EARLY IN HER tenure at Facebook, Sheryl Sandberg had a series of conversations with Chamath Palihapitiya. They knew each other well. He was a family friend, part of a regular poker game with her husband, Dave Goldberg. The two of them would often go on man trips to Las Vegas. Palihapitiya, himself a father, was great with Sandberg’s kids. So Sandberg was open to hearing what the thirty-one-year-old executive had to say.
Palihapitiya was at a crossroads at his time at Facebook. He had joined barely a year before, leaving a job as a venture capitalist. Previously, Palihapitiya had been a VP at AOL, the youngest person to hold that position at the company.
That job had been one more step in a heroic rise for Palihapitiya, who had emigrated from his native Sri Lanka at six years old, when his civil-servant father got a post in Canada. When his father lost the job a few years later, the family went through hard times. As Palihapitiya told The New York Times in 2017, they moved into a 400-square-foot apartment above a laundromat. His father had trouble finding a job, and “there was drinking and stuff,” Palihapitiya recalled. Though his mother was trained as a nurse, she worked as a housekeeper and nurse’s aide.
Palihapitiya worked, too, first at a Burger King. He found it more profitable to run a blackjack game in his high school lunchroom, though, raking in $40 or $50 during some breaks. He’d then take that money to a casino—he had a fake ID that gained him entry—and try to augment his winnings. He had a knack for high-stakes poker, and eventually would compete in the World Series of Poker.
After graduating from the University of Waterloo with an electrical engineering degree in 1999, he worked for an investment bank for a year. He later called his job trading derivatives “the most insipid, idiotic use of my time.” (No one in Silicon Valley could match Chamath Palihapitiya when it came to erudite invective.) He quit the bank and began applying for jobs in the online sector, ultimately moving to California to work for a music start-up based on the WinAmp music player that had inspired Zuckerberg and D’Angelo back in high school. Though WinAmp was seen as a cool application, Palihapitiya came to view it as a platform; outsiders were free to customize it with decorative “skins” and creative plug-ins that boosted its power. (Synapse was an example.) The platform concept enabled talented engineers and designers to improve their products at no cost, making them more valuable and harder for competitors to beat them. WinAmp was where Palihapitiya met Sean Parker, and naturally the two got along famously.
AOL bought WinAmp in 1999 and Palihapitiya joined what was then the world’s most popular online enterprise, a year before its disastrous purchase of Time Warner. And now, in 2005, Parker was calling, saying, “I’ve just become president of a company called Thefacebook, and I want you to meet me and learn about it and meet the founder.”
Parker and Zuckerberg flew to Dulles and met Palihapitiya and another AOL executive, Jim Bankoff. Parker did most of the talking, as Zuckerberg was in his silent mode for most of the meeting. But Palihapitiya was impressed. He later told Bankoff that AOL should consider buying Thefacebook, but AOL was still sifting through the rubble of its train-wreck merger with Time Warner and couldn’t consider an offer. Not that Zuckerberg would have gone for it.
AOL wound up doing a small deal with Thefacebook that linked AIM to the start-up’s website, so that Facebook friends could find one another on the chat service. Palihapitiya often referred to it as solely to AOL’s advantage.
But the biggest outcome of the deal was the connection between Palihapitiya and Zuckerberg. In 2005, Palihapitiya quit AOL—one takeaway from his experience there, he later said, was that “most people at most companies are really shit”—and joined the venture-capital firm Mayfield, in Menlo Park. Every couple of months or so, he and Zuckerberg would get together. Zuckerberg’s combination of chutzpah and shyness appealed to Palihapitiya.
Not that Palihapitiya necessarily agreed with the lionization treatment that the young CEO was enjoying in the press and in Silicon Valley’s inner circles—he never subscribed to the pervasive meme in the tech world and the business magazines that successful founders were like gods in the Pantheon. To him, they were more like opportunistic beneficiaries of favorable economic and social conditions. Maybe if Zuckerberg had gone to Ohio State instead of Harvard, he thought, none of this would have happened. (The chip on Palihapitiya’s shoulder that came from not attending Stanford or an Ivy seemed big enough to leave a permanent depression on his clavicle.)
Still, the boisterous Palihapitiya and the more introspective Zuckerberg had similar views about business and tech. Inevitably the idea arose that Palihapitiya might join Facebook. Nothing formal. But Zuckerberg, in the throes of post-Yahoo! dissatisfaction with Van Natta, wanted someone in his top management rule who was more an ally.
So in early 2007 Palihapitiya went to the University Avenue office—just to check things out, mind you. Zuckerberg wasn’t there and he wound up talking to Moskovitz, who asked him why he wanted to work at Facebook. Being regarded as a supplicant offended Palihapitiya. “Hold on a fucking second,” he said. “I’m not interviewing with you.” Then Palihapitiya proceeded to tell Moskovitz everything that was wrong with Facebook. The list was long, but Palihapitiya wasn’t implying that Moskovitz and Zuckerberg were idiots—they were doing the best they could, considering that they had zero experience in business.
But what did impress Palihapitiya—and make him amenable to actually taking a job there—was that Zuckerberg and his coterie seemed open to listening to suggestions that might fix things. To Palihapitiya, it was an opportunity worth taking. Even so, Palihapitiya drew the process out; at one point during his Hamlet-esque indecisiveness, Zuckerberg even called in Roger McNamee to tout the company. “I think [Chamath] already decided he was going to do it, and he was just playing with me,” says McNamee.
When Palihapitiya finally joined the company, he was given considerable, if amorphous, duties under the umbrella of “product marketing and operations.” These were basically responsibilities yanked away from Van Natta, who, after his lack of loyalty during the Yahoo! episode, was essentially demoted from chief operating officer. (Facebook temporarily retired the title.)
He had an immediate impact. Director of monetization Tim Kendall was impressed with Palihapitiya’s talents—and appalled by his demeanor. “He was an incredibly shrewd leader,” he says. “I learned quite a bit from him. And I would never want to work for him again.”
Palihapitiya thought that the haphazard hiring process over the past two years had brought dead weight into the company, and he introduced a “forced ranking” process to identify and dismiss the laggards. While that made sense, his public ranting about people not pulling their weight was scary even to some high performers. Most of the older, more experienced people who came into Facebook had learned to adjust to the callow culture. Palihapitiya doggedly remained himself—embracing his own worldliness, his tough upbringing, his contempt for the conformity of the striving-for-Harvard crowd.
And Palihapitiya could be a bully. He would humiliate people at meetings, criticizing their appearance. He mocked one barely middle-aged executive for his receding hairline. Another former executive, who would only talk about Palihapitiya if I turned off my tape recorder, was almost in tears as he recounted the verbal abuse he took from Palihapitiya. It was as if he was still afraid that Palihapitiya might emerge from behind the bushes and resume his battering.
When I asked Palihapitiya about this, his reaction was unapologetic. “Get. Out. Of. My. Way,” he says. “They must feel really bad as they lounge around in their chinchilla blankets and their multimillion-dollar mansions right now.” He explains that the workplace is not a family, and if people were looking for touchy-feely responses from him, “that was not going to be a good meeting, those people probably felt intellectually bullied.”
On the other hand, he could be inspiring. He would stand on a table and tell people how big Facebook would be. At a time when people would be buzzing at the prospect of Facebook being valued at $10 billion, he would proclaim that Facebook would be worth ten times that. (Later, when that no longer sounded crazy, he would talk of it being worth a trillion dollars.) He also had no compunctions about dismissing people he felt were not sufficiently fast-moving and innovative to help Facebook meet those marks. Some of those people had been regarded by coworkers as subpar but lingered in their jobs because Zuckerberg could be timid about firing people.
Yet, for all his talent and bluster, Palihapitiya in his first few months was not a smashing success at Facebook. He had not really added much to the Platform team. What he was most associated with was Beacon, one of the company’s most traumatic moments. He felt that his unfortunate quote in The New York Times—the one about Beacon not transferring data against a user’s wishes—was misinterpreted. (“I learned not to get into technical details with reporters,” he now says.) Nonetheless, it eroded what was already a fragile relationship with Zuckerberg.
By the end of 2007, even Palihapitiya acknowledged he had done a miserable job. Maybe he was one of those people whose success was simply due to lucky circumstances. “If I were you, I’d fire me,” he told Zuckerberg in a post-Beacon meeting. They agreed that he should figure out a more focused task for himself and take one last shot at things.
NOW, IN SANDBERG’S office, Palihapitiya was about to throw a Hail Mary: he had something in mind that he felt would be instrumental in Facebook’s future success. He didn’t anticipate that the same effort would cause its failures as well.
At that moment—early 2008—Facebook’s growth had slowed. A similar trough had hit Facebook more than a year earlier, before Open Reg and the News Feed kicked in, but in a way, this was more alarming, because there were no similar groundbreaking products on the horizon. And no one had any idea about the cause. “Everything stopped,” says one executive. “And to this day we don’t know why.”
“Growth had plateaued around ninety million people,” Zuckerberg recalls. “I remember people saying it’s not clear if it was ever going to get past a hundred million at that time. We basically hit a wall and we needed to focus on that.”
Palihapitiya came offering a solution: a high-energy team with a long leash whose focus would be accumulating and keeping users. He felt he had identified the North Star of Facebook, the most elemental aspect of how it defined its business and its financial health. And that was the concept of the Monthly Active User. Other Internet businesses counted how many people were on the site each day, or how many had signed up in total. But monthly was a better indicator, because someone consistently on the service for a full month was likely there to stay. Thus the number took into account the “churn”—how many people were leaving Facebook. Palihapitiya proposed to be utterly obsessive about MAUs—to look at every part of Facebook’s business in light of this metric, to learn what can drive MAUs, to fix things that don’t increase it, and to build new parts of the company to boost MAUs even higher.
Sandberg was intrigued. “What do you call that?” she asked.
“I don’t know,” said Palihapitiya. “The job would be to grow things—all to grow MAUs.”
“Maybe you should just call it growth,” she said. (Sandberg says she doesn’t remember that exact meeting—she had a lot of them with Palihapitiya—but confirms that something like that exchange occurred.)
Palihapitiya honed his ideas and presented them at the next board meeting. Palihapitiya claimed he could double or triple the user base by using aggressive techniques, and go far beyond that by making Facebook’s whole Platform itself an engine for growth.
The board wasn’t terribly excited. Just give me some time to execute, said Palihapitiya. If they didn’t like what he was doing after a few quarters, then maybe they could all agree on his exit.
Palihapitiya was doubling down on a mania for growth that had begun well before he arrived. Even as Sandberg gave Palihapitiya the green light to lead his team, she felt that Facebook was already a focused growth machine. Facebook’s origin story was a fable of growth: in 2004, in the space of a month—February, the shortest month of the year!—Zuckerberg expanded Thefacebook from Harvard to other campuses. In 2005, an early employee named Noah Kagan (who later got fired for leaking product plans to the trade publication TechCrunch) once suggested to Zuckerberg that Facebook sell tickets to events listed on the site’s pages. Zuckerberg’s answer was to walk over to one of the omnipresent whiteboards and scrawl one word with the marker: Growth. “He said that if any feature didn’t do that, he was not interested,” wrote Kagan. “That was the only priority that mattered.”
Indeed, even in 2005, the company decided to hire specialists who would dive into the information Facebook gathered to draw more users. That was to be the main role filled by Dan Plummer, the scientist who died in a bicycle accident in January 2006. His replacement came to Facebook by a chance encounter involving Zuckerberg’s sister Randi, whom he’d asked to come work for the company. At her going-away party in New York, Zuckerberg recognized an engineer named Jeff Hammerbacher as someone he’d been in a seminar with at Harvard. (Hammerbacher’s girlfriend had been pals with Randi and had dragged him along.) On Zuckerberg’s suggestion, he applied for work at Facebook. His plan was to establish California residency for a year and go to grad school there paying in-state tuition. But he was impressed with the Facebook engineers who interviewed him. And he was intrigued when he saw that Adam D’Angelo’s business card read “data mining.” Hammerbacher was all about that. And so was this little Palo Alto start-up, even in 2005.
Hammerbacher, usually collaborating with others, analyzed behavior on Facebook, with an eye to see if the data they unearthed could help growth. More important, he built a system where Facebook would collect huge amounts of information and analyze it to come up with conclusions to help Facebook work better. He set up a system to log every single click made by Facebook’s users. He would later refer to it as an “Information Platform.” On the first day, he had 400 gigabytes of information.
Working with Naomi Gleit and Matt Cohler, he looked at schools where growth took off and compared them to ones where growth had stagnated, seeking to understand the ingredients for success and failure. When Facebook unveiled Open Reg, he drilled down on the data to find how users came to Facebook, and discovered that the biggest source came from a program an engineer had built that imported people’s contacts from Microsoft’s Hotmail.
The program was called “Find Friends,” written by Facebook engineer Jared Morgenstern. It worked with Hotmail, Gmail, and Yahoo! mail. Users would provide their login and password, and Facebook would scan all the people who were contacts or connections and import those into its database. Those who matched with people already on Facebook would be sent friend invitations. Others would be presented to the user and a simple check alongside their names would trigger an email inviting them to join Facebook. Upon completion, Facebook would delete the login information.
Microsoft had objected to this. “They were blatantly stealing, trying to build their social network on the backs of others,” says a Microsoft executive involved in the discussions. Zuckerberg shrugged it off when confronted with the charge. “He was like, Yeah, I know it’s kind of annoying, if it bothers you we’ll stop,” says the executive. “But he didn’t stop.” In Microsoft’s view, it not only violated their terms of service but was arguably an unethical data grab. Just because you were someone’s contact on Hotmail, that didn’t mean that you would be okay with winding up in a Facebook database. Find Friends created a tension with Microsoft that was resolved only on the eve of the larger company’s investment in Facebook in 2007.
Such techniques were as vital as oxygen to Facebook, because people often did not bother to fill out their social graph. “A huge percent of Facebook users barely even sends friend requests,” Zuckerberg told me in 2011. “They won’t even set up the most basic relationship. They only accept incoming connections.”
To keep that growth going, Facebook needed to scrape not only Hotmail but numerous other services. The process had to be done separately for each email provider, a time-consuming project that could not possibly be executed by the single engineer Facebook had assigned to the task. An early Facebooker named Jed Stremel, who had been an ace dealmaker at Yahoo!, took care of the problem. He discovered that the global wizards of contact scraping were found in a two-person company in Malaysia called Octazen. Stremel quickly made a deal for them to write their little data gobblers for Facebook. He recalls that he paid about $400 for the deal. “It was in keeping with the spirit of ‘Move Fast and Break Things’—just getting something done quickly is what mattered,” he recalls. (Facebook would buy Octazen in 2010.)
Some early Facebookers contend that email scraping was the single most valuable factor in helping Facebook grow when for the first time it found itself competing not in closed networks but in the world at large.
By the time that Palihapitiya moved to Growth, Hammerbacher had grown doubtful about the mission he was fulfilling at Facebook. He left the company in September 2008. “It was turning from a place to explore to a place to exploit,” he once told an interviewer. He became a co-founder of Cloudera, a company that stored data in the Internet cloud, and later became involved in trying to solve cancer with data analysis. Though he felt no animosity toward his former employer, at times he has expressed its motivations in phrases that speak volumes. In 2011, assessing the jobs of data scientists at Facebook and its peers, he made a remark to a BusinessWeek journalist that would reverberate for years: “The best minds of my generation are thinking about how to make people click ads,” he said. “That sucks.”
IN THE OLD television show Mission: Impossible, each episode began when the leader, Jim, flipped through the dossiers of spies, strongmen, and honeypots, putting rejects in one pile and tossing the photos of the talented ones who were perfect for the mission into a stack on his coffee table. Palihapitiya did something similar in handpicking his team, plucking his diamonds from the ranks of other groups as well as outside prospects. He had a great eye for talent. Among them were Naomi Gleit, one of the first few Facebook employees; Javier Olivan, an engineer from Spain; Alex Schultz, a UK-born marketer; Danny Ferrante, a data wizard; and Blake Ross, a star hacker who had been the co-creator of the open source Firefox browser. They were a diverse bunch. A majority had either been born overseas or at least had a parent from outside the United States. Two were gay. One of the leaders, Gleit, was a woman. They were a team of outsiders, a data-driven Dirty Dozen, armed with spreadsheets instead of combat rifles. The choices proved brilliant, especially Oliver, Gleit, and Schultz; more than a decade later, that threesome would still be at Facebook in the most powerful leadership cadre, “the small group.”
“Chamath had the rare ability to do the crazy-speech thing and get everybody fired [up] but those were like the three musketeers,” Mark Slee, an early Facebook engineer, says of the trio. “At the end of the day it was the three of them working with a boatload of engineers and product managers to do a lot of really unglamorous work. Everybody wanted to work on the cool product stuff, but they were doing the stuff that actually gets more people on Facebook.”
As one of the first employees of Facebook, Gleit had a quiet moral authority. In a sense she had helped own Facebook’s growth since the start, working, as everyone in the company did, to map Mark Zuckerberg’s bottomless ambitions onto products and initiative. “We always had growth projects,” she says, noting that from the very start Zuckerberg began spreading Facebook to other colleges. When Gleit joined in 2005, Facebook was moving to high schools and already preparing for Open Reg. Some people on the team would say that it was she as much as Palihapitiya who provided the vision for this new initiative.
Palihapitiya wanted Growth to be a power center in the company with special status and a distinctive subculture. To distance it from any other group at Facebook, its members would not refer to themselves as a “team,” but a “circle.” The Growth Circle. They had their own little grotto in one of the Palo Alto buildings, and in 2009 when Facebook moved from downtown Palo Alto to a new headquarters, Palihapitiya claimed space apart from everyone else, in a dark corner of the ground floor.
“Even the name reflected how special it was,” says Gleit. “We were a growth circle, we were tight-knit, we were the A-plus team. It was just magical, almost.”
It also could be brutal. Palihapitiya saw himself as more worldly and pragmatic than the idealistic and largely homogenous workers at Facebook. He had a sort of contempt for the charmed lives they led. The way that he saw it, the Facebook hiring process rewarded the kind of people who spent their lives making checkmarks on life-achievement boxes—astronomical test scores, Harvard or Stanford, enviable internships. You could see it in their Facebook feeds, as each post seemed to unlock one more level on a video game. But the problem with those box-checkers was that they were locked into following predetermined courses.
The idea of going off course, abandoning the neatly arranged checkboxes entirely, didn’t fit their mental model. They couldn’t handle it. So Palihapitiya undertook to mess with their minds, doing psychological tricks to rewire their brains so they wouldn’t be so fixated on validation and affirmation. Calling his team a circle was part of that—an attempt to jar them out of the idea of a hierarchy where the most important person sat at the head of the table. The idea was to empower anyone on the team.
While the circle was supposed to equalize the power dynamic in the team, Palihapitiya’s personality still dominated. His style was geared to provoke reaction. Every third or fourth word out of his mouth seemed to be an expletive, usually some variation of the synonym for fornication. When he felt someone did or said something stupid, a reliably frequent occurrence, he would unleash a fusillade of insults. Such assaults weren’t the kind of thing that you would take to HR, though. At Facebook that behavior was regarded as Chamath being Chamath.
And what happened in Growth Circle stayed in Growth Circle.
The core members of Growth still speak highly—reverently—about Palihapitiya. But even within this group, the feeling could be mixed. “I didn’t love working for him, and he knows that,” says Alex Schultz. “But he was generally right. And so I really thought as a leader he was pointing us in the right direction.”
Ultimately, Facebookers shrugged off Palihapitiya’s behavior because he clearly had the company’s interests at heart. “I had a good relationship with Chamath,” says Chris Kelly, whose job included reining in the Growth Circle’s excesses. “We would have conversations about where the boundaries were, usually ahead of time. Transgression wasn’t their goal, growth was their goal. Maybe there was a little bit too much comfort with transgression, but growth was the goal.”
EARLY ON, GROWTH identified a lot of low-hanging fruit to help Facebook’s numbers. One of them was search engine optimization (SEO), the practice of raising the visibility of content in Google search rankings. If a Facebook virgin using Google came across a friend’s profile, he or she might be inclined to join the service. But Facebook’s SEO was so weak that when you typed the word “Facebook” into Google, you had to scroll far down the rankings until you saw the actual site.
In the previous year (2007) Facebook for the first time allowed its users’ profiles—or an abbreviated version of them—to appear in search results. But they weren’t appearing high in rankings, in part because they weren’t easily found within Facebook, and Google’s web crawlers would have to burrow deep into Facebook to find them. Schultz and Gleit put together a directory for Facebook that interlinked people’s profiles in a manner that was catnip for Google. It resulted in the profiles being ranked higher, and when people came across them, they could ask to friend those people right there in the Google search engine. It got Facebook some new users.
“It wasn’t like doubling [new users] but was it five percent, maybe ten,” says Schultz. “But the thing is, each of those things you try, even if it gets one percent, and you get one percent and one percent and one percent [from other tricks], it really adds up. Doing a lot better in terms of growing as a company. And so we were willing to fight for each of those individual wins.”
“We were very open about trying new things,” says Palihapitiya.
But the masterpiece of Growth—its Mona Lisa, its “Like a Rolling Stone,” its Godfather 1 and 2—is a feature that became almost as much a part of News Feed as weddings, vacations, and political outrage. It’s called People You May Know, referred to internally by the acronym PYMK. Officially launched in August 2008, People You May Know is a feature that identifies personally selected prospects for one’s friend list. PYMK proved to be one of Growth Circle’s most effective tools, and also one of its most controversial ones, a symbol of how the dark art of growth hacking can lead to unexpected consequences.
It wasn’t a Facebook invention—LinkedIn, a growth-crazy company in its own right, did it first. (Reid Hoffman would later put a ribbon on the growth-at-all-costs phenomenon and dub it “blitz-scaling.”) But Facebook took the idea of presenting current members to new and current users to dizzying heights.
On its face PYMK seems innocuous enough: a carousel of profile pictures on Facebook presumably connected to you, but somehow not your Facebook friends. Its impetus was to address an imperative that the Growth team’s researchers had unearthed: a new Facebook user is likely to abandon the service if he or she doesn’t connect with seven new friends—fast. For someone without a core list of friends, Facebook is like playing soccer solo.
Facebook did have some tricks to use for newcomers who had yet to connect with friends. “At one point we created these fake stories called fluff stories,” says Palihapitiya. “We were like, Why don’t we create stories about people’s birthdays, or suck in an interesting article or whatever? It’s not like [friends] generated a story I’m seeing, it’s system-generated. The point is, it’s fluff.” The idea was to stave off entropy until people made enough friends to have real stories.
But that’s no substitute for actually having Facebook friends. The Growth team’s data scientists did a study that emphasized how critical it was for Facebook to find friends for newbies—especially active ones. As the study called “Feed Me: Motivating Contribution in Social Network Sites” stated,
It is vital for developers of social networking sites to encourage users to contribute content, as each individual’s experience is dependent on the contributions of that person’s particular set of connections. It is particularly important, if rather difficult, to encourage continuing contributions from newcomers.
Thus PYMK was essential for Facebook. Exposing potential friends is a way to improve a member’s experience; it increases the chances they will share more, and, most of all, it makes people less likely to bail on Facebook.
For many people, PYMK is a welcome feature: a helpful prompt to get in touch with connections who would help them get value from their Facebook experience. But sometimes PYMK can be unsettling, raising questions of what caused those cameo appearances on your News Feed by people whose connection to you was obscure, and sometimes downright unwelcome. A sex worker found Facebook recommending her clients, who did not know her true identity. A sperm donor got a suggestion for the biological child he never met. A psychiatrist learned that Facebook was recommending that some of her patients friend each other on the service. And millions of people went Ew! as Facebook suggested they develop relationships with friends of their children, spouses of their casual acquaintances, or disastrous blind dates of a decade ago.
Journalists who studied the feature—notably Gizmodo’s Kashmir Hill, who spent part of a year trying to get to the bottom of the mystery—were never able to get Facebook to divulge exactly how the product works. Hill was the one who unearthed the story of the woman who got a Facebook suggestion that she friend her long-absent father’s mistress. And Hill herself was stunned to find that someone on her own PYMK suggestions turned out to be a great-aunt she’d never met. Facebook did not provide her the information she requested on how it made this connection.
Later, Hill would also write about the psychiatrist who discovered that PYMK was suggesting that her patients make friend connections with each other—even though the psychiatrist did not friend her patients on Facebook. Hill conjectured that it might have something to do with the fact that the psychiatrist at one point shared her phone number with Facebook, and the company scraped her contacts, as well as the contacts of her patients. Once again, Facebook would not provide an explanation.
Neither would Facebook respond to Hill’s queries about whether PYMK’s instant suggestions for new users means that it was storing data on people not signed up on Facebook, and making use of “shadow profiles” when someone joins. Years later, Mark Zuckerberg would testify in Congress that the company does not engage in that practice. It does keep some information on nonusers, he said, but only for security purposes, to fight fake accounts. (Zuckerberg did not mention his early cogitations in the Book of Change about dark profiles.) In a more elaborate explanation provided later, Facebook said, “We do not create profiles for non-Facebook users,” though it also says it keeps certain data, like what device and operating system version a nonuser has, for things like “optimizing registration flow for the specific device” should someone decide to join.
But Palihapitiya now indicates that dark profiles did exist, and the Growth team took advantage of them. He says that Facebook would take out search ads on Google using the names of Facebook holdouts as keywords. The ads would link, he says, to those dark profiles of nonusers that supposedly do not exist. “You would search for your own name on the Internet and you’d land on a dark profile on Facebook,” he says. “And then you’d be like well, fuck it, you’d fill it in and then PYMK would kick in and we would show you a bunch of your friends.”
Some of the mysteries of PYMK were addressed in a 2010 talk by Facebook data scientist and engineer Lars Backstrom. Reporting that the feature “accounts for a significant chunk of all friending on Facebook,” Backstrom went through the technical process of how Facebook chooses its suggestions. The most important hunting ground is the “friends of friends” region, according to the presentation. But that is a very large set.
Users have an average of 130 friends, he said, each of whom has their own 130 friends. (This is close to the so-called Dunbar number, named after the sociologist Robin Dunbar, who discovered that most people can reasonably maintain relationships with no more than 150 people.) So the typical user has 40,000 friends of friends (FoFs), and a power user with thousands of friends might have 800,000 FoFs. That’s where the other data comes in—to find signals like the number or closeness of mutual friends and mutual interests, along with “cheaply available data” to identify which ones are likely to cause someone to click when spotted in a PYMK list. As the data gets refined, Facebook uses machine learning to make the final suggestions.
Backstrom also revealed that one’s behavior on PYMK helped determine which suggestions Facebook would offer—and how often it would show you the list. Once Facebook determined you fell for the feature, it would keep coming back, stuffing your friend list with weak ties.
The Backstrom presentation omits any specific information about what data sources besides FoF analysis Facebook uses in the feature. To be sure, those sources have evolved steadily since Facebook introduced PYMK in 2008. It’s almost certain that Facebook watches your email and sees whom you are contacting. Probably your calendar as well, to see whom you’re meeting with. Other sources have indicated that if someone glanced at your profile, that act might increase the odds that the person might appear on your PYMK list. It’s doubtful that simply thinking of someone is enough to put that person on your PYMK lineup. It just seems that way.
Alex Schultz says that a lot of guesses about what data goes into PYMK are conspiracy theories. He says that people often misremember that they permitted Facebook to use their contact lists or email (maybe they got insufficient notice they had given permission?). Anyway, he says, the biggest reason for someone showing up is that they are friends of friends deemed likely to know you.
Or, as Cameron Marlow, who once led Facebook’s Data Science team, puts it, “The goal is to try and find relationships that you have that already exist on Facebook, but that you were not aware of.”
As troubling as PYMK is, the scary thing is that it could have been worse. Facebook’s chief of privacy, Chris Kelly, says that he blocked the use of some questionable techniques that the Growth team had suggested. “There had to be some rules,” he says, declining to share the ideas he snuffed.
Other problems with PYMK are subtle but no less troubling. (Warning: we’re going down a rabbit hole here.) The early Facebook executive Dave Morin came to view PYMK as an insidious means of boosting retention numbers at the expense of a good user experience. Since a key goal of PYMK was to boost the value of Facebook for new users—making sure that they had enough friends to fill up their News Feed—the suggestions were tilted to help those newbies more than the people they friended. Particularly valuable to Facebook would be suggestions of users who posted promiscuously, because (as the “Feed Me” study proved) early exposure to super-active users will influence newcomers to share more throughout their Facebook life.
As Morin puts it, “When Facebook shows you people you should connect to, it can make a choice as to how that algorithm works. It can either show you people you’ll become closer to and who will make you happier if you add them to your world. Or it can show you people that are advantageous for Facebook, the system, to show you, because it increases Facebook’s value and wealth and it makes my system better.” He says that Facebook takes the latter course, benefiting itself at the expense of its users.
This might give the experienced user a worse experience. The News Feed is zero-sum—people view only a limited number of stories. Facebook would prioritize stories from your newer, weaker ties that it wanted to keep on the service. And you would see fewer things from people you did care about. “The system knew that if I said yes to you, you would become more engaged,” says Morin. “You’d be effectively stalking me because I’m like a person distant in your social graph who you want to know. It’s almost like watching a tabloid.” Morin says this semi-stalking factor “became the primary variable in PYMK.”
Some people pushed back on Palihapitiya on this issue, arguing that such behavior was not Facebook-ish. Palihapitiya said since the ultimate goal is getting everyone on Facebook, it doesn’t matter in the long run. Though he said it more colorfully than that. “He was basically like, Go fuck yourself, and he’d walk out of the meeting.” says Morin.
Eventually Morin would quit Facebook and start his own social network, called Path. The idea behind it was that it would limit one’s social network to only meaningful connections; following the dictates of Dunbar, one could have 150 friends and no more. (He later increased the number.) Despite winning plaudits from critics, Path ultimately failed, unable to compete with Facebook.
Zuckerberg defends PYMK, and the way he does it illuminates his thought process and product acumen. When I bring up the above conundrum to him, he gets very serious. “This gets to a really deep philosophical thing about how we run the product,” he says. He concedes that if users take the hint from PYMK and friend their weak ties, their experience might be somewhat degraded. But there is a more important issue at stake, he argues—the health of the network in general. “We don’t view your experience with the product as a single-player game,” he says. Yes, in the short run, some users might benefit more than others from PYMK friending. But, he contends, all users will benefit if everyone they know winds up on Facebook. We should think of PYMK as kind of a “community tax policy,” he says. Or a redistribution of wealth. “If you’re ramped up and having a good life, then you’re going to pay a little bit more in order to make sure that everyone else in the community can get ramped up. I actually think that that approach to building a community is part of why [we have] succeeded and is modeled in a lot of aspects of our society.”
Furthermore, Zuckerberg believes that by friending your weak ties—which includes people you hardly know—you become closer to them. Facebook might even violate the physics of social interaction by stretching the number of meaningful contacts that people can handle. “There’s this famous Dunbar’s number—humans have the capacity to maintain empathetic relationships with about 150 people,” he says. “I think Facebook extends that.”
In a social-science sense, that would be like surpassing the speed of light. But if anyone could do it, it would be Facebook’s Growth team.
SOON INTO THE Growth effort, Palihapitiya met with Sandberg and told her he was going to focus on the most fertile territory for growth: the international market. While Facebook had captured a significant portion of the North American market, it had yet to make a huge impact on the rest of the world. He wanted to expand the team to aid the push to other countries. She suggested he look at what other companies, like Yahoo! and eBay, did when they drew users internationally. Palihapitiya came back and told her what he wouldn’t do—he wasn’t going to hire white Ivy League grads. He wanted street fighters who knew the local language.
He tapped Javier Olivan to lead the effort. Olivan was a Spanish engineer who’d arrived in the United States in 2005 for a Stanford MBA. His dream was to start his own company. He was struck immediately with the way that Facebook had captivated campus life, and decided, basically, to clone the idea. Working with friends, he started a college social network in Spain while he was still at Stanford. One day Zuckerberg himself spoke to one of his classes, and Olivan chatted with him after the session about expanding to other countries. Soon after, he was at Facebook, on Palihapitiya’s team. “The assignment I had was international growth,” he said. “The most obvious thing to be done was to get the site in as many languages as quick as possible.”
Facebook had already embarked on the process some months earlier. A small team working on internationalization had created an app called Translate Facebook. It enabled people in various countries to translate words from the original English into their home language. Of course, Facebook wasn’t waiting around for people to volunteer. It checked its logs to identify people who were using the English version of Facebook in some foreign country. Those chosen—by algorithm of course—would see a message on top of their News Feed asking if they’d like to help to translate Facebook. These unpaid helpers (another benefit of crowdsourcing) would do a first draft, setting up a scaffolding that identified the pitfalls of moving Facebook to the individual language.
After that first step, Facebook might open up the process of translating the terminology to everyone. Sometimes it would prompt people for hints: a native speaker might get a pop-up saying, “Hey do you speak this language, can you help us with these ads?” Sometimes the crowdsourcing would be used to refine a machine translation.
The hurdle was getting good translations. In order to be put into production, a translated version of Facebook needed verification that the translation was accurate and nuanced, not to mention avoiding any ugly American faux pas. And often even translating simple words could be difficult in certain languages. Some languages couldn’t easily make the metaphorical transition of a “wall” from a structural element in a building to a virtual bulletin board on a profile page. And forget about getting an easy translation to “pokes,” which even English speakers had difficulty grasping. One engineer on the team who had been a huge fan of the Reddit website suggested that Facebook adopt the Up and Down buttons that Redditors used to sort content.
The crowdsourcing approach ran counter to accepted wisdom. The tried-and-true method was known as the 80/20 rule, where resources were concentrated, and professional help sought, on the key languages known as FIGSCJK—French, Italian, German, Spanish, Chinese, Japanese, and Korean. That would get you the vast bulk of the online audience. “As a company we’ve always had the mission of connecting the world,” says Olivan. “And the 80/20 rule is not going to cut it. We really wanted to make sure it was available for everyone.” To do that Facebook needed to improve its crowdsourcing tools so those professionals weren’t needed. The idea was to have Facebook everywhere. Even in many languages that no one at Facebook was familiar with.
Crowdsourcing would only go so far, though, and Facebook had to accept that it needed professional translators for what were deemed the most important languages, which they calculated by a country’s gross domestic product. Someone from the Internationalization team would set up a meeting with native speakers who taught in English in a given country, and then spend days verifying the translations of certain words used on the site for a given language. Still, the priority was shipping rather than perfection. “We had lists of the words that were used most often on the site and we would be like, Well, let’s just only ask them about those,” says Kate Losse, an early employee on the team. “And then there’s hundreds of words that only appear once and we don’t care if that one’s wrong, because you just have to triage. Essentially it was proofing that the translations that were coming into the app were going to be good enough to launch.”
There were limits to this low-cost, high-velocity approach. When Facebook officially announced its entry into Japan, for instance, it hit a wall of criticism that seemed to stun the high-level executives who traveled to Tokyo for the event. Zuckerberg himself was there, on the way back home from his 2008 personal journey to India.
Facebook Japan’s announcement uncovered a culture clash that would play out in many countries. On one hand, Facebook was proud of its process. Over three weeks, 1,500 “amateur and professional translators” took a stab at translating the English text on Facebook—pages, dialogue boxes, and messages—into Japanese. Then they debated and revised. But local sentiment was mixed. Some Japanese users found it offensive that Facebook thought a superficial translation, and not a thoughtful product redesign, could make its American product meaningful to the unique Japanese audience. “It did not set up a Japan office or select a local representative to lead the effort,” complained the Japan Times. “In fact there isn’t a single unique feature about Facebook Japan.”
When the Japan Times asked Javier Olivan, who attended the launch, about this, he confirmed their fears that the company was exporting American values along with their product. “Facebook is exactly the same all over the world,” he said. (Facebook Japan would eventually vanquish its competitors, reaching three-quarters of the social-media market in Japan by 2016. By mid-2019, however, that share would plummet by two-thirds.)
Under Olivan’s direction, the process worked astonishingly well. When it deemed the translation ready, the Internationalization team would push it in a given country and watch the usership explode. Then they would plant another flag in a corner of their office where they had a small cluster that marked the national symbols of the countries where Facebook spoke their language. In a few months Facebook went from no languages to hundreds of languages.
In 2012, Facebook hired a woman named Iris Orriss, who helped further refine the process. She had begun her career in software testing, then moved to engineering teams, and realized that her most satisfying tasks were working to extend technology to other countries. She was passionate about language and cultures and globalization. It was almost a calling for her. When Facebook recruited her from a job at Microsoft, scaling speech software to other languages, she researched the company. She concluded that something about internationalization at Facebook was different from the previous places she worked. There was a mission driving the entire enterprise.
And the key to it was growth. Other companies view international expansion as an operational exercise. “If it’s operations, it’s more a cost-center mentality: ‘Get me as many translations as cheaply as you can,’” she says. “If it’s in growth, it’s all about the opportunity: ‘How can we actually integrate this and make connections to everybody and open up the world and make it smaller at the same time?’”
Orriss understood how crowdsourcing would help Facebook pursue that mission, particularly in tongues spoken by small populations. Take the African language of Fula. It has two flavors, one in northwestern sub-Saharan Africa and another spoken around Nigeria. They are sufficiently distinctive that it’s almost two different sets of languages. There is no way Facebook could spend resources finding linguists who knew those dialects and understood technology. So Facebook unleashed its translation tools.
It was all worth it. Even in countries where many languages were spoken but official business was conducted in one language or dialect, the local vernacular made for better, and more profitable, Facebook users. People engaged a lot more when the language was theirs. “Local is what’s relevant to them,” says Orriss. “Who wants to see something translated of what’s going on in the other half of the world? They want to know what’s going on right here.”
But one aspect of this expansion presented a problem that Facebook largely ignored: by crowdsourcing the process of internationalization Facebook started operating in regions where none of its employees, or almost none, actually spoke the language. This meant that Facebook could not provide suitable customer support, or oversight on what people posted. Posts that violated Facebook’s standards could be dangerous or lethal. Even when local users reported violations, Facebook was hamstrung in responding, because its people didn’t speak the language.
IN 2013, THOUGH, it wasn’t those consequences that worried Facebook about its international growth. It was the concern that even with its best efforts, the company might soon hit the ceiling of its possible member base. Yes, it was amazing that Facebook could enlist a billion or two people into its community. But that wasn’t the whole world—just the easy part of it, the ones who had Internet connections and enough money to pay for data. Many among the not-yet-Facebooked billions were not in that cohort. As Facebook saw it, they were either too poor or did not have a way to connect to the Internet. Or both.
The solution? Make a version of Facebook that was cheaper to use. And take the initiative to get them connected, even if it meant creating infrastructure.
Mark Zuckerberg embraced this idea with a passion. Though the impetus came from his Growth team as part of its drive to get everyone in the world connected to Facebook, he claimed to view the venture as sort of a philanthropic effort that would transform the lives of those billions. He would come to call the project Internet.org, invoking the top-level domain suffix usually reserved for nonprofits and foundations.
In 2013, Zuckerberg would write a ten-page white paper laying out the vision, entitled, “Is Connectivity a Human Right?” The answer was a resounding yes. “Everyone will benefit from the increased knowledge, experience, and progress we make from having everyone connected to the Internet,” he wrote. “Connecting the world will be one of the most important things we all do in our lifetimes.”
When he briefed me about it that August, I asked why we shouldn’t view the initiative as just a way for Facebook to get more customers.
“We do theoretically in some way benefit from this,” said Zuckerberg. But that wasn’t why he was doing it, he insisted. “I think the thing that’s kind of crazy [about people thinking Facebook is doing this for profit] is that the billion people who are already on Facebook have way, way more money than the next 6 billion people combined. So if we really wanted to focus on just making money, the right strategy for us would be to focus solely on the developed countries and the people already on Facebook and increasing their engagement rather than having these folks join.” Facebook, he said, may never turn a profit from its efforts. “But I’m willing to make that investment because I think it’s really good for the world.”
Facebook devised a number of initiatives to spread the Internet—and itself—to what was called “the next few billion.” First, it addressed the problem of many underserved regions: people could not afford to pay the data charges that came from using the mobile Internet. Facebook had two ways to deal with that. The first was to create versions of Facebook that consumed less data. Second, it would partner with telecom carriers to deliver a selected slice of the Internet with no data charges. Facebook, of course, would be included in that slice.
The most ambitious part of its plans, though, was to actually create the means of delivering Internet to the next few billion users. For several years at his F8 keynotes he would sketch out his plans with pride.
One scheme was the satellite that would beam connectivity to sub-Saharan Africa. That was the one that blew up on Elon Musk’s launchpad when we were in Nigeria.
Zuckerberg had another means of broadband access in mind, though, an even sexier one: beaming down the Internet from solar-powered, super-lightweight drones. By mid-decade it had become almost a fetish for him. The Facebook air force! The drones would be capable of circling at high altitudes while directing broadband signals Earthward. (Google had its own high-altitude Internet-delivery scheme involving modified weather balloons.)
This was a geeky dream that had gotten some serious attention: a company called Ascenta was actually building such an aircraft. Its CEO had formerly created a thrill ride for the Jurassic World theme park. Facebook bought the company for a reported $20 million and began building a prototype drone, dubbed Aquila. Its wingspan, covered with solar paneling, was the same as an Airbus A320 weighing nearly 100,000 pounds, yet the exotic materials of its frame kept the weight down to under a thousand pounds, less than a standard sedan. Aquila became Facebook’s unofficial mascot. There was a period when Zuckerberg would lead visitors to a piece of an Aquila wing he just so happened to have hanging around, standing taller than he was, and he would lift it like a kite.
After a few years of relentless fanfare, Aquila was finally ready for a private test flight in 2016. Zuckerberg flew to the Yuma, Arizona, test site and watched as it took to the air. Upon his return, Facebook summoned a reporter from The Verge to write about how brilliantly the plane performed. Later, it emerged that the plane had a “structural failure” that led to damage on the landing and triggered a National Transportation Safety Board investigation. Zuckerberg had failed to mention it in his narrative. A second test flight reportedly went better but Facebook gave up on the idea in 2018.
The immolation of the satellite and the crash of Aquila were physical representations of wreckage in Internet.org’s policy initiatives, particularly in the company’s biggest target, India. This country, with a population of more than a billion, was a shortcut to the Growth team’s goal of worldwide ubiquity. Zuckerberg made it a personal crusade. He visited India in 2014, schmoozing with the prime minister and visiting a rural classroom. Time magazine’s Lev Grossman accompanied Zuckerberg, and while acknowledging the value of connecting the underserved, he noted that there was another way of looking at the program. “However much the company spins it as altruistic, this campaign is really an act of self-serving techno-colonialism. . . . Facebook, like Soylent Green, is made of people, and it always needs more of them.”
While the drones and satellites were meant to provide access to the 15 percent of the world without a signal, the real foundation of the initiative was the Internet.org program that partnered with local telecoms to deliver services with no data charges. Facebook launched it in India in 2014, as well as in other countries. The program drew criticism because its inclusion of a group of selected apps—including Facebook—put competitors at a disadvantage. This seemed to violate the principle of net neutrality, which dictated that all developers should have equal access to the Internet. In April 2015, some publishers abandoned the program.
Facebook announced that it would open Internet.org to all developers, but the criticism persisted. Zuckerberg recorded a late-night video from Facebook headquarters and posted a plea that begged critics to back off. Universal connectivity can coexist with net neutrality, he insisted—no one was blocking any sites. Of course, the critics weren’t complaining about that but about Facebook giving itself an advantage by including itself in the “free” tier. Zuckerberg’s answer for that? “If someone can’t afford to pay for connectivity, it is always better to have some access than none at all.”
The critics were not quelled. Zuckerberg renamed the program Free Basics to eliminate the complaint that Internet.org falsely promised the entire Internet. But in February 2016, India banned the service. It ran into trouble in other countries as well.
The Internet.org debacle had elements that would later reappear in the Facebook narrative: officials and the general public rejecting a seemingly benevolent initiative with collateral benefits to Facebook. Before challenging Facebook was cool, they did not trust that the company’s motivations were pure.
Facebook now claims that the program is successful, with around a hundred million people using it. But even the successes would turn out to be woes, as regions unprepared for sudden access to a huge unmoderated speech platform were exposed to those who used Facebook to manipulate, spread misinformation, and promote violence.
By then Mark Zuckerberg would be keynoting much less on Internet.org. He would have other products to tout—and new mistakes to defend.
PALIHAPITIYA LEFT THE Growth team in 2011, and Javi Olivan took his place. As Chamath’s protégé, he was just as single-mindedly obsessed with boosting Facebook’s numbers. But he was also an engineer, and under his leadership, he expanded Growth to include coders outside the circle who previously implemented initiatives while reporting to other organizations.
His personality was different. When asked what the differences were, Alex Schultz replies, “Javi is very kind, very considerate, just . . . ethical and a great guy.”
The other big change was something more profound. To date, Growth had focused on adding users and engagement for the Blue app, the core Facebook product. With Zuckerberg’s blessing, Olivan expanded that domain. Critical portions of the company came under the Growth umbrella.
Growth became the lens through which Facebook viewed almost everything it did. “The way that the Growth team thinks about the product is unlike the way any entrepreneur or any businessperson in the world ever thinks about a product,” says Rob Goldman, who joined Facebook in 2012. “I’ve never seen it before, which was a deep bedrock assumption that everyone in the world should probably be using the product every day.” While Zuckerberg extolled the mission of connecting everyone in the world, the actual operations of doing so were owned by the Growth organization—so one could make a credible argument that the company’s real mission wasn’t connectivity but growth. Facebook’s PR person, Brandee Barker, recalls a meeting she had with Zuckerberg around 2009 where he told her that Facebook needed more publicity to increase growth and engagement. “That was the only reason we were supposed to be doing it!” she says.
“In the beginning, things were a lot smaller,” says Naomi Gleit of the Growth team. “We were only focused really on growth—we owned the registration experience, we owned the invitation experience, we owned that new-user experience. And then, over time, our scope grew. Mark was like, Okay, can you take on this? Can you take on growth and Messenger? And then can you take on . . .” She laughs, knowing there are a number of things she could list to fill in that blank.
Growth was in charge of internationalization. Growth would own mobile. And in recent years, when it became clear that Facebook was going to have to spend vast resources to win back trust—trust that in part the practices of the Growth team had eroded—Growth was put in charge of that too. Some people found it ironic that the Growth organization, notorious inside Facebook for flirting with the algorithmic dark side, was now charged with integrity. Even the area of what Facebook calls “social good” was turned over to what was Facebook’s increasingly powerful Growth organization.
Gleit explains the reason is that her team’s characteristic approach to tackling problems scales well to other initiatives. “What do these things really all have in common? It’s about taking this data-driven, product-driven approach to problems. It really is all just taking the same approach of understand, identify, execute.”
The data-driven, take-no-prisoners DNA of Growth, the Palihapitiya way, is baked into all those initiatives. Palihapitiya left the company in 2011 to start a venture capital fund. (Investors included Facebook, and several of its current and former employees, though not Zuckerberg.) In his farewell memo, he said that the journey was all about winning—everything else comes in second, he said—and warned Facebook folk to be alert to spot “the company you don’t know,” whose big ideas might displace you.
All good, but the memo was remembered at Facebook for one final Chamathism as he slipped out the door.
Don’t be a douchebag.