of human behavior and decisions and behavioral economics and things like that. So I think Duncan went into it at a much deeper level and with much more rigor and much less smoke and mirrors and with many fewer people watching. I think that there’s this natural progression of these two pieces. Usually the order is, “Oh, wow, look at this parlor trick. Something just spread to millions of people and everything is connected and things cascade.” And then you say, “Oh, how does this work?” Well, there are mathematical properties in networks and that’s interesting to look at. But then there’s each node. In the case of an email forward, the nodes are people or a site, like Black People Love Us or Rejection Line.
And then how do people make decisions? How do they decide that of all the things that come into their inbox, they’re going to forward this Nike email, but they’re not going to forward other things? Then you start thinking about human psychology and how people evaluate content. How do people decide whether something is worth passing on to their friend or not? This kind of gets back to education in a way, or trying to understand how the brain or mind works.
It was this giant, unfolding, complex problem with different areas that you could spend a whole lifetime thinking about and working on from the structural mathematical sense of networks. All those things working together started to feel like the way the media industry’s going to work. That’s the way content and ideas will spread. It doesn’t necessarily have to be a broadcast pipe or printing press and trucks driving around newspapers. It feels like there’s this new, different way that people are just beginning to understand.
Even now, websites look kind of like print publications and we’re still stuck in lots of legacy ways of thinking. But now there’s this possibility of, “Oh, it could work a different way.”
FS: And all of this just went automatically through your head the minute you saw the Nike email go viral?
JP: No. It’s like a constellation of connected things.
FS: Okay, so this is one of the very small number of actual, genuine questions I wanted to ask. Obviously the Nike thing was lightning in a bottle—just randomly happened. And then you have Black People Love Us and Rejection Line.
JP: Rejection Line and then Black People Love Us.
FS: Were there others in there which just didn’t work, or did you have a hundred-percent hit rate?
JP: The Nike email, and the next was Rejection Line. That was the only other thing. Then Black People Love Us. I was teaching a class, and it was a design class, so Chelsea [Peretti; Jonah’s sister, currently starring on “Brooklyn Nine-Nine”] and I farmed out ideas to the students and said, “Your job is a design challenge.” We were the clients, basically: “Make these things.” Some of those ideas were things that we came up with because we needed one for every student, and some of the things we were excited about. Black People Love Us is the one where we actually took the pictures, developed the stuff, you know. But there were others in that cohort of students that didn’t do as well.
FS: You’re basically just completely taking advantage of these poor design students to further your own nefarious ends?
JP: Actually, I stopped teaching the class that way. I subsequently taught the class and said, “You have to come up with your own ideas.” I did an RFP. Part of the issue was that the students at Parsons [The New School for Design], some of them were pure design people, so they were like, “What? Come up with an idea for some viral thing or whatever? I know Photoshop. I know Illustrator.” Maybe that’s changed since then.
With this class, we had them do a design challenge of our ideas and then make their own projects. So we said, “Learn by doing one of our projects and then do your own.”
FS: But your ideas were better than their ideas.
JP: Yeah. There’s all these things you can do to ensure that something doesn’t work. I’m thinking back to what other things were going on at that time. The first one was sort of an accident. The Nike thing was an accident and I collaborated on that one with a customer service representative at Nike—a non-consensual collaboration maybe. The next one with Chelsea was Rejection Line and that was the only one we did and it really took off. And then Black People Love Us was the one we really focused on in terms of these design challenges.
I think it’s hard to do that and the only reason we were able to do three in a row is because no one was trying to make things that were viral. There was no competition. There were things accidentally happening that sometimes would go viral, and then we were like one of maybe a few dozen people trying to actually make viral web culture, when nobody was doing it. So the networks were completely open in the sense that no one was even trying to make content that intentionally would go viral. A few years later, everyone is like, “How do I make my YouTube video go viral and how do I make this and how do I do this viral marketing project?”
There’s sometimes moments where networks are so amenable to spread. Duncan uses this forest-fire analogy, which is if there’s a forest where the underbrush is wet, the trees are far apart, there’s not many dead trees, you could take a flamethrower to it and it won’t burn. If the forest is dry and it’s been hot and the trees are close together, you can just drop a match and the whole thing will burn. I think there was a period between 2001 and 2003 when the dry forest was ready to burn. If you made something that was pretty funny and you made something that had certain qualities that caused people to want to share and talk and discuss, then things would spread pretty far. Now you see people do a really cool project or a cool Tumblr and they don’t end up on the Today Show. We were on Good Morning America for Black People Love Us. We had the front page of Sunday Styles for Black People Love Us. The Rejection Line, we were on CNN and in People and in Elle. I think that some project like that today, would not have had the novelty to get the mainstream attention and would have a lot more competition on the web of cool things, and the rate at which they spread has been compressed a lot so things pop for a day or two. The Nike email was three or four months and the Rejection Line was like maybe a month and a half and then Black People Love Us compressed to a little less than a month.
So things were spreading faster and the network was more lubricated or more connected, but also there started to be more and more people making things designed for these networks. So it’s kind of like cable existed and there was only three channels on cable and you start a new cable channel and you’re like, “Oh, people love my channel.” But it’s really that the cable networks were just built out and Viacom doesn’t exist yet so you succeed by default.
FS: It’s interesting that the sign that you got it right, the sign that these things succeeded, is, you can say, “Oh, I was on the telly. I was on the front page of the Styles section.” It’s this sort of ratification of digital culture by the mainstream media, which was very important back then.
JP: Yeah, for sure.
FS: But it still seems that these were basically, for lack of a better word, Media Lab projects? They were still just kind of you messing around on the Internet and seeing what worked or was there actually some kind of commercial kernel?
JP: It wasn’t commercial. In fact, often the mainstream media would cover this stuff — we were on CNN talking about Rejection Line and they think it’s kind of a cool project and then they’re like, “So, are you making any money on this?” We were, “Oh, no, it’s not a commercial project.” And they’re like, “Ha ha ha.” They were laughing. “You’ve got a lot of time on your hands.” That’s a phrase people would use. “Oh, wow, you must have a lot of time on your hands,” and laughing that it doesn’t make any money.
For a lot of people, having something make money is the thing that legitimizes it. Maybe mainstream media legitimizes it to a certain point, but mainstream media plus actually making money is, “Oh, this is real.” For me, I was just trying to understand how these networks work. And for Chelsea, she was just like, “Oh, this is awesome. I have a massive venue for my comedy.” For her, I think she saw it as more of a platform for comedy and for me it was like a lab to understand the way networks work and content spreads and ideas spread.
FS: When did you first hear this word viral? This odd epidemiological metaphor that seems to have taken over the world?
JP: I’m not sure when I first heard it, but it’s one of those words that started to have a worse and worse brand, if you can think of a word having a brand. Viral marketing companies started to use it to describe a phenomenon that is actually not viral, which is making something and buying a million views for it and then saying, “This got a million views, it’s viral.” And then they’d say a million views means viral. We were thinking in terms of an actual epidemiological definition of viral, with a certain threshold of contagion that results in it growing through time. Instead of exponential decay, you get exponential growth. That is what viral is.
FS: I didn’t know how to react to Black People Love Us. I think I got the joke, but I’m not American, so maybe I didn’t. I certainly got the message, which is, “Let’s laugh at all of these people who are getting offended by this.”
JP: Well, right. As trolling, you understood it.
JP: So, some context. My sister lived at Sojourner Truth House at Barnard College and was one of the few white students who lived there. If you think back to university, like UC Santa Cruz, which was a very progressive university but with mostly white students, there was a phenomenon of liberal, progressive students who were trying to show how not racist they were. I was just thinking about this. I had a friend who had dreadlocks and anywhere she went, white guys would offer her weed. She smoked weed occasionally so it was actually kind of nice. She never had to pay for weed. But it was like, “It’s so cool. I’m smoking a joint with this cool black girl.”
There was this excessive attempt to show that you were progressive and you had black friends, so I think part of the humor of it comes from this sort of late ’90s or early 2000s progressive university culture that was starting to evolve beyond this naive and performative attempt to prove that you’re not racist through your affiliations with black people. That was at the center of it. The targets of the satire were a particular kind of young, white, liberal person. The reality of these web projects is that they then reach all these other groups.
We were continually shocked at how different people were reacting to it. One thing Paul Berry noticed when he was doing his Dog Island project is that there’s always this matrix where you have the four quadrants. One dimension is, “Do people get it or do they not get it?” and then another dimension is, “Do they like it or do they hate it?”
FS: You need a certain amount of don’t-get in order for these things to work?
JP: Yeah, probably.
FS: Were there any white liberals in the not-get half or did they all get it and either hate it or like it?
JP: That’s the other thing. You could put every email we got into this matrix, too, and then we’re the ones who don’t get it or get it. We’d get emails from someone saying, “Oh, yeah, I love the site and I have a lot of black friends.” Are they playing along, or are they liberal and they’ll get it, or are they just clueless?
FS: This reminds me a bit of, I guess from the same era, was what you might call early snark era Dave Eggers before he renounced such childish stuff and became a post-ironic do-gooder.
JP: What if that whole debate is really just about life stage? It’s like, “Oh, no no no, snark and trolling and being relentlessly critical is good when you’re younger…”
FS: But this is super-high irony, right? That’s something which feels like early 2000s in a way. You don’t have that mode so much any more.
JP: It’s harder to pull off too because there are so many ways of validating things. People are so good at saying, “Oh, let me do a Google search. Let me see where else this has appeared. Who’s shared this on Facebook? Has BuzzFeed written about it? Has The New York Times written about it?”
FS: It’s also a little bit cruel, deliberately cruel. I mean it’s a worthy target.
JP: This is also post hoc. With hindsight, if you think that Chelsea and I would be doing this today, it would be more cruel. But at the time we were doing it, Chelsea was a little nervous about making sure that the humor was right and that it would communicate what we wanted to communicate and not actually have the opposite effect that we wanted.
FS: You wanted to find people all over the quadrants?
JP: We didn’t know these quadrants existed then. The effect we wanted was to, at least in my mind, satirize shallow liberals and their relationships with people of color that they were using for their own vanity instead of something real. So it was like a critique of a particular type of college liberalism that had gotten out of control. The idea was that it would satirize those people and maybe annoy them a little and then for more enlightened people, it would be a funny joke and they would get that it was a satire of those people. But it was made for this smaller group and it blew up in a much bigger way than we anticipated.
We thought it would go viral within one group, and we saw it start jumping from different groups. It would be on hair-care message boards but also a car message board, where people are talking about different kinds of cars and then people are on there debating it and white power groups. The thing that was pretty eye-opening to me, just from a technology standpoint or from a network standpoint, was that you could look at your referral stats and you could see traffic coming from these different places.
On Being too
Soon in Tech
“The thing is, if you care about having an impact on the world, the too-early mode is the highest leverage point.”
Back to Top
Felix: You have this charming narrative. “I was teaching in New Orleans and I went to the MIT Media Lab, and then I was living in New York.”
Jonah: And none of it’s true. I know you won’t do any fact-checking because it’s too much work.
FS: Oh god, can you imagine.
JP: No record of him ever living in New Orleans.
FS: Probably, if this was a Playboy thing, they would do that.
JP: Medium’s fact-checking team is—they might actually have fact-checkers at Medium. I wouldn’t be shocked. [Ed note: We do. Don’t be shocked.]
FS: Tell me a little bit about what’s going through your head, and how you’re making a living, when you were in Boston and when you were in New York before HuffPo and full-on media life.
JP: In Boston, I had saved a little bit of money from teaching. I made $24,000 a year in the first year, which in 1996-97 in New Orleans was amazing. I had a shotgun apartment, I could go out to eat whenever I wanted. It was my first job and it was awesome. It took maybe twelve years to get to the point where I had an apartment as nice as I had in New Orleans. Then the MIT Media Lab gives a scholarship to every student, so you’re essentially a paid researcher. I could live like a grad student and break even. I think I spent $1,000 from savings or $1,500 from savings in the two years I was there. So that was awesome. And then I got hired to be the director of R&D at this urban technology non-profit in New York, called Eyebeam.
FS: Eyebeam, of course, yes.
JP: So I worked at Eyebeam and the Nike email had taken off and the Rejection Line was just sort of between Boston and New York. We hadn’t, I think, launched it yet until we were in New York. Then at Eyebeam I was able to have my work be these projects, so Black People Love Us and Rejection Line were done at Eyebeam. Then there was Fundraise, which was another thing that was the first geo-coded campaign-contribution database that allowed you to type in your building and see who gave in your building.
FS: I remember that.
JP: That also was a very viral thing that I did with this guy, Mike Frumin, at Eyebeam. By the second week people in Milwaukee were like, “The Brewers gave to Bush and this is how much.” That was all work I could do at Eyebeam because Eyebeam was created by John Johnson and he was from the Johnson & Johnson family and he had been a filmmaker, an artist, and a patron of the arts, and he built Eyebeam as this creative space where artists and hackers and engineers and activists could come together and do projects.
I ended up building the Artist In Residency program there and then starting something called the Eyebeam Open Lab, which had a free printer and a laser cutter and electronic soldering benches and did residency programs for a year, renewable for a second year. I wrote a grant to the MacArthur Foundation and they funded the Open Lab. The concept was R&D for the public domain. Instead of a private research thing where you figure something out and then the IP is owned by the university or by a corporation, we would make things and then publicly license them, whether it’s through an open source license or Creative Commons or just DIY instructions for hardware.
So the MacArthur Foundation was like, “Oh that’s cool. If we fund that, then other organizations can use the work and it can have a bigger impact.” And then I would do these creative technology groups and contagious media groups in the evenings. I would order from Giovanni’s Pizza in Chelsea and it was usually groups of six to eight people and we would brainstorm ideas, talk about projects, build projects, show each other projects. Josh Schachter was building Delicious, and all he ever talked about at these meetings was Delicious. And we did Fundraise.
We did Re-blog, which was this server-side RSS reader that let you read your RSS feeds from any computer. And then we had this little Re-blog button and if you clicked it it would push the post onto your blog or onto a public-facing blog and it would say, “From: kottke.org re-blogged by Felix Salmon.” And if someone was reading that, they could do it and it would show the chain of attribution. We built that because Eyebeam couldn’t afford to have a full-time blogger writing about stuff happening at Eyebeam, so we’d have a guest re-blogger and they would come in and they would have a feed and they would just re-blog the stuff they liked. That was the first instance of that particular form of sharing. We had several people come to us and say, “This should be a company. Dude, turn this into a thing.”
I think we were just like, “Oh, we’re not really interested in starting a company” and “That sounds terrible that we’d have to maintain all these servers and host this,” and like, “The idea is already out there and that’s what it’s about.”
FS: This was all too early, wasn’t it? I mean Dodgeball was too early. In a way, Blogger was too early. And then Tumblr just sort of hit at the right time.
JP: Tumblr was starting a little bit after that, or maybe even overlapping, but Tumblr didn’t have re-blog.
FS: Oh really? When Tumblr launched?
JP: Yeah, it didn’t have re-blog when it launched.
FS: I thought that was there from the beginning.
JP: No, it was based on tumble logging, which was this convention that some people were doing that [Tumblr founder] David Karp saw. It was short snips of text and images and pieces and that would be your tumble log, but it wasn’t social. So he kind of took the re-blog idea and the tumble log idea and an amazing design and incredible focus and turned that into something that was sort of between the two of them, sort of a social network or sharing thing. But there were so many things like that in that era where people were building these little half-baked things.
Like [Computer scientist] Jeff Hahn was making multi-touch displays and he would come to the technology group things, and he was working on these cool hacks, and then, all of a sudden, they’re on CNN and soon after that they’re in everyone’s pocket in the entire universe. I don’t know whether Jeff Hahn licensed technology that ended up in the iPhone or CNN thing. I think that he was doing the CNN one for them, but he was mostly too early.
The thing is, if you care about having an impact on the world, the too-early mode is the highest leverage point because you can have an idea, build a mock or a prototype of it, and then have those ideas find themselves in products that other people build that then scale up to massive.
FS: You’d rather be Xerox than Apple?
JP: People always talk about Xerox as a sad story.
JP: I mean some people do. If all you value is money, then it’s a sad story. But if you think that the graphical user interface is a cool thing and you worked on the graphical user interface at Xerox, you can feel like, “This has a big impact on the world.”
FS: It’s a way of looking at the world through a lens of capital rather than labor. I’m sure the people who invented the graphical user interface at Xerox are doing very well for themselves right now.
JP: Right. And some of them have a sense of personal satisfaction that they had a big impact, even if they didn’t profit from it at the scale that they could have. Like Delicious was so influential and now people don’t really talk about Delicious. It was Union Square Ventures’s first investment of that kind. It essentially informed their whole thesis of a network of engaged users, watching Delicious and watching it grow. I don’t know if you were ever a big Delicious user.
FS: I kept on trying and I never got it. It was hard for me. I don’t know why it was hard for me but it was.
JP: It was hard. You compare it to Tumblr and Tumblr was just so much easier and better designed. [Delicious founder] Josh Schachter is brilliant but obtuse. But when people were using it actively, if you cared about the kinds of topics they were talking about, you could look at the stream for the tag and you would find just all these things on a topic you’re totally interested in, streaming in from a tag. It then led to tagging in Flickr. Flickr was like, “Oh, Delicious is adding tags—let’s add tags in Flickr.” So anyway, I’m feeling very old talking about all this stuff.
FS: And everyone wound up selling to Yahoo in the end.
JP: Yes. Flickr sold to Yahoo and then I think, in part, said “Delicious was where we got some of our greatest ideas from.”
FS: That’s always a good way to kill a company.
JP: Well, there was a moment of people being really excited about, “Oh, Yahoo has Flickr and Delicious,” and then that moment passed. But there was a pretty generative early tech scene in New York and Eyebeam was—
FS: In the middle of it. You did that physically, you just found those key nodes and said, “All right, you guys are coming over for pizza.”
JP: The other key thing is I had a very modest budget, but I had a credit card and so I could decide that it was a good use of budget to order beer and pizza and invite six or eight people over every Thursday night or Tuesday night or whatever it was. Josh was working at Morgan Stanley at the time and Paul Berry was working at Charter Mac, which was a publicly-traded real estate company, he was their CTO. [Vlogger] Ze Frank was doing his own Ze Frank stuff, and I’m trying to think who else. Jeff Hahn was at NYU as kind of a researcher-in-residence. Then we’d have guests come through too. The other thing was, Eyebeam was a great place for people to come and share ideas and they would share ideas so freely because it was a non-profit.
Seth Goldstein [co-founder of electronic music company DJZ] was the one who really wanted to invest in this re-blog idea; he’d crash meetings: “Hey, why don’t you turn this into a business?” And it was too early in that nobody really wanted to make any of this stuff into businesses and saw that as a distraction. It would lead to not being able to generate as many ideas.
FS: That’s [technologist] Anil Dash’s whole theory about the web we lost. The minute it all became a business, it all died, in a way.
JP: I don’t know I’d totally agree with that. Scaling things, and building a business, and the data that you have when you grow something to a large scale, does allow you to learn certain things that you can’t learn in a lab. The thing that bothered me about Eyebeam was that you’d do some amazing project or event and it would get attention and people would love it and it would be a cool idea and would make people think about new ideas and get excited, and then, at the end of the project, you would start back at zero or you’d have to go write another MacArthur grant, which would take two years.
What I learned first at HuffPost is that if you do something and make a splash and build something interesting, then people will give you money to do more stuff. They come to you and say, “Why don’t you take this to the next level? Let us invest.” And then you generate revenue, and that allows you to explore more ideas. Then you start saying, “Oh, wow, we’re at a scale that starts to be significant relevant to the web as a whole. So we can see, based on that, some things about how people behave and how the ecosystem works.”
I did become a convert to building businesses and start-ups. But at the time that I was at Eyebeam, I wasn’t really interested in that. I wasn’t interested in business and I was almost like, “Oh, this is just something that constrains you and doesn’t let you explore ideas as freely.” That’s remembering how I thought then, not how I think now. At Eyebeam, I would do a project, it would go well, and then at the end of the project I would have zero budget again and have to start back at zero.
Kenny [Lerer] was the one that got me excited about doing business. I wasn’t interested in Huffington Post primarily as a business. I was like, “Oh, it’s a cool new opportunity. It’s something different. I’ve been at Eyebeam for a long time. We’ve had this Bush guy in office for a long time.”
“What Ken Lerer said to me was, ‘I know business, you know the Internet, let’s build something together. Let’s start
Back to Top
Felix: You’re about to become a publishing person and join the Huffington Post.
Tell me the one about how Kenny finds you living out of a cardboard box under a bridge somewhere and says, “I want you to co-found a company with me.” What’s the story there?
Jonah: Kenny was doing some political projects, political work. He called together a group of luminaries who knew all about the Internet to help him with his political projects. He had John Borthwick invite all these people. I was not one of those people, but Duncan Watts [principal researcher at Microsoft] was. Duncan was a good friend of mine, and Duncan said, “Oh, there’s this rich guy and he’s having all these people come to his office to talk about political stuff. What he really wants is you. You actually make things that spread on the Internet. You should come along.” I was like, “Well, I’m not going to crash this thing.” He’s like, “Come on. Come with me.”
So [journalist] Jeff Jarvis is there talking about blogging and someone else is talking about MeetUp. There’s someone there who does mobile stuff and there’s someone there who does ad stuff.
FS: Arianna’s there? Not yet?
JP: Arianna’s not there. Arianna’s in L.A. or wherever she is.
We ended up working together on Tom’s petition. That was a project that Kenny was doing with the Brady Campaign. Tom was Daniel Mauser’s dad; Daniel died in Columbine. We built this six degrees of separation kind of visualization so that he could email his friends, they could email their friends, you could track it around the country. Kenny was paying for it. He just believed in the cause.
It was interesting for me because I had done things with no connections in the world at all. Everything we made spread from the ground up. We launched with a Bob Herbert column in The New York Times about the initiative. It was super interesting for me to see the way that you could do things on both sides. You could have a model that was based on people sharing things, certain things propagating on the Internet, but you also could have an important cause and call up a reporter that you’ve known for twenty years and have them write up a story. There was this world that I didn’t know anything about that Kenny was enmeshed in.
What he said to me was, “I know business, you know the Internet, let’s build something together. Let’s start something.” There wasn’t an idea of what it was going to be, defined. It was just like, we’ve worked together on these things that have been successful. We bring in different pieces to it. It’s very complementary.
FS: And you said yes without really knowing what it was going to be.
JP: I said yes with the idea that we’d figure out what it is in the future. Then he went to L.A. to this event that he almost skipped, and ends up meeting Arianna. Was just struck by how many people she knew in different worlds, and who looked to her to have intellectual conversations and discuss important things of the day, and realized that that was happening in her house in Brentwood, but not happening on the web. The explosion of blogging was happening at the time. It felt like, “Oh, what if this was happening on blogs? What if this was happening on the Internet and not just happening in her house in Brentwood?”
He came back very excited, but just started talking about like, “Oh, our company with Arianna.” I’d be like, “What?” I remember not really knowing who she was, and Googling, and being like, “Wait, that’s who were doing this company with?”
FS: Did Arianna know what an Internet was at that point?
JP: She had the site Arianna Online. She would put her syndicated columns up there. I think both Kenny and Arianna learned an incredible amount from building the Huffington Post. A great way to learn about the Internet is to build one of the biggest sites on the Internet. It’s hard to remember how much we’ve all changed, because we’ve changed so much. In their case, they were more fully formed than I was, but they changed most in their familiarity and embrace of the Internet. I’m, like, day-to-day fully immersed.
HuffPost, like any other company, had to keep reinventing itself, and changing its model, and evolving, and building technology, and moving from linking out to other sites on the web to making more content.
FS: This is the bit where it gets very interesting, because this is the first time that you even do news.
JP: I didn’t do news myself. I was focused on tech and product and traffic and that kind of thing. When we started, Andrew Breitbart was a partner.
Andrew had worked previously with Arianna and helped her with Internet things. He went and worked for Matt Drudge after working for Arianna. When Kenny met Andrew he was like, “This guy is an idiot savant of Internet news.”
FS: He was the Neetzan Zimmerman of his time.
JP: Andrew was the guy who spent half the day, eight hours a day, posting the links on the Drudge Report. The other eight hours Drudge posted. They could have sixteen hours of coverage. Andrew took the morning shift.
We learned a lot from Andrew in the early days, about how the Internet news cycle works. He knew how to go late at night to the British sites to get the morning papers as they were coming out. He knew how to write headlines that were irresistible and how to position stories with a headline. He was, early on, the one who was sort of obsessed with the news.
I was just trying to solve this problem, early on, of how do you make something that is both contagious and sticky.
I knew that Arianna’s friends, who were on television but not on the Internet, were going to create a sensation if they blogged. You remember when celebrities first started tweeting? Everyone freaked out: “Oh, I can’t believe a celebrity’s tweeting.” Blogging at that time was all about people in their pajamas, the person who couldn’t be on TV, the little guy who finally has a voice. I knew that people would freak out if Larry David or Tina Brown or a senator or a congressman started to blog. We knew we could make that happen. We already had all these people committed, from Alec Baldwin to Larry David and Tina Brown—all going to blog for the first time for us on Huffington Post in the first week that we launched.
I knew that that would create a sensation, that some people would hate it, and that some people would love it, but that it would be an incredibly exciting development that every blogger would have to watch, and be excited to watch, and want to form an opinion on. That was what would generate the viral spike. The question was, all my previous projects did a spike and then crashed.
So I was like, “What keeps people there?” One of the stickiest sites on the Internet was the Drudge Report. With always-fresh headlines and splashes and things like that, people would come back every single day. That site was built over a really, really, long period of time, and had the Lewinsky scoop and other things that drove its usage. Tons of people have created clones that never took off, because they’re sticky but not contagious.
There were these two models that we just kind of bolted together. One was to make the site itself viral, which was celebrities blogging. I was very focused on making sure that they used the default blogging tools of the Internet. I think that everyone expected us to have some Flash site that wasn’t a real blog. I went to Ben and Mena Trott, who had started Six Apart. We used Movable Type, which at the time was the premiere blogging platform. We modified it a lot in order for it to work, but it had permalinks it was reverse chronological. It had all the things that blogs were supposed to have so that people who knew about blogging would see it and say, “Oh, Larry David is blogging.” Not, “Larry David’s doing some weird new thing that Arianna Huffington invented.”
We knew that was the piece that was going to make it take off and be contagious. Then Andrew posting links and headlines that were constantly updated would be the thing that made it sticky. You’d come to see the celebrities blogging, you’d say, “Wow, what does this mean? That blogging has evolved in this different way.” And then you would say, “Oh, there’s a good link here. There’s a good link here.” And you would just keep coming back every day. Even if Larry David didn’t blog again for three months, you’d be checking the site because you’d have great links to content around the web. That was sort of the idea.
FS: This is about the same time as Sploid, which was trying to do the same thing.
JP: Right, Sploid had the sticky part, but not the contagious part.
We had both, but I didn’t know if it would work. We launched with Arianna on the Today Show, big traffic spike, hit a peak a couple days later, decline which I expected, and then I was like, “Where is it going to plateau?” It did plateau at a pretty high level. There was actually a little bit of investment from John Johnson at Eyebeam. He was like, “Should I invest?” I was like, “I don’t know. I don’t know if it’s going to crash down and not work.” Then the plateau was very high, so I said, “John, you should invest.” John was the last money in, and invested in the very first round, in part because I told him that.
FS: Most of the money was Kenny, presumably.
JP: No, there was a bunch of investors. It was a friends-and-family kind of round, but I didn’t have any friends or family who could put any money in. There was no institutional investor.
FS: How big was that first round?
JP: It was, I think, around a million dollars, something like that.
FS: Okay, so small.
JP: Well, I guess small to you.
FS: You had, what—
JP: Like a dozen people.
FS: You can’t support twelve people for very long on a million dollars.
JP: We raised again once the site was off and running. It was a small team and it wasn’t like people were making huge salaries. It was a small little upstart. It’s hard to remember that Huffington Post used to just link out and didn’t have any article pages.
It was really only in the run-up to the election where we started to realize, “Oh, we need to add things other than politics.” It was all politics when we started. Right after election, we’re like, “Oh, people are coming, traffic’s growing, but we need entertainment, we need media, we need other verticals.” BuzzFeed was around at that point and I would put a little BuzzFeed widget on posts to see whether the audience would click on it. People clicked on celebrity just as much as politics. I was like, “Oh, I bet celebrity will work well in Huffington Post.”
FS: So your original vision of grabbing people with boldface names and then keeping them with news—was that a success?
JP: That was a success. It wasn’t a model of where the company went, but it was a success as a way of launching the company and getting to a plateau and then getting to grow.
I think that’s also why HuffPost, to this day, has a large front page audience. In part, it started it that way. Part of its DNA was thinking of how to build an entry point or front page to the web.
FS: The last major media company based around the front page.
You do look at these things in hindsight, but certainly the point at which I started really paying attention to HuffPost, it was full-time staffers writing and editing content, who were the meat and potatoes of the site.
FS: At what point do you realize that blogging was not going to cut the mustard on its own?
JP: I’m trying to remember what the evolution was. The site was growing but the election was a big part of it. We knew we had to launch other verticals and sections so that after the election, people would see HuffPost as being more than just a political site.
We launched those verticals and had to freeze development on all other features. It was a long delay. It was very painful. Everyone had things they wanted built or done. We couldn’t do any of them because we were launching a model to have five or six front pages. One for entertainment, one for media, one for … I forget what the first verticals were, entertainment, media, maybe lifestyle, there was a few others. We launched them and there wasn’t really—they didn’t really generate much traffic.
Part of the problem with a home page model is that you have one home page and people don’t really like to go to the secondary pages. We did a lot of work and it didn’t have that much of an impact except that we had hired editors for each vertical and we could put some of those links on the main home page and those were getting clicked and people were seeing those and seeing that the site had done other stuff.
What we did after that is we realized, okay, now this vertical section needs to start having more original content and aggregated content. We started to have the entertainment editor use photo wires plus do original writing plus link out to places, and make their own page, and that was something that started to transform the company from being an aggregator on the front page, to being an aggregator on the B page.
From there it went from to carrying original content as well. That was sort of an evolution.
Around that time, at BuzzFeed we had figured out that you could rapidly swarm a breaking news topic, particularly about a person, place, or thing that was new, like a beauty queen who loses her crown and no one’s heard of this beauty queen. If you make a great page about that thing, you often could get to the top of Google results just as searches were surging. It was partly because Google got faster indexing at that point. Google was slow indexing and then all of a sudden became quick, and BuzzFeed figured that out in the lab, but then HuffPost editors got really good at it and we’d swarm stories very quickly and often be the first news source to create a comprehensive page for what was happening, linking out to other multiple other sources. Those pages became huge growth generators for the site.
FS: That was the high point of SEO as a business model, basically.
JP: It was something where there was a model that HuffPost had that other people didn’t really fully understand, and that HuffPost got good at.
FS: You stayed one step ahead of everyone else for a while on the SEO front.
JP: At its best, it was pages that were pretty valuable to the user. Heath Ledger died and people wanted to know what was happening. They do a search and the HuffPost page would have regular updates and links out to the most relevant stories that were happening other places. It was a great page to keep informed of what was happening everywhere. It was clearly written, there were six dedicated people all working on it. At its best, it was great for users and great for the business because it was allowing the company to scale and grow really quickly.
When you look at that period, the Huffington Post was much, much, much bigger than it was when it was a blog and a left-leaning Drudge.
FS: Obviously, at its worst, it’s “What time does the Super Bowl start” and—
JP: I don’t know if that would be the worst. What time the Super Bowl starts is actually useful, if you want to know what time the Super Bowl starts. I don’t think that was ever the best example to make fun of the Huffington Post. Now Google just puts that information natively at the top of the search results.
At its worst, SEO is someone finding an inferior article instead of a superior article because the site does a better job of writing for Google, but a worse job for writing for humans. I think that’s when, in my view, you’re over-optimizing. You end up reaching people who are not satisfied with what they’re getting from you. Which isn’t good for a brand.
FS: That was the standard criticism of HuffPo, was that it would aggregate these stories and put long lists of wonderfully SEO-optimized tags on the end of them. It would do no real linking and writing, but it would just be parasitical on everyone else.
JP: I think there’s an interesting tension between what’s good for the user and what’s good for the industry. That was really created by Google. Say The New Yorker writes a really long 12,000 word piece on Scientology. That takes lots of reporting and lots of investment. That’s important work that our
industry should embrace and should find ways of supporting economically.
The average person who hears about that story doesn’t want to read the whole story. They’re at work, most likely. They do a Google search because they’ve heard about this Scientology scoop or long form piece. Their first result is the HuffPost link, not a New Yorker link. They look at it. It summarizes what the article is about. It says, “Here’s what was in it, here’s what was notable about it.” Has a few tweets from people. This is how people are reacting to it, and if you want to read it, here’s a link and you can go read the article.
The problem with that example is that from the perspective of the user, it’s a better experience to land on the summary, to see a little bit of the reactions, and have the option of reading it, because that’s as much as most people want. From the perspective of the industry, it would make much more sense for people to go to The New Yorker article so that they get the traffic, as modest as that ad revenue would be, they get the traffic and they get the people onto their site. There is some conflict between Google saying, “Well we want to serve the consumer,” and sending people to the article that the consumer likes the best. Or is Google supposed to send people to the article that costs the most to produce and supports the industry the most? Does that make sense? There is a little bit of a conflict, or a little bit of a tension.
FS: How much of HuffPost’s success do you ascribe to tech, you being able to do stuff on the tech side which no one else could do?
JP: People always overestimate their importance to the success of the company. When you talk to the people who are on the sales side, they say, “Well, you know, we drove revenue. That allowed us to invest in all these things. None of the rest of the company would have even been possible if we hadn’t driven that revenue.”
You ask the tech people, the product people, they say, “That’s the competitive advantage of the company. All the other companies had great editors but we had the better tech.” Then you ask people who are on the editorial team and they say, “Well, if you get a scoop, people have to link to it no matter where it is. Great editorial content is really what drives the traffic. The CMS, it can be broken and then stop you from being successful, but if it’s good enough, then editing really is the key and so we really drove a lot of the success.”
When there’s a startup that sells, for example, or there’s a startup that’s super successful and is growing, people’s view of who drove the success is very highly correlated to who they know at the company. Chris Dixon will say, “Oh, HuffPost is really a tech company and Jonah was a big important part of it.” Because he knows me. But then someone who’s friends with Arianna will say, “Arianna’s a force of nature. She is constantly on TV. She was the name and the voice of the site. Her blog posts were constantly in the news cycle. That’s where the site mattered.” If you talk to someone who knows Kenny, they’re like, “Oh, Kenny was behind the scenes, building this whole thing and planning this out. He’s done it before, he’ll do it again.”
That’s true, not just for HuffPost. It’s true with most companies. The thing you’re closest to, you think has the biggest impact. That, I think, is the cognitive bias that makes it hard for me to answer the question. The system as a whole matters so much. Any one of us who thinks that they’re solely responsible for the success of the company is, by definition, wrong, and the relationship between all the different pieces of the company has to be strong.
In some cases, there’s things that aren’t even measurable. Like maybe just having tech, edit, and business teams communicating effectively, is more important. The lines might be more important than the dots.
FS: But at the same time, if you’re part of a bigger legacy organization, then you want less communication, broken lines.
JP: It depends on how much they can change at all. I think that Janine Gibson coming out to New York and doing something new, a little bit away from Guardian headquarters, is probably a good way to innovate. BuzzFeed’s video operation is in Los Angeles, far from New York, and it’s really thrived in a tremendous way, and part of that is that it wasn’t in the early days seen as a companion to the kind of editorial content we’re doing in New York. Sometimes that helps. To feel like “This is a new startup,” as opposed to “This is an offshoot.”
FS: I think we’re doing that at Fusion, as well. I think that Fusion is being set up in Miami, which is quite a long way from the more conventional media centers. The Fusion digital team in New York, again, is away from the Fusion TV bit in Miami. The distance can help. It can allow you to be a little bit more innovative and dynamic.
JP: Yes. If you’re trying to build something that is disruptive in the [Harvard professor] Clay Christensen sense, within an existing organization, his research, at least, shows that it’s better to do it in a different physical location, even if it’s in the same city, because it’s just too easy to think, “Oh, let’s look for synergy. Let’s bundle these things together. Let’s do joint ad sales across digital and print. Let’s make sure that the digital stuff is easy to package into stuff that can be printed.” You spend lots of time with your mind focused on the relationship between the existing business and the new business you’re building, instead of just thinking about how to reach the full potential with the new business.
FS: Which is one of the lessons we can draw from the success of All Things D, right? That’s something where Kara [Swisher] and Walt [Mossberg] were absolutely right. They insisted on having basically no contact with the mothership. They were on the other side of the continent, and they did their own thing, and they did it very successfully. In large part because they weren’t constrained by the politics of News Corp.
JP: And they built what to them was a big business, but which to News Corp looked like a very small business. The problem is, when you’re embedded in a company where what you’re doing looks like a very small business, it doesn’t get the attention, and even people who work there have trouble taking it seriously because they’re like a little flea compared to the other activities going on in the business.
FS: What was your business, at HuffPost? You were working with Kenny and Arianna and Andrew to create this site. Was it always, “We are going to build a site which will have lots of traffic and then we’re going to sell that traffic to advertisers, that’s the business model”? Was that always there from day one?
JP: We never talked in those terms ever. It wasn’t like we were two Stanford computer science graduate students who were like, “Let’s go build a new way of indexing the web.” Or something where there’s a much more aligned kind of view of the world. The worlds that I came from, and Arianna came from, and Kenny came from, and Andrew came from, were very different worlds.
FS: You were there all the way through the sale.
JP: But I wasn’t there full-time all the way through.
BuzzFeed as Willy
“It wasn’t easy raising money for BuzzFeed. It was always, ‘Is there any way you can do this without having any writers or content creators or journalists?’”
Back to Top
Felix: When did BuzzFeed start? How did BuzzFeed start? Was that after HuffPost started taking off,? You’re having lunch with Kenny one day, he’s like, “You know what? You should just create this little viral factory.”
Jonah: I kind of missed, a little bit, the Eyebeam environment where you could just experiment with ideas. HuffPost was a real business and we had goals. We were hitting them, and we were growing the company. There were certain ideas that I wanted to play with, particularly because the company that Kenny and I ended up doing, HuffPost was more of a news company and a little bit less of a social content company.
FS: It becomes a news company, basically.
JP: And it’s more serious. It was hard to do playful experiments on a site that was very serious, without it conflicting with the mission of the site.
BuzzFeed was really like, “Oh, okay, I have a lab.” It doesn’t cost much to run. John Johnson put the first money into it, and Kenny. I can play around with ideas there. Sometimes when there’s an interesting thing we figure out I can just apply it to HuffPost, which we did on various occasions. It was not really seen as a separate company, it was more like an R&D lab that was separate, with some of the same people involved.
FS: It was operated as a for-profit company in terms of funding and stuff, from the beginning.
JP: It was an LLC. It didn’t have any institutional investors.
FS: Was it non-trivial, raising money for BuzzFeed?
JP: Yes, it wasn’t easy raising money in the early days of BuzzFeed. It was always, “Is there any way you can do this without having any writers or content creators or journalists? Can you make this automatic? Could you detect what was trending and then grab stuff from other places and turn it into an article synthetically where the cost of content creation would be zero?”
FS: That’s always the big thing in Silicon Valley. They buy Nick D’Aloisio just because it’s all algo based.
JP: It can be algorithms or it can be the user doing the work for free. Twitter and Facebook are exciting because their cost of content creation is zero. You can imagine them scaling without having to hire lots and lots of people. WhatsApp is exciting to them because you have fifty-some employees and reach massive scale.
The kind of leverage you can have if you’re a tech company is really phenomenal, but when you look at media companies through history, they’ve had a different kind of leverage. If you’re a newspaper that has a circulation of five million there’s a lot more leverage than if you’re a newspaper that has a circulation of one hundred thousand. Media businesses can have leverage, but it’s different.
FS: At what point did you start concentrating increasingly on BuzzFeed?
JP: I went from being one day a week on BuzzFeed and four days a week on HuffPost, to being half and half, and then to being one day at HuffPost and four days at BuzzFeed. Although, it was hard to just to go to the manager meetings on Mondays at HuffPost, and it was very hard to not spend lots of my time when I was at BuzzFeed thinking about HuffPost. HuffPost was scaled up and there was all of this stuff going on, the stakes were really high. There was politics. Navigating all these things, my brain cycles were kind of stuck on HuffPost more than they should have been. When HuffPost sold, it was possible to say, “Oh, I don’t need to think about this stuff at all.” All of that energy and focus went into BuzzFeed and what should BuzzFeed do.
We had this little widget on the front page of Huffington Post, and that was sending traffic to BuzzFeed. I was like, “Okay, we’re getting the tie-in at HuffPost which benefits BuzzFeed too, so I don’t have to feel that guilty.” But at a certain point, that widget was removed. I was like, “Okay, now the traffic’s going to drop at BuzzFeed—now I’ve got to figure out how to make BuzzFeed stand on its own two feet.” That’s when I went half-time.
FS: This is actually something you’re incredibly well suited for, because what you do is create things which grow, and this is what people want to invest in, things which grow. Rather than, “This is a nice corner store which I can use to support my family for the foreseeable future.”
JP: Yeah. Right or wrong, Wall Street and VCs and angel investors are very focused on growth. It makes a huge difference to the valuation of the company if it’s growing fast or if it’s growing slowly.
FS: [Venture capitalist] Paul Kedrosky, a few years ago now, did this study for the Kauffman Foundation, where he says that, if you look at all the fastest growing companies in America at any given time, only 5 percent to 10 percent of them are taking VC money. Having an emphasis on very fast growth is probably necessary to get that money, but it’s definitely not sufficient. Most fast-growing companies don’t. That’s probably a good thing, really. Because the VCs, as you’ve discovered, have a habit of biting off parts of your anatomy, which you’d really rather—
JP: I like venture-backed companies. I think VCs in general are great to be partnered with if you have the same goals. If you want to build a giant company relatively quickly, in an emerging space where the market isn’t mature enough to generate a lot of revenue early on, then VCs are the perfect partner.
If you want to own 100 percent of a company and build it over twenty years, even if it’s a fairly fast growing company, in a market where revenue is booked quickly and ideally where the revenue comes before you have to sell the product so you have nice cash flow to build a business from, then it’d be great to do it without VCs. If you want to build a business that generates $10 million a year in revenue and $2 million a year that goes to you and your partner, and that’s what you’re interested in, then you should figure out how to bootstrap. A lot of agency businesses are that way. When you look around New York, there’s a lot of really successful fifty-person boutique agencies, where the people who started them are making a lot of money and no VC would ever touch them.
If you read the tech blogs, they converge into this idea that the only way to build a good company is a venture-backed company and it probably should be in Silicon Valley, but maybe it could be in New York or a couple other places, and no other company matters. That is clearly false. If you’re a budding entrepreneur who wants to start a business, depending on the idea and depending on your temperament, it might be great for you to take VC money, or it might be terrible for you to take VC money. It just depends.
FS: I feel like you like the discipline, you like the VCs.
JP: I like that they’re all steeped in, or a lot of them are steeped in, Clay Christiensen’s philosophy. I like to be working in sectors where there’s no clear model to find yet, where things like social and mobile and online video or things that are still evolving so quickly that there’s not a clear, stable business there yet. I like to be in the kind of places where things are being formed. I like growth-oriented companies. It’s exciting to me to have a different job every year because the scale of the organization is different and the types of things that you’re able to do are different and the kinds of problems you face are different. That’s exciting to me, more than having a stable company. I like the dynamism of that. VC-backed is great.
But it’s not for everyone. For example—this is a totally different example— 4chan should not have been VC-backed. [Founder] Chris Poole has resisted the idea of funding it, even though it’s a big site and you could expand it in various ways.
FS: And famously, Mashable never took outside funding until a couple months ago. Which may or may not have been the right decision.
JP: Yeah. It’s hard to know. It depends on what your goal is.
FS: Your goal was… what? BuzzFeed has been around a long time. It started as a crazy skunkworks ideas thing.
FS: You then raised a round from SoftBank, and now it’s a real thing. Now you’re like, “We are going to be this thing which finds memes and just really manages to glom onto whatever people are caring about the most on the Internet.” This is before the whole concept of social.
JP: At that stage, the site was a proof of concept for the technology. We were thinking of building a technology platform, and then the site was a proof of concept. It wasn’t like, “Oh, we want to make the site big.” But if the site didn’t grow, the proof of concept wouldn’t work. We even had a conversation with The Washington Post about them using our technology to optimize The Washington Post.
FS: The technology was, basically, just trying to work out what people were buzzing about.
JP: We had something called Buzz Bot, which would IM you links. We built something called the Trend Detector, which we crawled from Jason Kottke’s blog to a bunch of different blogs. Then we would look to see acceleration of URLs being cited in blogs. Then that would show up on the Trend Detector. Then Peggy, our founding editor, would look at the Trend Detector for interesting stories. If the number one thing in the Trend Detector was, “Wordpress has a new plugin,” she’d ignore it. If it was a news story or a new trend or a new product, she would then write a little summary of the story and then put the best links below it.
The format for the summary was, imagine you’re at a cool party, you go up to someone and the first sentence is, “Did you hear about X?” It just sort of describes what X is. If they haven’t heard about it they kind of get it, and if they have heard about it, it doesn’t take that long. “So, have you heard about degentrification, where the prices of a neighborhood start going down?” or whatever. That’s the first sentence. The next one to four sentences were opinion or point-of-view or commentary or a joke or something. Sort of what you would say afterwards.
That was the beginning. It was just all these simple little “one sentence plus three sentences,” describing something that usually she found in the Trend Detector or sometimes she found in Greenpoint, hanging out with friends of hers. Six or eight or ten links about degentrification or about this new kind of shoe, or this new kind of fashion, or this new idea. Some of them were more trendspotting, and some of them were driven by this technology. That was the original BuzzFeed. We did five things a day. It was just these five things.
In that memo I wrote about Time magazine and the history of media, that was the period when we were kind of similar to what Time was in the early days, which is: There were a few editors subscribing to lots of newspapers, reading the newspapers, and then summarizing them into a more digestible form in Time so that an intelligent reader with not that much time on their hands could quickly digest the important stories of the day. Peggy was just doing the equivalent.
She was reading all these things on the Internet and using the Trend Detector, and then putting it in a very digestible format so that you could be informed of what were the trends happening on the Internet, what was propagating and getting shared across the Internet.
She was a former computer programmer and wrote in a very clear language, but also had a sense of humor. That was the very first product.
FS: Were you surprised, or not surprised, at how newsy it was? How much of what you might conceivably find in a newspaper would be in those five things? And how much of it was dogs on skateboards?
JP: It wasn’t really dogs on skateboards. In those days, it was more people, places, things that were trending, which often was news: the beauty queen losing her crown for some scandal. That was something that unexpectedly ended up being a top search result. We’re like, “Whoa, Google works totally differently now. It’s like, this has done as much traffic as a front-page Digg link, and it’s some story that just happened. Google used to take a week to index anything that was breaking news, and now they’re doing it right away. Oh, that’s interesting.”
It would be like, “Militaries are now using unmanned aircrafts to do strategic killings of potential terrorists,” then some little commentary on it, then links to all the stories about it. But it wouldn’t be, “Eight people die in Iraq in a bombing and the US responds with a drone attack on a terrorist and….” It wouldn’t be like a whole news thing.
FS: At this point, it’s more reactive. You’re not creating stuff which is designed to go viral. You’re just identifying the stuff which is already viral and amplifying it.
JP: Exactly. That’s exactly right. That was true for the first couple years.
FS: You can make a distinction between that and HuffPo, where HuffPo wasn’t chasing those things in quite the same way, although it knew what people were searching for.
JP: HuffPost was contagious and sticky. BuzzFeed was just a little sticky. Often, television producers would read BuzzFeed, and they’d say, “I’m getting these new trends and I’m always ahead of the curve.” People would describe it as a secret. Their friends would say, “How come you always would send me all these interesting stories?” None of them are BuzzFeed stories, but they were finding them on BuzzFeed. Every day, they were learning from Peggy what the cool things were on the Internet and finding them, clicking through to them, and then that’s what they would share.
In a way, it’s similar to Huffington Post starting with a front page linking out to other sites, and then making content themselves. We were doing a similar thing. Then we started to have a really good understanding of what people share and why people share things and what stories are interesting. And we said, “Oh, why don’t we make our own? Why don’t we make some of these things? Instead of just tracking culture, why don’t we influence culture by making great entertainment, great news, and content that people want to share?” Initially, it wasn’t news, but making great things that people think are worth sharing and passing on to their friends.
That became a focus. The editorial team grew to about twelve people. A lot of those few people, like Matt Stopera, who started as an intern, started making these lists and really narrating a story in such a way that it would get people excited and people would love it enough to pass on to their friends and share.
Around that time HuffPost sold, and then soon after we hired [former Politico writer] Ben Smith, and that was really the moment we went from being a tech company experimenting with content to being a company that took editorial seriously. Not just serious editorial, but that took entertaining content seriously, and news content seriously, and had a more ambitious editorial mission. A more ambitious sense of what it means to make content, or make original content.
FS: Did you hire Ben because you felt that it would give you a certain amount of respectability or gravitas that you needed in order to get your brand clients to take you seriously?
JP: It wasn’t clear to me that Ben would generate a bunch of revenue through gravitas. It was really intellectually driven in the sense of I was trying to understand, and am still trying to understand, why certain ideas spread and other ideas don’t spread, why certain things become popular, why other things become popular, other things don’t become popular, how people learn, how people make decisions, and to use that knowledge to inform people and entertain people.
We had this huge hole; it was obvious we had this big hole. It was particularly obvious to Kenny, who never loved the fun, playful Internet comedy. He wanted the site to be more serious too. It was internally very evident. Matt Stopera was saying, “we have one arm tied behind our back because we’re missing news and reporting which should be part of the mix of the stuff that matters, particularly when there’s a big news story happening and everyone is turning to it and interested in it.” That was a big part of it. It was like, “Okay, the web is growing up. The social web is growing up. We need to grow up. And we need to add capacity to do all kinds of stuff that we’re not doing now.”
Ben came in and instead of saying, “We’re not going to do any of the fun stuff,” he said, “we have to do more of that. We have to do it faster. We have to do it better.” While simultaneously admitting it wasn’t his expertise. He really empowered people to step up and do more entertaining web culture type content, while also hiring reporters first for politics and to go after the election cycle and then in other areas.
FS: Ben kind of got it naturally?
JP: No. He said, “I get what you’re doing. It seems really interesting. I’m not your guy.” He remembers the conversation as, interesting ideas, plus a bunch of jargon about the social web. Really it was Peter Kaplan talking with Ben, and also Kenny talking to Ben, but in particular Peter talking to Ben about how every election cycle, there’s a new site that emerges that captures the spirit of the election, and that BuzzFeed could be that site and here’s how I would do it. Peter had a meeting with Ben where he really made Ben switch from thinking “This is a good idea for someone else” to “This is a good idea for me.”
FS: Do you think that worked? Do you think in hindsight that BuzzFeed kind of nailed the 2012 election?
JP: I think we did. But it wasn’t until later, it wasn’t until the Boston bombings, that we saw people coming to us for news. We had reporters on the ground and we were the first site to authenticate suspect No. 2's Twitter account by looking at his avatar, looking at who he was following, getting two of his followers on the phone to confirm that they went to high school with him.
There were things that we were able to do because we were very savvy with understanding the structure of the social web, and because we had reporters. We could wade through Reddit and Instagram with more knowledge than a lot of traditional publications, but we also had reporters on the ground interviewing people, which was something that a lot of the pure web companies didn’t have. From that point, we said, “Okay, we see that breaking news is going to be a bigger part. Not just people coming via Twitter, but people coming to BuzzFeed for news, particularly when there’s big news.” That’s when we hired Lisa Tozzi [formerly of The New York Times] and then, after that, expanded our foreign coverage with Miriam Elder from the Guardian, hiring foreign correspondents. Now we have two people in the Ukraine.
We are still at the beginning of the investment in news, but it really started as breaking political scoops and earning the respect of insiders, to then having younger people coming to us for their news, to now expanding into a broader news mission.
FS: The election, you’re getting most of your traffic from Twitter, which we know is always a tiny fraction of the fire hose that is Facebook. For me, mentally, as Twitter is to Facebook, in terms of size and importance, so is news to the rest of BuzzFeed, the core viral stuff. It’s important, and it’s the kind of stuff which media people tend to notice. But in terms of attention or page views or whatever metric you want to use, it’s still always going to be the tail, not the dog.
JP: I think that’s true, not just at BuzzFeed, but if you look at a television network, American Idol has higher ratings than the evening news.
FS: Even so, it is using a massively disproportionate amount of your resources. You’re investing a huge amount in news, compared to cats.
JP: We’re definitely making a big investment in news. It is generating considerable attention and traffic and influence. I think you’ll see more of it.
How to Win the Internet
“What would this be if the readers and the publishers were not focused on making something similar to print?”
Back to Top
Felix: How much A/B testing do you do? On a scale from one to Upworthy?
Jonah: We do a lot of testing and we look at lots of metrics, but ultimately the raw material is people coming up with creative ideas. We have lots of meetings with five or six people siting in a room brainstorming about what they could create. You have a lot of reporters saying like, “Oh, what stories are we going to cover?” None of that is directly tied to any metrics. Although having seen what people’s response is in the past, it actually might give people a gut feeling.
FS: You have to have some kind of a model. If you’re asking yourself, “What stories are we going to cover?” you need a criterion so that once you’ve decided to cover a story, you can then go back and look at that and decide whether it was successful or not. That’s going to be metric based. You can be like, “That was a good idea to cover that story because it did really well” or “That was not a good idea to cover that story because no one cared.”
JP: I’m trying to think how to describe this. It’s not that we celebrate maximizing the metrics. We want the stuff we do to reach the maximum audience it should reach, no less and no more. If we make a wonky political scoop, we want every political wonk to read it. If we make “Which state should you actually live in,” we want everyone who lives in a city to read it. Most A/B testing is just user experience. Is this page loading fast? Do the share buttons work well? It’s not A/B testing the content, it’s A/B testing the site and the platform.
When we have something that’s a hit, usually our response is not, “Let’s do more of those.” Our response is, “Let’s figure why this is a hit and make variations of this.” This was successful because it was tied to someone’s identity, it was successful because it had cats in it, or it was successful because it had humor, or it was successful because it tapped into nostalgia. If you’re making entertainment content, which is a big part of what we do, you look at that hit and you say, “Why was that successful? Can I do it again? Can I make something else that people really love and want to share?” And you try to vary it, even though you know doing something derivative would work. Long term, you want to have a deeper understanding of how to make great things. That’s really the focus. That comes from people in a room talking and saying, “Oh, let’s try this, let’s try that.” And valuing people doing new things, not just valuing people doing big things.
You see a very massive skew between different items. A problem I think in people writing about the media now is that people take averages and then they make statements like, “The average piece of branded content gets this many Facebook shares” or “The average longform story gets this many views.”
FS: Everything online is barbelled, right?
JP: Everything is just very skewed. When you say barbelled, it depends on what the graph is. But if you rank things in order of whatever the metric is and then you do a graph, it’s always this very skewed distribution. I feel like what you see in the industry now is people jumping around and trying to find the God metric for content. It’s all about shares or it’s all about time spent or it’s all about pages or it’s all about uniques. The problem is you can only optimize one thing and you have to pick, otherwise all you’re doing is making a bunch of compromises if you try to optimize for multiple things. So you pick the one that matters and maybe you have minimum thresholds for a few others. The problem with that is that the natural inclination, if one metric is seen as the important, true metric —
FS: Is to game it.
JP: Is to game it. And then when you game it, you essentially are creating a fake version of that metric. So page views are a metric of how many stories people want to read—and then you split the story in two. You essentially are doubling your pages for that story, or not quite, probably, because not everyone will click. But you create pyrite page views. If you’re looking at clicks, if there’s a story that says “Malaysian airliner, new developments,” and you want to know about new developments in that story and you read the story and there’s a new development, then you’re happy and the click rate is actually pretty accurate. If you click and there’s actually no new developments or the story is actually about something else—
FS: Or if it just takes you to a web page, which has a link to another web page with new developments—
JP: So the bigger issue is not which metric is the most important metric. You’ll see some sites talking about time as being the metric.
FS: That’s [Upworthy founder] Eli [Pariser’s] thing.
JP: He said Upworthy were switching. The challenge there is that, like you said, if you create a long, meandering, boring story that’s just good enough to keep people reading, they might spend more time on that story than the short, condensed one that just tells you what you need to know. If you use time on that one, it will tell you to do the wrong thing.
FS: The thing about time, with Eli in particular, is it’s skewed toward video content. People love being able to click Play and then sit back and watch a seven-minute video. It seems to me to be less engaged than if you’re actually concentrating and reading something for seven minutes.
JP: One of the reasons reality TV became so dominant was because people looked at time as being the metric. And the reason that reality TV works well for time is that the classic reality TV formula, in the beginning, was the tribal council and somebody getting eliminated. So you could have 50 percent of the show being boring filler and you’re kind of wanting to change the channel but you’re like, “Oh, but I wonder if my favorite person’s going to get eliminated.” So you have to watch to the end to see the elimination. In a way, that was a way of gaming time. You could look at that and say, “Oh, they spent an hour watching this show, including the commercials. That means it must be a really high quality show.” But it also might just mean that they figured out a hook that incentivizes you to watch to the end and then did a lot of mediocre content in the middle.
So there’s not going to be one metric that you look at. I love metrics and I love thinking about optimization, but I think that the optimal state is being slightly suboptimal because as soon as you try to actually optimize, particularly for a single metric, you end up finding that the best way to optimize for that metric ends up perverting the metric and making the metric mean the opposite of what it used to mean.
FS: Where does the discipline come in in terms of deciding not to do things?
You can take that viral value, and you can have the quick hits which people love to share, and the deep thoughts that people love to share, and kind of ignore a lot of the boring day-to-day stuff in between, which is eat-your-greens news. But at the same time, you want people building quizzes for years before you finally manage to get them right, and they suddenly take off, and you’re like, “Wow, quizzes, wow.”
Is there any kind of discipline involved in saying, “No, we’re not going to do that”? Or is that not part of this process?
JP: I don’t know that there is a list of like, “These are things we don’t do.” I think there’s some cases where it’s clear that someone is trolling for attention and they have fabricated some hoax story that is designed to get picked up everywhere and are pitching it to every editor of BuzzFeed and we send an internal email saying, “Don’t pick this up. Nobody touch this. Don’t get tricked by this person.” Even though you could just—there’s lots of sites who will run that story and get a lot of traffic.
We also have our no-haters orientation. We tend to be enthusiastic and we tend to avoid snarky articles about mediocre things.
It’s not like there’s some hard rule. In general, we tend to avoid a post that is designed to make the author feel smart and superior and the reader to vicariously feel smart and superior because a Hollywood film is mediocre or because something in culture is mediocre.
FS: Honest enthusiasm is a sort of default stance at BuzzFeed.
JP: If there’s something that is worth someone’s time that is interesting and is worthy of being excited about, we should cover that. If there’s an egregious miscarriage of justice or corruption or fraud or something that needs to be investigated, those are both strong things. In the middle, there’s a lot of things that are kind of a waste of time. Mediocre things that you can write cynical comments about.
FS: Is the church-state wall important to you?
JP: Yes, it’s important. Definitely it’s important.
JP: Well, particularly with reporters it matters a lot, but even with people making entertainment content. The idea that we are making a quiz that involves characters from a TV show, and then we are also pitching business to market that TV show—it creates a conflict of interest. If you put the consumer first and you say, “We want to make stuff that people love,” then the editorial team should not have to stress or worry about the business side at all. They should just make stuff that they think consumers will love.
That’s why the wall’s important, but when it comes to platform and technology then there isn’t a wall. It goes across the whole organization and company. When you abstract away from a particular piece of the content to how do we serve the content, and how do we optimize the content, what formats are there, that’s stuff where you want to make great formats for people whether it’s for a piece of branded content or whether it’s editorial content. There’s not a wall when it comes to things like data science and product and tech.
FS: Do you think this is going to turn into success for Ezra’s [Klein] thing? Brands want to be “thought leaders.” They love the idea that there’s a formula out there where they can get their ideas out into the world and they can explain things to the world more boldly and have that position.
JP: Yes, if he’s better than the market at that. Some of what you were describing earlier about digital publishers being small relative to the traditional media and relative to television, actually it’s because early-stage digital publishers have stayed too close to print. They look like print. Their basic unit is the same kind of article structure. Some of them might be shorter or longer, but the front page is programmed almost like a newspaper. The formats of the articles are more like a newspaper. And it’s like, “Oh, let’s add a little video,” but when they add video it’s like they are trying to be TV, but it’s not quite as good as regular TV.
The way to break through and to make something that can actually scale into something big is just to say, “What would this be if the readers and the publishers were not focused on making something similar to print?” If they said, instead, “What should this be if mobile is the most important thing; if things can be more visual; if things can be more shareable; if length can be anywhere from 140 characters to 12,000 words? In that kind of world, where things can be interactive, like quizzes—in that kind of a world, what should a media company be?”
I think if people can figure that out, they can build really big businesses. If you look at the branded content that you were just describing — the informative kind of branded content — it’s often like a special advertising section, and it’s the kind of thing that reads like a magazine article and it’s always like a page long and feels so of the world of print; whereas so much of BuzzFeed is sort of scrolling on a smartphone and that’s the best way to consume it.
Publishers should figure out how to re-invent what the informative branded content would be. We’ve had some success with list formats that are very informative. In a grand sense it needs to move away from mimicking print to doing what is natural for the web.
FS: What we seem to be talking about here is a world where what you’re selling is domain expertise. “We really understand what works in this world, for this audience, on these platforms, and if you want to create something which works in this world, for this audience, on these platforms, come to us and we can help you.” Which is very different from a traditional print or television media company where they’re selling an audience. They’re saying, “Here’s an audience, do with it what you will. We’ll give you a page; put something on that page. We’ll give you thirty seconds, put something in those thirty seconds. Do you need to be selling expertise more than inventory in order to really make it work in digital?
JP: When BuzzFeed started, people would ask me for expertise a lot because of the Nike email, the Rejection Line, Black People Love Us. It was, “Oh, you have this expertise on how to make things go viral — make my things go viral.” Part of the idea of BuzzFeed from the beginning was can we get to the point where the platform is more valuable than the expertise? If you actually have an idea or an insight into how things are shared or why they’re shared or what works, can you build that into either the technology platform or the data science or the culture of a team of people?
Or build an audience that reads the site and engages with the site in a certain way that it gives you an advantage to be able to launch something that will take off? If you can do that, that’s like much better than being a consulting team where you drone on about, “This is why something goes viral, that’s why something goes viral.” A lot of the focus that we’ve had is to figure out do we really know something, and then how do we build that into a platform and how do we build that into a process? If someone said to me, “Help me market a movie, or help me market a new product, but just do it as a consultant,” I would do a much worse job and there would not be as much that I could do as if I said, “Let’s do a partnership with BuzzFeed.”
Because when they do a partnership with BuzzFeed they get a big audience. They get reach, there’s technology, there’s different formats, there’s people that have lots of creative ideas that I wouldn’t have come up with that work at the company — and all those things together will result in a much better marketing campaign for an advertiser than if I was just peddling expertise. I think the expertise matters, but the expertise needs to find its way into a platform. And also having reach makes a huge difference.
FS: So the reach in that sense is necessary, but not sufficient?
JP: Yes. We could just put a bunch of ad networks on our site and lay off our business team, and we’d be a much less interesting business.
FS: Conversely, if I have a site which gets loads of traffic, but I can’t offer that deep understanding of the domain, then all I can do is sell a bunch of programmatic ads for three cents.
JP: Yes. It’s one of the reasons why there was some benefit to struggling to raise money in the early days, and trying lots of things, and experimenting in lots of different directions. We learned a lot during that period. I think it’s hard when you see certain companies that they launch and six months after they launch they get this crazy hockey stick and their service is huge. Then they’re running this huge thing and they’re not sure why it works or how it works. It’s hard to experiment when you have tens of millions of users.
FS: Was there a conscious decision made at any point at BuzzFeed that you would avoid the curiosity-gap thing?
JP: Well, at HuffPost I remember there was a very talented editor who now works at the Daily Mail who figured that out. You could show a picture of like an older guy at the beach and be like, “Guess whose body this is?” Then you click and it’s like, “Oh it’s Giorgio Armani” or whatever, and you could get a tremendous clickthrough rate on headlines that didn’t tell you what the story is about. The problem with that is that if you’re just getting clicks that would have gone to another headline on your front page, it’s sending people the content that might not be as good, because they’re clicking because they want to know what’s there. They’re not clicking because they’re interested in what’s there. If they knew that it was Giorgio Armani — if you just did a post saying, “Here’s a picture of Giorgio Armani on the beach” — people who care about that sort of thing would click and people who didn’t wouldn’t. You end up with lots of people who don’t actually want to see Giorgio Armani in a Speedo on the beach clicking that and then feeling like, “Oh god, why did I do that?” Like, “That was a waste of time.”
The main problem for us is that when you think from the perspective of the reader, if headlines are all devoid of information and you have to click them to find out what they are about, all the social streams out there would become much less useful and much less valuable. When you think from that perspective it’s like, “Whoa, let’s just make headlines that describe what’s in the article and that’s better for the consumer and it’s better for the ecosystem as a whole. Then let’s make articles that people really want to click because they’re interested in them, not because they’re wondering what it’s about.”
Then in some rare cases there’s something that say is impossible to describe and beautiful and weird. You might say, “This is beautiful and weird and you have to see it to understand it,” and in that scenario that’s not really anything else you can do.
FS: John Herrman [then of BuzzFeed, now of The Awl] wrote that piece about headlines.
JP: John wrote a piece saying the Internet provides many headlines for every article, sometimes better than what the author wrote, and that there just shouldn’t be headlines on articles. Headlines are generated on Twitter and all these different places and it is something that we think a lot about. What do people add to a story when they share it? In some cases it’s better than the headline that our team wrote and in some cases shows why content matters to them. Because you say, “I’m sharing this” and explain why you’re sharing it.
FS: What BuzzFeed doesn’t do, which I see a lot in the rest of the media world, is the secret-project-then-launch thing.
JP: Well, partly it’s that secret ideas rarely work. Coming up with some brilliant secret idea and then launching it and having it work the way we planned, that would be nice if that happened, but that essentially never happens anywhere. Maybe it happened with a string of Apple product releases that Steve Jobs was able to pull off—and was that a rare genius or was it luck or was it both? It’s hard to say, but it’s not a good model to operate a business on for most people.
FS: I think far too many people think it’s all about ideas.
JP: Sometimes you see people who say ideas don’t matter. Some VCs say it’s all about execution. There are other search engines and Google beat them and there were other social networks and Facebook beat them and it’s all about execution and ideas don’t matter. Sometimes you see people saying, “When did you come up with the idea for BuzzFeed,” and, like, “Wow how did you come up with that idea?” And I’m like, “I don’t even know what you’re talking about. I don’t know what idea you mean because there was a space that we were exploring and there’s a constellation of related ideas.” Being able to generate new ideas that help improve what you’re doing is really important for a company, but the execution is also super important.
FS: Nick Denton says he’s in a blood feud with you. There are definitely a lot of media types — and Nick would completely admit to being one of them — who like to define themselves against the competition and think of it as a race with winners and losers. How important is it to you to beat the competition? And who is the competition?
JP: It’s possible that it becomes more important as we get bigger, but for almost every startup the competition doesn’t really matter that much. The competition matters usually at the level of public companies or larger companies where they’re starting to bump up into the total size of the market. Then maybe the market’s growing at 6 percent or 10 percent a year and there’s competitors in there.
FS: You have 100 percent of the market in BuzzFeed ads.
JP: Well in BuzzFeed ads yes. There is an emerging space—it’s unclear how big the space is going to be if you think on the ad side. If you think the kind of advertising were doing, we have a good leadership position. It is less about competing against another company and it’s more about building a larger ecosystem and making something sustainable.
That could mean that it’s better for Vice and Vox to thrive because it starts to create a new guard of ambitious purely digital companies that will help transform the landscape. When they have conversations with brands and talk about how you can do big things in video and big things online and big things in social, that helps shift the budget into a larger market that we are in. That’s a good thing for us.
FS: My feeling about Vice is that they do a better job of selling how cool they are than they actually do of selling the brands who come to them.
JP: I don’t know what research they do, but I think they do take it pretty seriously. And I think that you can get to a certain level with smoke and mirrors, but if you don’t have something real I don’t think you get into the hundreds of millions of revenue.
FS: What’s their secret?
JP: They have a very strong and clear brand and even when the brand has moved and evolved it’s been very consistent at any given time.
They’re very good at making content for a particular consumer that is not as broad a mainstream consumer as BuzzFeed or more pop kind of site, but that there is certain people who are obsessed with what Vice does. I have met a young person who knew all their videos and were like, “It’s so badass and I can’t believe they did this.”
FS: Those people do exist.
JP: If you watch their stuff, it’s good quality. To me it’s, like, more motivating to see other people doing good stuff. If you’re in a space where everything sucks and you’re pointing to another competitor and saying, “Oh, look at how terrible they are,” then how does that motivate you to do good work? I think you look at Vice and you say wow. You can watch some of the documentary video that they’ve made and say they understand what they’re doing. They have a clear brand and they’re telling a good story.
“Life is tricky because it happens once and there’s no
opportunity for A/B testing.”
Back to Top
Felix: This is BuzzFeed 2014. You have this job that you love, which changes every year. What’s the big difference between this and where you were a year or two ago?
Jonah: One big difference is international, and another big difference is video. BuzzFeed UK is a decent size operation now. We are close to a billion video views in the first 14 months of BuzzFeed Video being done.
What Ze is building there has really been interesting and changing, so video is becoming a bigger and bigger part of the business. The ability to do lots of things at once has increased. It used to be that we could only really do one thing at once and now we’re doing many things in parallel. Managing that is challenging, but there’s so much more we’re able to do. The biggest concrete things are video and international. Then the mobile app—it’s just growing. It’s grown over 400 percent in the last year. It’s more and more consumption happening on mobile. Those are the biggest shifts.
FS: The revenue model, is that going to stay the same for the foreseeable?
JP: The basic revenue model is having an understanding and a technology platform for the distribution of social content. That model will be the same. We’ll be more on mobile, we’ll be more video, it’ll be more international, but the general model will be that brands become more like publishers and the best guide for them is a publisher—that really has a deep understanding of how this stuff works, and has a technology to support it; that if they want to go deep with someone to really take branding content to the next level, it should be us. That’s sort of the plan.
FS: I read your memo, which was a real rah-rah.
JP: Too rah-rah for you? Too American?
FS: It was too earnest. Don’t you think?
JP: Would you rather work for someone who’s cynical, or someone who earnestly believes in what they’re doing?
FS: I really like the stuff you would often read in the back of Wired magazine, or Spy magazine, or any time that Denton talks about himself, that very English sort of, “We don’t take ourselves too seriously. We laugh at each other and decide that we have no idea what we’re doing.” It’s a little bit of fun modesty, but also, there’s a humanity to it.
JP: I don’t think we take ourselves that seriously. I think the challenge is how do you be humble about what you know, because there’s nobody to learn from if you’re arrogant. And at the same time, be ambitious and say, “We’re going to try to do big things.” They’re kind of at odds, but I feel like the best way to operate in a fast-growing startup is to be able to do both. It is a tricky balance. I just think people having fun and people who are enjoying their work and people who are laughing about what they’re doing and feeling like you can make a mistake and it’s not like you get yelled at or you get fired, are all pretty important. And underappreciated by a lot of companies.
FS: You’ve got this very digitally-native, digital-only franchise. I’m sure that people have come up to you saying you should do a BuzzFeed TV program on cable TV or something.
FS: What that doesn’t really address is the bigger picture which is where, does digital fit into the broader media landscape? Are we still, for the foreseeable future, basically living in a television world where everything else is an afterthought, and you’re fighting for a share of an afterthought? Or are things converging?
JP: The Internet is certainly not an afterthought. If you include all the Googles and Facebooks and social platforms and tech companies—
FS: In terms of total audience hours or—
JP: I think in terms of revenue. Certainly in terms of market cap of the company. Facebook has a bigger market cap than the biggest media company. Google has a bigger market cap than Facebook, and if that’s a vote on the future of the industry I don’t think the web is going to be an afterthought. The question, though, I think you’re asking, more specifically, is content companies or people creating content, news, and entertainment on the Internet. Already in terms of scale of audience, BuzzFeed reaches more people each month than MTV or Comedy Central or a lot of the big cable networks. The amount of time people spend on television is still much higher. We reach a lot of people, but they don’t spend as much time on BuzzFeed as they spend watching reality shows on Bravo or something.
Depending on whether you measure by time or unique audience, we’re bigger by unique audience and they would be bigger by time, but I think we’re seeing a crop of significant businesses that are built on the web or primarily on the web. Whether it’s Vice or BuzzFeed or Vox Media or Hulu—
FS: Which is an interesting one—
JP: Which is a kind of hybrid kind of thing. Or Netflix. It’s another one where there’s starting to be companies that are in the media space that don’t own cable boxes and don’t own broadcast pipes and transmission towers or any of the infrastructure of the industrial media age, and that are purely or primarily in the digital space and are doing real revenue. I think you’ll see that trend continue.
It’s hard to predict the future, but it doesn’t feel to me like the space of web publishers or web content is going to just be the cheaper faster version with low rates and low margins—maybe a good business, but a small business. I think we’re going to get past that. At least that’s what we’re betting on.
FS: Vice is an interesting one. Did you see Shane Smith coming out and saying he could be worth more than Twitter?
FS: In a way they’re closest to you. They’re basically a digital marketing agency, in terms of where they get their revenues.
JP: Well, I think you’ve seen them under the same roof in a lot of other places too. MTV did a lot of this in the early days of MTV, where they would work with brands and do integrated stuff. Meredith Publications makes a lot of branded content and has whole kitchens and facilities to make great spreads for brands.
FS: Vice and BuzzFeed and Vox — it seems to me that you’re monetizing the paranoia of brands and of marketers who are like, “Oh shit, young people don’t watch TV they are on their phones all the time they are on the web the whole time—we don’t know how to reach them.”
JP: It’s not paranoid to think that the audience watching broadcast television is old. And it’s not paranoid to think people, particularly young people, are spending a lot of time on their phones and a lot of time on the Internet. It’s accurate to say that media consumption is changing in a pretty dramatic way and that if your marketing stays the same you essentially will be marketing to people who are consuming media the way people consumed media ten years ago instead of the way they’re consuming today.
FS: Is there anything you kick yourself about—that you feel like you had the opportunity to do something and you should have done it and you didn’t?
JP: I think in general it’s like life is tricky because it happens once and there’s no opportunity for A/B testing. It could be that you are living your best possible life and that if you re-play Felix Salmon’s life hundreds of other times, that this life you’re living is the best or among the top 5 percent of lives that you would have lived, and in lots of other ones you’d end up in an alley or in an unhappy relationship or with a job where you’re not intellectually fulfilled, and that you have found this amazing path.
It’s also possible that you’re not even in the top 50 percent of lives and that your life is really tragic and that despite all the wonderful and impressive and amazing things you’ve done, that you had the potential to do all these incredible other things that would have been either bigger in scale or more fulfilling or more modest and simple, but more pleasurable or whatever. That there were all these other paths that would be better. It’s, I think, hard to say whether there is something I missed that would have made things much better. In general, I’m pretty happy, and all these imagined alternate lives, I wouldn’t know how to even begin to speculate on how they’d compare.
This interview with Jonah Peretti was conducted by Fusion senior editor Felix Salmon. The piece was edited by Mike Benoist and Evan Hansen, fact-checked by Hilary Elkins, and copy-edited by Tim Heffernan. Illustrations by Young Professionals.
Follow us on Twitter
Like us on Facebook
Subscribe to our newsletter