Part IThe Story
The Bench at the Caltrain Station
In early April 2012, the most coveted photo-sharing company on earth was, for the length of one train ride, a very tall man sitting on a bench at the Caltrain station in Palo Alto. Kevin Systrom was twenty-eight and six feet five. He had once worked behind an espresso machine a few blocks away, and he had once worked as a product marketer at Google a few miles away. Now he had just left Mark Zuckerberg's house carrying an offer. He was waiting for Mike Krieger, his cofounder, who was on his way down from San Francisco, so that the two of them could decide what to do with it.
The offer was $1 billion: $300 million in cash and the rest in Facebook stock, made weeks before an initial public offering that analysts expected to value Facebook at around $100 billion. Instagram was eighteen months old. It had thirteen employees and no revenue. It ran out of a cramped former Twitter office in San Francisco's South Park neighborhood, a room with dark gray walls and no windows that the team called "the cave." The week before, it had released an Android app that was downloaded more than a million times in less than a day. It had also nearly closed a $50 million round from Sequoia, Thrive, Greylock, and Benchmark at a $500 million valuation, which counted as a wild number at the time. Zuckerberg had looked at that number and doubled it.
The bench is a useful place to begin, because nearly every line of Systrom's life runs through it. Seven years earlier, in the summer of 2005, Zuckerberg had taken a Stanford undergraduate named Kevin Systrom to dinner at Zao Noodle Bar on University Avenue and asked him to skip his senior year to build a photo service for a site then called The Facebook. Systrom said no. In February 2012, about two months before the bench, Zuckerberg's interest came back through Matt Cohler, a former Facebook vice president of product who had become a Benchmark partner and an Instagram investor. Cohler texted Systrom what Zuckerberg had told him: "i'm not sure if this is a good idea yet, but i think maybe facebook should buy instagram, what do you think?" Systrom answered with three questions, typed like a man drafting a risk memo:
a) will he go into destroy mode if I say no b) will he understand if we choose to raise instead c) will he understand that I don't want to shutter the product and that doesn't align with what FB does with companies
Cohler wrote back that on (a), probably, and probably also if they just didn't engage at all. On (b), no, "he'll go harder into destroy mode then." On (c), he guessed Zuckerberg would most likely want Instagram to become "a stand-alone mobile facebook photos app, like beluga turned into facebook messenger." Earlier in the same exchange, Systrom had written that he was "not interested really — even at the right price I don't think so."
There was a second suitor, too, and the record on it is unflattering. According to people on both sides of the talks, whom The New York Times later quoted, Systrom and Krieger had met with top Twitter executives as late as March and had verbally agreed weeks earlier on a price of $525 million in cash and Twitter shares. On March 20, Systrom told Twitter that he and Krieger had decided to "remain independent." Less than three weeks later, Twitter learned about the Facebook deal along with everyone else.
So the man on the bench was not innocent of the game. He had already said no to Zuckerberg once, to his face, and had already said no to Twitter in writing. Somewhere in his phone was a text saying he wasn't interested at any price. He was holding a number that was almost twice what Twitter had offered and exactly twice what the venture market had just agreed Instagram was worth. And he was waiting for the one person who could tell him whether he was about to do the thing he had spent seven years not doing.
The answer was yes. The bench is where the later paradoxes start, though, and the rest of this story moves outward from it: the rocket boosters, the zero of three hundred, the witness stand in Washington thirteen years later. There is an old Silicon Valley question that the moment hands over in its plainest form. What does a person with taste actually own once he has sold the thing he made?
By the Numbers
The Systrom Ledger
25,000Users who signed up on Instagram's launch day, October 6, 2010
13Employees when Facebook agreed to buy the company
$1BAnnounced price, April 9, 2012: $300M cash plus stock
~$400MSystrom's reported take, on a stake of roughly 40%
1BMonthly active users by June 2018, three months before he quit
0 of 300Incremental video headcount Instagram received in 2017, per his own email
3BInstagram users today, inside a company that once asked him to drop out
The Boy Who Pestered the Record Store
Holliston, Massachusetts, is a town of about 14,000 people some twenty miles west of Boston. It has well-kept lawns and long driveways, and its parents are mostly professionals with clear ideas about where their children ought to end up. Systrom grew up there in a two-story house framed by trees. His father, Doug, was a vice president of human resources at TJX Companies in Framingham. His mother, Diane, worked in marketing and caught the first tech boom early: Monster.com, a dot-com called Swapit, and later Zipcar. His family name is pronounced "Sisstrum." Hometown reporters in 2012 thought that worth explaining, which tells you how far he had traveled.
The family bought its first computer when he was twelve. He has said many times, and with the fixed earnestness of a man describing a conversion, that the turn came from Doom II. "I was like a Doom II guy," he said in a video interview the Boston Globe dug up after the Facebook deal. "Doom II all the way… I credit Doom II for everything." He built his own levels. He learned QBasic. Once he had AOL, he wrote small Visual Basic programs to knock friends offline and others that made it look as if he had hacked their Instant Messenger accounts. These are the ordinary hobbies of a clever adolescent boy. Their whole appeal was making something on another person's screen look different from what it actually was.
The teenager who comes through in the archive has two distinct sides, and the more interesting one has nothing to do with code. In high school he got obsessed with deejaying and decided he needed to work at Boston Beat, a vinyl shop on Newbury Street. "I would e-mail the store every other day to get a job there," he told Fortune. "I pestered them and pestered them and finally they let me in, and I worked a couple hours a week." Before long he was opening for real DJs at Boston clubs, relying on older friends to get him inside because he was not yet eighteen. He piled records in a corner of his bedroom. "One thing about my background," he said, "is I get very obsessed with something."
He ran cross-country at Holliston High in eighth and ninth grade and won rookie of the year. Phil Ryan, who had coached there for thirty-eight years, remembered him with an old coach's highest compliment: "He was never late, not Kevin." In tenth grade he transferred to Middlesex School in Concord, a boarding school whose alumni include Steve Carell and William Weld. He would later be described as captain of the lacrosse team there. At Middlesex he took every computer course offered. "They waived my requirement to do biology," he recalled. "I still haven't to this day taken a biology class." He put an antenna outside his dorm window and started a radio station, and then found out that broadcasting from his room was illegal.
Photography ran alongside all of this. As a child he played with his father's Polaroid and asked for a new camera every year. At summer camp he took a Photoshop class. In an essay he wrote as a junior at Middlesex, preserved and later quoted by the head of school, Kathleen Giles, he called photography his way "to show my outlook on the world to everyone." He hoped it would "inspire others to look at the world in a new way." When his father was asked about him in 2012, he did not mention computers first. "I remember that art history and photography were two classes Kevin really enjoyed," Doug Systrom wrote in an email to the local paper, "while focusing on computer science and advanced math classes."
Read with hindsight, the whole kit is already there. He had the patience of a crate-digger who will flip through a thousand records to find the one that sets a mood. He was persistent enough to email a store every other day. He was on time for practice. And he wanted, as he put it at sixteen, to change how other people looked at the world. Nothing in Holliston told him this could become a business. "That idea that you could get rich really quickly off of starting a start-up didn't really exist in Massachusetts, on the East Coast, during that time," he told Kara Swisher. He graduated from Middlesex in 2002 and went west, and as the Boston CBS affiliate noted with a touch of regional hurt, he never came back.
The Offer He Refused First
He applied early decision to Stanford intending to study computer science. He found the classes "more academic than applied," struggled in at least one of them, and switched to Management Science and Engineering. That program was heavier on finance and economics and lighter on proofs. He kept writing code in his spare time anyway. He built a campus version of Craigslist that about 8,000 Stanford students used. He started a digital radio station, because apparently he could not stop. With his Sigma Nu fraternity brothers he set up a site for sharing party pictures internally. Out of that came Photobox, a tool for sending large photo files between people. "I saw how social photos were a part of the fabric of college life," he said later.
Photobox caught the attention of two people who would keep showing up in his life. One was Mark Zuckerberg. The other was Adam D'Angelo, whom Systrom had met at a Sigma Nu party over beers in red plastic cups. D'Angelo was a 2006 Caltech graduate who became Facebook's chief technology officer before going off to start Quora. He was soft-spoken, precise, and an engineer's engineer, and he later explained Instagram's whole thesis in one sentence: "You can look at Facebook as this bundle of so many different things, but it turns out that people just like photos more than anything else." He would also be the person Systrom phoned the night everything broke.
The Zao Noodle Bar dinner took place in the summer of 2005. Zuckerberg wanted Systrom to leave school and build photos for Facebook. Systrom turned him down. "Working at a startup to make a lot of money was never a thing, and that's why I decided to just finish up school," he told Forbes in 2012, over $4.50 cups of artisanal coffee at Sightglass in SoMa. "That was way more important for me." His account to Fast Company in 2011 was less self-assured: "I would've loved to have been part of Facebook's growth over the years, but it was the first time I met those guys."
What he did with the senior year he had protected is the part his own mythology leans on hardest. In the winter of his third year he went to Florence to study photography. An Italian professor there showed him a Holga, a cheap plastic camera, first popular in China, that had developed a cult following for its soft, vignetted, light-leaking pictures. Its flaws were the reason people loved it. Systrom adored the look. The Holga is the earliest object in the Instagram story. It made ordinary moments look nostalgic because it handled light badly, and the people who loved it treated the bad handling as character.
He was one of twelve students chosen that year for Stanford's Mayfield Fellows Program, a work-study course in how technology companies grow, and the program required an internship. He wanted to work at Odeo, the podcasting startup run by Evan Williams, who had founded Blogger. To get Williams's email address he ran a WHOIS lookup on Odeo's domain. Williams didn't reply the first time, or the second. After the third email he agreed to meet. It was the Boston Beat method moved to Silicon Valley. "In fact," Systrom recalled, "I think the day Jack started his first day as an engineer at ODEO was my first day as an intern." That was Jack Dorsey, who went on to build Twitter at Odeo after Systrom left and who later invested in Instagram. "That was my first foray into social media," Systrom said.
The scene that best captures the arc of these years comes from the spring of 2006. Systrom was working behind a hissing espresso machine at Caffé del Doge in Palo Alto. Zuckerberg came up to the counter looking puzzled. The prior summer he had offered this young man millions in Facebook options, and Facebook was now worth $500 million. Here was the young man making cappuccinos. "I had been like, 'No, I don't want to work at this thing,'" Systrom remembered, "and here I am working at a cafe." Forbes later put the point about this moment in one line, and it applies to the whole story: in Systrom's case, "the place you end up is exactly the place you turned down."
After graduation he joined Google as an associate product marketing manager and worked on Gmail, Calendar, Docs, and Spreadsheets. He later moved into corporate development. At Google he learned, as he put it, "how to talk about products." He also learned what peak comfort looked like. Years later he described standing at the No Name Cafe on the Googleplex watching chefs shuck oysters and thinking: "We've got to be at a moment in technology, or a moment at this company, where it's the best it's going to get." He was wrong about Google. He was right about himself. After about two years he left for Nextstop, a small social travel-tips startup started by former Google colleagues, where he stayed a little over a year. At night he taught himself to build things.
Burbn, or the Art of Subtraction
The project began on October 21, 2009. That day an account called @burbn posted to Twitter: "Well hello there good lookin'." Burbn took its name, in typical Valley whimsy, from Systrom's interest in fine whiskeys and bourbons. It was a location check-in service very much like Foursquare and Gowalla, the two companies then fighting over who would own the idea of saying where you were. Burbn layered several things on top of the check-in. You could post future check-ins called "plans." You earned points for spending time with friends. You could post pictures.
Systrom has explained the technology choice with refreshing candor. "So the dirty secret of Instagram," he told Kara Swisher in 2017, "is that I didn't know how to make an app so I just made a website instead and I went around marketing it to VCs as like, 'Hey, HTML5 is the future. Forget about apps, this works on Android and IOS.' But the dirty secret was I didn't know how to make an app." He later summed up this quality in an interview headline: "I'm dangerous enough to code and sociable enough to sell our company."
The selling worked. Ronny Conway, son of the angel investor Ron Conway and a former Google colleague, had moved to Andreessen Horowitz and asked Systrom to meet his bosses. Systrom already had a $250,000 commitment from Steve Anderson at Baseline Ventures. He had barely a product and no deck. Marc Andreessen co-wrote Mosaic, co-founded Netscape, and had built his new firm on the view that 97 percent of venture returns in a given year come from about fifteen companies. He and Ben Horowitz looked at the Burbn prototype and insisted on investing $250,000 of their own. On March 5, 2010, within roughly two weeks of quitting Nextstop, Systrom closed a $500,000 seed round. Both investors told him the same thing: find a cofounder, because you have to learn to build an app.
The cofounder was Mike Krieger. He fits the pattern of Systrom's people closely and also complicates it. Krieger was born Michel Krieger in São Paulo in 1986 and moved to California in 2004 to attend Stanford, where he majored in Symbolic Systems, a program combining computing, psychology, and cognitive science. Early on he had wanted to be a journalist. He took nearly every human-computer interaction class Stanford offered and stayed on through the Mayfield Fellows program, which is how the two met. After graduating he became an engineer and user-experience designer at Meebo. He was one of Burbn's first beta users. He joined the project on May 19, 2010, and Burbn's Twitter account announced it with an exclamation point. Systrom had grown up with the East Coast's assumptions about how a career goes. Krieger was an immigrant who would later tell Bloomberg that getting his visa took longer than building Instagram did. Inside the company he came to be seen as the "heart and soul."
They studied the data, and it was humbling. Burbn had a few hundred users at most. Asked at a Stanford talk in 2011 how many people had used it, Systrom counted the hands that went up: "Like three people… So that's why we started working on Instagram because that's basically the number of hands that went up in the room when we were working on it." The users ignored plans and points and check-ins. What they did was post a lot of photos. Systrom and Krieger looked for apps that did photos well and found that the photo apps were not social and the social apps did not handle photos well. So they did the hard part of any startup, which is removing things. Krieger later recalled that they went back to what the Mayfield program had drilled into them: too many features, and users did not want a complicated product.
The filters came from a beach. During a trip to Todos Santos, a bohemian beach town in Baja California Sur, Systrom was sketching app ideas when his girlfriend told him she didn't want to post her pictures because they didn't look good enough coming off a phone camera. The fix was the Holga's logic in software: if the camera produces poor images, make the poor quality look intentional. Systrom wrote the first filter himself and called it X-Pro II. It is still in the app. He has described this with some pride and some sheepishness. The filters "effectively" hid how inferior the photographs were.
The name joined "instant camera" and "telegram," Polaroid and wire, and the app came out square because the Kodak Instamatic and the Polaroid were square. Three decisions settled the new product, and every one of them was a subtraction. Sharing a photo did not require saving it to the camera roll first. There was one feed, one picture at a time. And you could push the image out to Twitter, Facebook, and Flickr in the same motion. Chris Dixon later named the strategy: "Come for the tool, stay for the network." The filters were free when competitors charged for them, which gave people a reason to download it. Sharing to other networks gave people a reason to tell someone else about it.
They built it in about eight weeks. On July 16, 2010, at 5:26 p.m., Krieger posted the first test photo: South Beach Harbor at Pier 38, with boats, water, and late light. Systrom's first image was smaller and more personal. The Guardian later summed up the origin in one line: Instagram "began… with a photo of a dog and his girlfriend's foot."
There was one problem with the investors. Andreessen Horowitz had also seeded Picplz, a competing photo-sharing service run by Dalton Caldwell, who had founded the music network Imeem and watched MySpace buy it and then bury it. When Systrom and Krieger showed the firm the new Instagram, the partners were not excited. "We were a little bit stuck," Horowitz said later. "He did a pivot into a company we'd already invested in." The firm that built its model on getting into the fifteen companies that matter put $250,000 into Instagram and then, as the Times put it, effectively bet against it. Systrom's other early backers doubled down at every chance. Andreessen Horowitz still earned an enormous return on the $250,000. In a place where everyone keeps score, the portion it left behind became the story.
A Warehouse Jutting Into the Bay
They launched after midnight from a dimly lit warehouse jutting into San Francisco Bay. The App Store listing went live on October 6, 2010. Within hours, thousands of people had downloaded it, and by the end of the first day 25,000 had signed up. The servers, if one tiny cluster deserves the plural, kept crashing.
Systrom described the feeling to Guy Raz as two sharp reactions back to back: "There was nothing more exhilarating than seeing all those people stream in and nothing more crushing than then seeing, you know, people posting on Twitter or on their blogs and saying like, oh, another startup that doesn't know how to scale. Like, oh, like so clowny." Krieger remembered it this way: "We were both, I mean, at that point, like, running on zero sleep for two days, just devastated. And I was like, this is it. We built this great thing, and we totally messed it up."
What happened next is the scene Systrom would later present to Stanford undergraduates as a parable. Standing in the warehouse with a collapsing system, he asked himself, "Who's, like, the smartest person I know who I can call up?" He scrolled through his phone and stopped at Adam D'Angelo, the red-cup acquaintance from Sigma Nu seven years earlier, who was by then a former Facebook CTO. "Adam spent like 30 minutes on the phone with us," Systrom recalled, "walking us through the basic things we needed to do to get back up." D'Angelo later helped them find engineers, set up databases, and work through features, and then he invested. Telling the story at Stanford in the spring of 2011, Systrom closed with advice that was really an instruction: "Make sure to spend some time after the talk getting to know the people around you."
Within a week, 100,000 people had registered. On December 12, 2010, the count passed one million. Krieger, who did most of the early engineering and user experience on his own, has a story from that stretch that shows exactly how thin things were:
One time I woke up and there was an email saying that the site went down, and I was like, who fixed it? Shane, did you fix it? Kev, did you fix it? No. And eventually on the terminal, you press up and you get to see what was the last thing you typed. So, apparently, at 3:30 in the morning I somehow managed to, in a completely drunken state, revive Instagram.
Josh Riedel had joined that October as community manager, and Shayne Sweeney had joined in November as an engineer. For most of that period the whole company was four people working in the cave. Systrom made the cave a rule. He banned meetings with anyone outside Instagram except the press and engineering recruits. "The first month when we would hear about Instagram killers or competitors, we paid a lot of attention," he told the Los Angeles Times in July 2011, when the app was adding thirty to forty users a minute. "Over time we heard it enough that now we believe the right thing to do is to focus on what we are good at, building the product we want to build. We try not to be too reactionary. At the end of the day, you are your biggest competitor."
— Kevin Systrom to the Los Angeles Times, July 2011Instagram is our baby. It wakes us up in the middle of the night. It makes us lose a lot of sleep. At the same time we love it, and there is nothing else we would rather be doing.
They added hashtags in January 2011 and asked users to make them specific rather than generic tags like "photo." The users responded with #ThrowbackThursday and #SelfieSunday. On February 2, 2011, Instagram announced a $7 million Series A led by Benchmark, with Jack Dorsey, Chris Sacca, and Adam D'Angelo participating, at a valuation reported at somewhere between $20 million and $25 million. The cap table read like an alumni directory of a few years in Systrom's life: the Odeo intern's first-day colleague, the Sigma Nu party, the Facebook VP who would carry Zuckerberg's text. The network that grew Instagram looked a lot like Instagram itself: people who knew each other, connected by images.
When it was all over, Systrom was careful about what credit he claimed for the speed. "I feel very lucky that so far we've managed to do a lot of things exactly at the right moment," he said in April 2012. "As much as entrepreneurs want to take credit for each and every little decision and say it made complete sense at the time, I promise you a lot of it is luck." Ask investors in 2010, though, and the luck had been hard to see. "They're like: 'Photos, why would you do photos? There's no money in photos,'" he remembered. "And they used to show this graph of exits in video and exits in photos, and photos was this tiny small graph and YouTube was huge."
In June 2011 Instagram passed 5 million users and 100 million uploaded photos. By September it had 10 million users. In December, Apple named it App of the Year. Registered users grew thirteenfold in 2011, a figure that would show up on a government exhibit fourteen years later. Snoop Dogg used it, as did Jamie Oliver, Justin Bieber, Kate Spade, Red Bull, and Starbucks. In his rare free time Systrom made mole with his girlfriend. "Cooking is essentially a social activity," he said. "It's also very analog. It has nothing to do with tech."
What the Billion Bought
Facebook announced the deal on Monday, April 9, 2012. It was Facebook's largest acquisition to date, and the comparison everyone reached for was the $35 million Yahoo had paid for Flickr in 2005. "It's the Web fairy tale that all start-ups dream of," a Forrester analyst told the Times. Zuckerberg said publicly that Facebook was "committed to building and growing Instagram independently." MIT Technology Review later reported that Systrom agreed only after convincing Zuckerberg to keep the Instagram brand and let him and Krieger run it.
The details of that period undercut the fairy tale. On August 14, 2012, Britain's Office of Fair Trading approved the deal. On August 22, the FTC closed its investigation. That same late-August week, Systrom testified under oath before the California Corporations Department at a fairness hearing Facebook had requested to speed up approval. A lawyer for the state, Ivan V. Griswold, asked whether Instagram had received any offers besides Facebook's at the time of the negotiations. "No, we never received any offers," Systrom said, according to the transcript. Griswold asked again: "Immediately before the negotiations, did you receive any offers from invest —." Systrom interrupted him: "We never received any formal offers or term sheets. No."
The word "formal" is doing a lot of work there. The Times, citing people on both sides and documents it had reviewed, reported a verbally agreed $525 million price from Twitter and the March 20 message about remaining independent. Facebook's general counsel, Ted Ullyot, had told regulators that potential acquirers showed a "lack of interest." Facebook and Systrom both declined to comment. Nothing came of it legally, and the Corporations Department said it had received no complaints. But this is the one place in the record where the young man who reliably showed up to practice on time can be seen choosing his words like a lawyer. The February texts make a moment visible that founders seldom put in writing: a fear of being destroyed, carried in the same thread as the habit of holding something back.
By the time the deal closed on September 6, 2012, Facebook had gone public in a rocky offering, and its stock had dropped. The final consideration was $300 million in cash and 23 million shares. CNBC later valued it at $715 million. Systrom's stake of about 40 percent made him worth roughly $400 million on paper. Thrive Capital, led by a twenty-six-year-old named Joshua Kushner, had been the second-largest investor in the round that closed only days before the acquisition, and it doubled its money in less than a week. In 2016, after Facebook shares had risen more than 500 percent, Forbes listed Systrom as a billionaire worth $1.1 billion. He had accepted a large share of Facebook stock when it was a gamble, and that stock became his fortune. To an unusual degree, the cofounder's wealth depended on how well the acquirer did.
Six years later, Ben Thompson made the bluntest version of this point at Stratechery. The date that explains the founders' 2018 departure, he wrote, "is April 9, 2012, and the people most responsible are Kevin Systrom and Mike Krieger." That was the day they sold. Whatever independence came afterward was granted by the buyer, and the buyer could take it back.
Matt Cohler, who had brought the first offer, described what Facebook was really purchasing. "This is the first thing I've seen that feels like it's truly native to mobile," he told Forbes. "To have scaled the product, the network of users and the infrastructure behind it is nothing short of extraordinary under any circumstance; to do it with such a small team is unique in the annals of technology." Facebook had been built for the browser era and was trying to rebuild itself for phones. Instagram had only ever existed on a phone.
Systrom's own explanations of the deal have stayed close to one image. "You can either be a rocket yourself or you can decide to strap on rocket boosters and fly even faster," he told Stanford students in 2020. "The thought of combining forces with one of the best social companies at the time and arguably still now made a lot of sense if I wanted Instagram to be in every single household in the United States." Then, without being prompted, he added the second half of the metaphor: "Sometimes the rocket you strap yourself onto wants to go in a different direction, and you have to compromise."
A Separate Company Inside the Mothership
The early test of the arrangement came on December 17, 2012, about three months after closing. Instagram changed its terms of service to say that "a business or other entity may pay us to display your username, likeness, photos… without any compensation to you." Users would not be able to opt out except by deleting their accounts before January 16, 2013. The backlash was immediate, and rival photo services took the chance to recruit unhappy users. The next day Systrom wrote a blog post that admitted the error before it explained it: "our mistake that this language is confusing." "It is not our intention to sell your photos," he wrote. On December 20, the advertising section went back to its October 2010 wording. The Verge pointed out drily that the original terms had always allowed Instagram to use users' photos in ads almost however it wanted. Still, Systrom learned something from it about the people who used his app: they believed the photos belonged to them, and the company had to keep up that impression even where the legal language did not.
Monetization arrived carefully after that. In April 2013 Instagram hired Emily White as director of business operations. In September she told The Wall Street Journal the company should be ready to sell ads by the following September. It started sooner. On November 1, 2013, the first sponsored image appeared in American feeds, from Michael Kors. By December, White had left for Snapchat. That was the first of several times an Instagram executive or idea would end up on the other side of that rivalry. Video ads followed on October 30, 2014, from Disney, Activision, Lancôme, Banana Republic, and the CW. In February 2016 the company had 200,000 advertisers. By March 2017 it had a million.
The team expanded to about 50 people in 2013 and roughly 150 in 2014, when Instagram reached 200 million monthly users. In 2013 Instagram grew 23 percent while its parent grew 3 percent, a gap that would matter a great deal later. Instagram moved onto Facebook's Menlo Park campus, a sprawling and spotless former Sun Microsystems site between the freeway and the marshy edge of the Bay, but Systrom kept it apart from Facebook's own buildings. Across the freeway, Facebook's Frank Gehry building held 2,800 people in a single room. Instagram had its own space. The walls carried posters: "Kick the shit out of Plan B." "Leadership is best measured by what people do when you're not around." "Looking for love in the great forest of like."
There is a revealing story about a conference room. In May 2014 Al Gore visited, sat in a standard corporate meeting room with a white plastic table, gray carpet, and black mesh chairs, and asked with obvious contempt, "This is an Instagram meeting room?" Systrom called a designer friend in San Francisco. The friend brought in green 1970s chairs, a huge white marble table, wallpaper made from employees' fingerprints, and prints of the team's favorite photographs. Systrom told the story to The Guardian in 2015 while being moved from one spot to another for a portrait, sighing "More photos!" in a tone of complaint that, the reporter noted, didn't quite convince. He wore a light gray jacket over a T-shirt. He was thirty-one and ran what she called a $35 billion company. It was his room, and the former vice president had embarrassed him into making it look like it.
This period also brought the founder into more official circles. In 2014 he joined Walmart's board and stayed until 2018. Fortune put him on its 40 Under 40 list and Forbes on its 30 Under 30. MIT Technology Review described him as an example of a new type of founder, "at the intersection of technology and the liberal arts," equally comfortable with runway models and Hollywood actors as with engineers. In February 2016 he met Pope Francis at the Vatican, and they talked about images connecting people "across borders, cultures and generations." He married Nicole Schuetz, whom he had met at Stanford and who later founded the clean-energy investment firm Sutro Energy Group, at a masquerade-ball wedding in Napa.
The world outside the room was rougher. In October 2013 Instagram deleted the account of Canadian photographer Petra Collins over a bikini photo. In January 2015 it removed an account belonging to an Australian agency, and in March 2015 it took down Rupi Kaur's menstruation photographs. The #FreeTheNipple campaign followed. In October 2015 Systrom explained the platform's rules by citing Apple's App Store guidelines and Instagram's 12+ age rating, an explanation many people found inadequate. In May 2017 a survey of 1,479 young people by Britain's Royal Society for Public Health ranked Instagram "worst for young mental health." The same survey credited it with helping self-expression and community. Instagram responded that keeping the platform safe and supportive for young people was "a top priority." Someone who had once hoped to "inspire others to look at the world in a new way" was now in charge of an app that had done that, in ways nobody had fully foreseen.
Breaking the Square
For six years the outside world described Instagram as "famously deliberate" and even "slow," a company that did not want to ruin its own look. Then 2016 arrived and Systrom started taking apart the features that defined the product.
On March 15, Instagram changed its feed from chronological to algorithmic, ranking posts by predicted interest. Users protested and begged their followers to turn on notifications. The company tweeted a reassurance and did not change course. On May 11, it replaced the beloved skeuomorphic camera icon with a flat gradient. On August 2, it launched Stories, which were twenty-four-hour vanishing slideshows. They were plainly modeled on Snapchat, the company Emily White had joined, which had become popular among young people in exactly the way Instagram once had.
Systrom's justifications are among the most revealing things he has said, because they turn the original logic of the product upside down. The square had been the frame that made Instagram what it was. "I think companies that fail are typically companies that look at themselves as a set of features," he told BuzzFeed News on Instagram's sixth birthday. "Companies that succeed look at themselves as mission-based companies… So if Instagram's mission is to make sure that everyone can capture and share the world's moments, and use them to form stronger relationships with one another, how do you say to someone, 'I'm sorry, you can't post that photo unless you crop it into a square and fit everyone in'? That's a ridiculous argument." The man who made millions of people crop their lives into squares had decided the square was in the way.
He didn't play down the copying. He said openly that Snapchat "deserves all the credit" for the format. When Recode pushed him on it, he described the whole industry as a long chain of borrowing: "Day One: Instagram was a combination of Hipstamatic, Twitter [and] some stuff from Facebook like the 'Like' button. You can trace the roots of every feature anyone has in their app, somewhere in the history of technology." He compared it to car companies, several of which can coexist. He also mentioned that Snapchat "didn't have filters, originally. They adopted filters because Instagram had filters." And he was specific about what he had improved: "one of the really annoying things about the format is that it just kept going and you couldn't pause it to look at something, you couldn't rewind. We did all that, we implemented that." Elsewhere he described the attitude as "remix": "all of these ideas are original when you remix them and bring your own flavour."
— Kevin Systrom, to Recode, 2016You can trace the roots of every feature anyone has in their app, somewhere in the history of technology.
The numbers backed him up, and that is the uncomfortable part. Two months after launch, Instagram Stories had 100 million daily users. By January 2017 it had 150 million. In April 2017 it reached 200 million, more than Snapchat's entire user base. By June 2018, 400 million people were using it every day. Instagram's total users went from 500 million in June 2016 to 600 million that December, a jump TechCrunch attributed partly to the algorithmic feed. In October 2016 the company celebrated its sixth birthday by moving into a new three-story building in Menlo Park, and Systrom posed for BuzzFeed in front of a photo wall at the Blue Bottle café inside it. He was, the reporters wrote, "upbeat — happy, even."
The obvious reading is that taste lost to scale. The more accurate reading is that Systrom's taste had always been a skill used in service of something else. The Holga was a hack that made a weakness charming. X-Pro II hid a bad camera sensor. The square imitated the Instamatic because borrowing the look of a beloved old object was the fastest way to make a new one feel familiar. He was never mainly a preservationist. He was someone who rearranged things, who could look at a feature and tell which part of it people actually loved. In 2016 he applied that to his own product and concluded that what people loved was sharing moments, and that the square was incidental. The posters on the wall had said so all along: kick the shit out of Plan B.
Zero of Three Hundred
In 2017 Systrom sent an email to Instagram's leadership that would be shown to a federal judge eight years later. Facebook was making a large push into video and had approved 300 additional hires for it. Instagram received none. "We were given zero of 300 incremental video heads," he wrote, "which is an unacceptable and offensive outcome." In another email, to Facebook CTO Mike Schroepfer, he wrote that Instagram had areas that were "'starving'" for investment.
The mood had changed, and the tension became something you could count. Instagram had become what Recode called "a consistent beacon of good news" for a parent company dealing with Russian election interference, Cambridge Analytica, and congressional hearings. In June 2018 Instagram reached one billion monthly users, about half of Facebook's total, with a small fraction of Facebook's staff. Fortune reported that Instagram was "reported to be worth $100 billion" and seemed to have a "Teflon-like force field" that kept it clear of its parent's scandals. That same year Fortune noted two milestones in Systrom's personal life: he became a father, and he earned a first-level sommelier credential. He posted both on Instagram, of course.
Inside Facebook, the beacon had started to look like a threat. In 2018, as Facebook's U.S. user numbers leveled off, Zuckerberg wrote a memo to his leadership team warning that Instagram's success could contribute to a "network collapse" of Facebook. He raised the question of whether it might be better to spin Instagram out entirely. The word "network" had run through Systrom's story from the beginning: the Stanford network, the red cups, the "come for the tool, stay for the network." Here it appeared in the opposite sense. The thing Instagram created was the thing Facebook was afraid of losing.
Over the following months the changes came one after another, and the reporting assembled them. In May, Kevin Weil, Instagram's well-liked VP of product, was moved to Facebook's new blockchain team. His replacement was Adam Mosseri, the former head of Facebook's News Feed and a member of Zuckerberg's inner circle. Zuckerberg reorganized all product divisions under Chris Cox, which added a layer between Systrom and Zuckerberg. Facebook removed Instagram branding from photos reshared onto Facebook. Zuckerberg sharply reduced the notifications and cross-posting links that sent Facebook users to Instagram. Systrom would later testify that an internal chart showed those integrations helped Instagram grow while having a neutral effect on Facebook. There was a long fight over whether IGTV, which launched June 20, 2018, as a standalone video app, would compete with Facebook's own video service. After Cambridge Analytica, Facebook invested heavily in data privacy and integrity, and Instagram, Systrom said, got no new people. COO Marne Levine left Instagram that September for a bigger role at Facebook.
Then Systrom went on parental leave. Kara Swisher, who had known him since right after the acquisition, reported in the Times that his first act upon returning on Monday, September 24, 2018, was to quit. Krieger left with him. The announcement was hurried. Facebook's communications people didn't know what was going on, beyond a brief story Mike Isaac had published in the Times. Casey Newton compared it to a flight attendant who drinks a beer and pulls the emergency slide on his last day. Swisher placed it alongside earlier exits: the WhatsApp founders, the Oculus team, and "a conga line of other founders of start-ups the social media giant has swallowed whole," most of whom, she wrote, "took the money and ran." The Instagram founders were different, according to several of her sources. "In a perfect world, they wanted to continue to build and were not bored at all," one said. "But they were frustrated by an inability to take it to the next level."
Systrom's statement went up on Instagram's own blog: "Mike and I are grateful for the last eight years at Instagram and six years with the Facebook team. We've grown from 13 people to over a thousand with offices around the world, all while building products used and loved by a community of over one billion. We're now ready for our next chapter." It closed: "We remain excited for the future of Instagram and Facebook in the coming years as we transition from leaders to two users in a billion." Zuckerberg's reply was short enough that Ben Thompson read it as revealing: "Kevin and Mike are extraordinary product leaders and Instagram reflects their combined creative talents. I've learned a lot working with them for the past six years and have really enjoyed it." On October 1, Facebook named Adam Mosseri head of Instagram.
Three weeks after quitting, Systrom went on stage at WIRED25 in San Francisco, following Jony Ive and Anna Wintour, with what looked like real relief. Lauren Goode pointed out that he was the only speaker who was technically unemployed. "I'm doing okay," he said, and the audience applauded. He set his own departure against a scale he clearly had worked out: "The PayPal guys, it was like two days… Maybe it might have been one day." Then he used the rocket image again: "it's kinda like a rocket. You have to aim it in the right direction, put the boosters on it, hope that it gets into the right orbit, but then take your hands off and let it go." The rest of that interview is more guarded. One headline from that week captured the line that wasn't guarded: "No one ever leaves a job because everything's awesome."
Two Users in a Billion
What followed looks, at first, like scattered sabbatical projects, and on closer reading is another round of the same obsessions. In early 2019 he decided his next challenge would be learning to fly. He had his first lesson that afternoon, in what he called a decades-old "lawnmower with wings," and got his pilot's license three months later. He got into competitive cycling and took something like a rule for living from it. "The thing I learned from cycling is that the best cyclists in the world are the cyclists who truly enjoy training and the pain of training a lot," he told Stanford students in 2020. "You have to choose an area of life where you enjoy the practice, you enjoy the pain, because you won't escape the pain anywhere." He told Tim Ferriss about Ray Dalio's Principles, Eliyahu Goldratt's The Goal, and Mortimer Adler's How to Read a Book, the last a nearly ninety-year-old manual on reading closely, offered by the man whose app had trained a billion people to scroll. He made angel investments, among them Superhuman and Clubhouse.
He began giving the kind of advice that only comes after you have lost something. "I tell first-time founders to make sure their product solves a core problem in people's lives," he told Sequoia in 2019. On becoming a father: "Being directly responsible for another human being makes you ask questions you might not otherwise — including about the products you're building." He mentioned the machine-learning systems Instagram had built against bullying and the tools for seeing and limiting time spent. "It made me think more about what our goals should be in terms of the utility of the product versus the time you spend on it." For new founders, his main lesson was about letting go: "If you try to be the factory that produces ideas, you'll always be constrained by your own time, ability, energy level, creativity. Instead, build that factory."
On April 18, 2020, as the pandemic hit, he and Krieger launched Rt.live, a state-by-state tracker of how quickly COVID-19 was spreading. It was the statistics enthusiast's side project, built on Bayesian estimates of a number the public was just learning to fear. On January 31, 2023, they launched Artifact, an AI-driven news app that used machine learning to recommend articles. It was in some ways a reversal of Instagram: text rather than images, reading rather than performing. The reviews were generous. In March 2023, the same year, he said publicly that Instagram had "lost the soul" it once had. In January 2024 the founders announced that Artifact would shut down. In April 2024 Yahoo acquired it to fold into its news products. Krieger, the immigrant engineer who once fixed Instagram drunk at 3:30 a.m. and couldn't remember doing it, joined Anthropic as chief product officer on May 15, 2024.
In a November 2019 interview with New York magazine, a reporter asked him what it felt like to have changed the world with Instagram and to see the world change to fit Instagram. He answered with unusual honesty: "It's really strange to me to walk down the street and see people using a product that I remember very clearly making, but what's weirder is it doesn't actually wow me. When we launched, I think it was 24 hours later, there was someone using it on the subway on the way home. I was blown away." He added: "We got fairly lucky in that, it turns out, what we wanted, a lot of other people wanted. That's not lost on me, but it is kind of lost on me." In the same interview he described going to a music show where his phone was taken away. He said there was a club he liked where they told him, plainly, that you could not use Instagram there.
— Kevin Systrom, New York magazine, November 2019We didn't set out to change the world; we just set out to make a good product that we wanted to build and we wanted to use.
The Witness Stand
On Tuesday, April 22, 2025, Systrom sat for about six hours in a federal courtroom in Washington, D.C. Zuckerberg had sat in the same seat the week before and told the court Instagram would probably not have become what it is without Facebook. The Federal Trade Commission was asking U.S. District Judge James Boasberg to unwind the 2012 acquisition. Columbia's Tim Wu had for years been calling that deal a felony, and an internal Facebook memo surfaced in 2019 had described its purpose as removing a potential rival. By then Instagram had nearly three billion users, was expected to bring in about half of Meta's U.S. ad sales that year, and contributed more than $20 billion a year to Meta's revenue.
FTC lawyer Bob Zuver brought out the exhibit charting 2011, the thirteenfold year. "The users, they just kept coming," Systrom said. He testified that Instagram could have built video and private messaging on its own. It had run on Amazon Web Services and didn't need Facebook's servers. "We would have been able to scale our problematic content screening fairly well," he said. "It wasn't rocket science." The phrase was striking from someone who had spent ten years describing his company as a rocket.
He talked about Zuckerberg with the controlled care of a man who has rehearsed his answers to provocative questions. "We were a threat to their growth," he said. Of the cuts to the integrations: "If Instagram didn't grow as quickly, Facebook wouldn't shrink as quickly, or plateau as quickly. I don't think he ever said it out loud that way, but that was the only reason we were having this discussion." Of the "zero resources" during the integrity buildup after Cambridge Analytica: "I thought that that was not appropriate given the scale of Instagram." And of the man who had asked him to drop out of Stanford: "He was always very happy to have Instagram in the family because it was growing so quickly and we did great product work. But also I think as the founder of Facebook he felt a lot of emotion around which one was better, meaning Instagram or Facebook, and I think there were real human emotional things going on."
— Kevin Systrom, testifying in FTC v. Meta, April 22, 2025I was working very hard for the company to make this a success and not getting resources back. It was in stark contrast to the effort I was putting in.
Under cross-examination by Meta's lawyers, he made concessions, and he made them the way a careful person gives up exactly what he has to. Yes, every team fights for headcount. Yes, Facebook "gave us many resources that allowed us to thrive." And on the central question, whether the company he sold would have made it without the sale: "It could have gone either way," he said. "The probability of us failing was low, but it could have failed."
All of it comes back to the bench, and to a young man in a café watching the founder who had offered him everything walk up to the counter puzzled. It comes back to an email to Evan Williams sent three times. It comes back to Doom II levels and a pirate antenna at boarding school, a sixteen-year-old's essay about showing the world his view of it, and a cheap Italian-sold plastic camera whose flaws were the point. Hundreds of billions of images later, after the posters and the marble table, after the Pope and Walmart's board and the masquerade ball in Napa, after the square was broken and Snapchat was copied and the zero of three hundred, after the cold office where at one in the morning he could see Krieger's breath, the first picture he ever posted is still the first.
A dog, and his girlfriend's foot.
Part IIThe Playbook
What follows comes from the record in Part I. These are not slogans from a stage. They are patterns that hold up across Holliston, Stanford, the cave, Menlo Park, and a federal courtroom. Some of them are about building, some are about selling, and a few are about what happens once you can no longer control what you built. The principles are uneven on purpose. One of the clearest lessons of Systrom's career is that the same instinct that makes a product will, under different conditions, cost you the company.
The principles
- 1.Subtract until only the existing behavior remains.
- 2.Hide the weakness of the hardware; flatter the user.
- 3.Borrow distribution from someone else's network.
- 4.Treat your personal network as infrastructure.
- 5.Wall off the cave.
- 6.Be dangerous enough to code, then find the real engineer.
- 7.Negotiate for autonomy, and price in its half-life.
- 8.Kill your signature constraint when the mission outgrows it.
- 9.Copy openly, then fix what the original got wrong.
- 10.Build the idea factory, not the ideas.
- 11.Try to prove yourself wrong.
- 12.End the chapter before it ends you.
K
The Decision Timeline
Eighteen years of choosing, refusing, selling, and leaving.
2005
Turns down Zuckerberg's offer at Zao Noodle Bar; chooses to finish Stanford.
2009
October 21: Burbn begins, a whiskey-named check-in app with too many features.
2010
March 5: $500K seed. May 19: Krieger joins. October 6: Instagram launches; 25,000 users day one.
2011
February: $7M Series A led by Benchmark. December: Apple's App of the Year.
2012
March 20: tells Twitter Instagram will "remain independent." April 9: sells to Facebook for $1B.
2013
November 1: first ad (Michael Kors). Instagram grows 23% while Facebook grows 3%.
2016
Algorithmic feed, new icon, Stories. The square and the chronology are both given up.
2018
June: one billion users. September 24: resigns on his first day back from parental leave.
2023
January 31: launches Artifact with Krieger; it shuts down a year later and Yahoo buys it.
2025
April 22: testifies in FTC v. Meta that Zuckerberg saw Instagram as a "threat."
Principle 1
Subtract until only the existing behavior remains
Burbn did not fail for lack of features. It failed because it had too many. Check-ins, plans, points, and photos added up to an app that "basically the number of hands that went up in the room" had used. What saved Systrom and Krieger was not some new insight about photography. They read the logs honestly. People were already doing one thing a lot, which was posting pictures, and nearly everything else was clutter.
This is subtler than "focus." Systrom did not pick photos because he loved photography, even though he did. He picked them because users were already sharing photos without being prompted, so the behavior existed before the product. He has said the same thing as an investor's rule: products that build on behavior people already have will beat products that try to create a new one. D'Angelo described the outcome from the buyer's side: "it turns out that people just like photos more than anything else."
Krieger made the point to Stanford students: the hardest part of a startup is finding the problem, not the solution. With Burbn they had a solution in search of a problem. Instagram was what remained after they removed every part that was solving their own problems instead of their users'.
Tactic: Rank every feature in your product by how much unprompted use it gets, and seriously consider shipping only the top one.
−
What Burbn Had, What Instagram Kept
The pivot was mostly deletion.
| Burbn feature | Fate in Instagram | Verdict |
|---|---|---|
| Location check-ins | Reduced to an optional geotag | Cut |
| Future "plans" | Removed | Cut |
| Points for hanging out | Removed | Cut |
| Photo posting | Became the whole product, plus filters | Scale |
| HTML5 web app | Replaced by a native iPhone app | Rebuilt |
| Social sharing | One-tap cross-post to Twitter, Facebook, Flickr | Scale |
Principle 2
Hide the weakness of the hardware; flatter the user
The filters came from a complaint. On a beach in Todos Santos, his girlfriend didn't want to post her photos because they didn't look good enough. The constraint was technical, early phone camera sensors, and the solution was emotional. X-Pro II and the filters that followed did not improve the pictures in a strict sense. They made the poor quality look like a choice. A Holga in Florence had taught Systrom that flaws handled stylishly come across as personality.
The strategic point is that Instagram did not wait for better phone cameras. It turned the camera's weakness into a reason to use the app. The user's ordinary snapshot looked like something made by a photographer with intent. People don't share tools. They share things that make them look good. Instagram's early growth was a vanity loop, and that same loop would later draw criticism over body image. The mechanism is powerful, and it does not have a built-in conscience.
Tactic: Find the biggest limitation your users struggle with today and design a layer that turns it into something they want to show off, instead of waiting for the limitation to go away.
Principle 3
Borrow distribution from someone else's network
From day one, Instagram let users post their photos to Twitter, Facebook, and Flickr. The New York Times found that early growth came because people saw Instagram links in their Twitter feeds and got curious. The app had no marketing budget. Its marketing was every user's existing audience on a larger network.
Chris Dixon's phrase "Come for the tool, stay for the network" names the two phases. The filters were free, which gave people a reason to download the app alone, before any of their friends were on it. Cross-posting turned each photo into an ad on another platform. Once enough people were on Instagram, the feed itself became the reason to stay.
This approach eventually turned back on him. By 2018, Facebook's cross-promotion, through notifications and links that sent its users to Instagram, was the borrowed distribution, and Zuckerberg could cut it off. The lesson cuts both ways. Use someone else's network to start growing, and plan for the day the owner notices.
Tactic: Build a one-tap export to the biggest network your users already use, and know in advance what your growth looks like if that export is shut off.
Principle 4
Treat your personal network as infrastructure
On launch night, as the servers went down, Systrom asked himself who the smartest person he knew was, and he called someone he had met over beers in red cups seven years earlier. Adam D'Angelo spent thirty minutes walking them through recovery and later helped hire, design databases, and invest. Jack Dorsey started at Odeo the same day Systrom began his internship there. Ronny Conway, a Google colleague, led him to Andreessen Horowitz. Matt Cohler carried Zuckerberg's first feeler. Instagram's entire capital and crisis structure was built out of relationships formed before there was any product.
Systrom has been unusually direct about this. "People are the connections that end up building your life," he told Stanford students in 2013. "They were really what helped us found Instagram." In 2011 his advice was simpler: stay after the talk and get to know the people around you. He got Ev Williams's attention with three cold emails and a WHOIS lookup. The network did not arrive by accident. He built it the same way he got the record-store job, by pestering.
Tactic: Keep an honest list of the five people you would call if your product failed publicly tonight, and invest in those relationships before you need them.
Principle 5
Wall off the cave
The first office had no windows, and Systrom chose to treat that as a benefit. He banned outside meetings except with the press and engineering recruits. He watched competitors for about a month, then stopped. "At the end of the day, you are your biggest competitor." When people called the photo market crowded and announced "Instagram killers," four people in a dark gray room kept shipping.
There is a disciplined version of this principle and an undisciplined one, and the difference is time. The ban on meetings protected the early period when product and team had to grow together without distraction. In 2016, at a company of hundreds with a billion-user parent, the same instinct would have been fatal. Systrom changed course then and copied Snapchat in plain view. The cave is a tool for one stage. It is not a permanent identity.
Tactic: For your first six months, refuse every meeting that doesn't recruit, inform, or ship, and write down the date when you'll reconsider that rule.
Principle 6
Be dangerous enough to code, then find the real engineer
Systrom pitched HTML5 as a vision when it was really a workaround: he did not know how to build a native app. The investors he won over saw through that and wrote the check anyway, on the condition that he find a cofounder. In his own phrase, he was "dangerous enough to code and sociable enough to sell." That was enough to produce a working prototype and to raise money, and not nearly enough to scale.
Krieger is what the founder brought in to cover his gaps: a Symbolic Systems graduate with real human-computer interaction training and engineering ability, who built most of the early product himself. Their partnership lasted from Burbn through Artifact, which is rare. Systrom later named mismatched expectations, about what you're building and how long you plan to build it, as the biggest cause of conflict between founders. Know enough to judge the work, admit where you fall short, and choose a partner whose strengths fill exactly that gap.
Tactic: Learn enough of the core craft to build a crude prototype yourself, then hire or partner with someone who will be embarrassed by it.
Principle 7
Negotiate for autonomy, and price in its half-life
Systrom reportedly agreed to Facebook's offer only after Zuckerberg promised to keep the brand and let the founders run the company. For about five years that promise held better than nearly any comparable acquisition in tech. Instagram had its own space, its own posters, its own marble table. Then the promise wore away, one decision at a time: Weil reassigned, Mosseri brought in, the Cox reorganization, branding stripped from reshared photos, cross-promotion cut, zero of three hundred.
Ben Thompson's critique is the most useful one: the departure traces back to the sale. Autonomy that the buyer grants is a policy, and policies change when circumstances change. The circumstance that changed here was success. The more Instagram grew, especially among young people, the more it looked like a threat to the main product. A founder who negotiates for independence should treat it as something with a half-life that runs inversely to how well things go.
The record also holds a smaller lesson. The February 2012 texts and the August 2012 testimony both became public. Anything you write during a negotiation is a document you may someday have to defend.
Tactic: When you sell, ask what your autonomy agreement would look like if your unit became your acquirer's biggest internal competitor, and negotiate for that scenario specifically.
⚖
The Promise and the Erosion
How the 2012 independence pledge played out by 2018.
| Domain | 2012–2016 | 2017–2018 |
|---|---|---|
| Brand | Kept separate, own space on campus Held | Instagram branding stripped from reshares Eroded |
| Reporting line | Effectively direct to Zuckerberg Held | Layered under Chris Cox Eroded |
| Leadership | Founders hire their own team Held | Weil reassigned; Mosseri brought in Eroded |
| Headcount | Grows from 13 to hundreds Mixed | Zero of 300 video hires; zero integrity hires Eroded |
| Distribution | Facebook cross-promotion fuels growth Held | Integrations cut back Eroded |
Principle 8
Kill your signature constraint when the mission outgrows it
The square was Instagram's look, and the chronological feed was how it felt. Systrom gave up both in 2016, and his reasoning is the most important thing he has said about product strategy: "companies that fail are typically companies that look at themselves as a set of features. Companies that succeed look at themselves as mission-based companies." If the mission is sharing the world's moments, telling a user to crop out a friend is, in his words, "a ridiculous argument."
The hard part is the timing. Constraints are often what give a product its identity early on. The square tied Instagram to the Instamatic and made a new app feel familiar. The founder has to recognize when a constraint stops shaping the product and starts limiting it. Systrom made that call in the face of real user anger over the algorithmic feed, and he didn't reverse it. The user numbers that followed suggest he called it right. The later criticism of "TikTokification" suggests that the next people to run the company took the same logic past the point of good taste.
Tactic: Write your company's mission in one sentence, list every feature constraint you consider sacred, and cut any constraint that would block a user from doing what the mission promises.
Principle 9
Copy openly, then fix what the original got wrong
When Stories launched, Systrom did not pretend it was original. He said Snapchat deserved credit, traced Instagram's own origins to Hipstamatic, Twitter, and Facebook's Like button, and argued that the whole industry works by remixing. Then he pointed to the improvements: you could pause and rewind, which the original didn't allow. Within eight months Instagram Stories had more daily users than Snapchat had in total.
Admitting it openly helped strategically. Denial would have made the story about copying. Admitting it made the story about execution. Copying without improving is just theft with good distribution. Systrom's version required finding the "really annoying" flaw in the original and fixing it. Underneath that is a cold calculation: when a competitor proves a format works, a larger platform with a better version will usually win, and the founder who refuses out of pride is protecting his self-image, not his users.
Tactic: When a rival's feature catches on, list its three most irritating flaws before you build your own version, and don't ship until you've fixed at least one.
Principle 10
Build the idea factory, not the ideas
The question new founders asked Systrom most often was how to scale their own time. His answer reflected having learned it the hard way: "Many projects failed because we founders didn't have the expertise, the time or both." The change came when they "learned to bet on our people," with the CEO's job being to keep everyone pointed at the mission and turn other people's ideas into reality. "If you try to be the factory that produces ideas, you'll always be constrained by your own time, ability, energy level, creativity. Instead, build that factory."
He also explained how he stocked the factory. His favorite interview question was about what candidates cared about outside work, because he had noticed a link between broad or unusual passions and creative thinking. One candidate told him about making salumi and drying meats in his basement. This is the same instinct that had him deejaying at sixteen and studying photography in Florence. He hired for interesting obsessions because his own had served him well.
Tactic: Every quarter, count how many shipped features came from someone other than you, and if the answer is close to zero, treat it as an organizational failure rather than a sign that you're indispensable.
Principle 11
Try to prove yourself wrong
Systrom's best line about decision-making is a statistician's: "I think a big part of knowing that you're right is working as hard as you can to prove that you're wrong. And if you can't, well, there's only one option left, which is you're probably right." The Burbn pivot came from exactly that, refusing to accept the founders' own story when the usage data contradicted it. The same reflex shows up on the witness stand, where he conceded that Instagram "could have gone either way" while still holding that the chance of failure had been low.
The habit matters most when you're winning. Rapid growth tends to confirm whatever theory the founder already had. Systrom's discipline was to test his conviction in the moments when it was easiest to skip the test. He also warned founders about advice from others, saying it "can go stale pretty quickly" and is "so often colored by someone's own experience," which applies the same skepticism outward.
Tactic: Before each major decision, assign one person to build the strongest case that you're wrong, and proceed only once you've actually answered it.
Principle 12
End the chapter before it ends you
Systrom's departure was messy in how it was handled and clear in what it meant. He compared a company to a rocket: aim it, add boosters, get it into orbit, "then take your hands off and let it go." With a billion users, revenue, and three offices, Instagram was in orbit. The fights over headcount and reporting lines told him that his inputs no longer mattered as much. He left on the first day back from parental leave, which suggests he did his thinking away from the building.
He has also been honest that he chose the timing in a way many people can't. "Sometimes those chapters are forced on us," he said at WIRED25, remembering people who told him the only difference between them was that they had been fired. Having the choice is a privilege, and a founder should use it while it's available. He stayed six years, longer than nearly anyone in Facebook's long line of acquired founders, then left before the company he had built became something he had to defend in public as its leader. Seven years later, when he did have to defend it, he was a witness, not the CEO.
Tactic: Write down now the specific conditions under which you would leave what you're building, and review that list every year as if a stranger had written it.
Free playbook
Get The Business Model Playbook
58 business models, one visual page each: how the money flows, the metrics that matter, and who runs it. Free when you join the Faster Than Normal email.
Free. No spam. Unsubscribe anytime.
Part IIIQuotes and Maxims
In their words
— Kevin Systrom, The Tim Ferriss Show, 2019I think a big part of knowing that you're right is working as hard as you can to prove that you're wrong. And if you can't, well, there's only one option left, which is you're probably right.
— Kevin Systrom, text to Matt Cohler, February 2012will he go into destroy mode if I say no
— Kevin Systrom, to BuzzFeed News, October 2016I think companies that fail are typically companies that look at themselves as a set of features. Companies that succeed look at themselves as mission-based companies.
— Kevin Systrom, Stanford Entrepreneurial Thought Leaders, 2020Sometimes the rocket you strap yourself onto wants to go in a different direction, and you have to compromise.
— Kevin Systrom, Seven Questions with Sequoia, 2019If you try to be the factory that produces ideas, you'll always be constrained by your own time, ability, energy level, creativity. Instead, build that factory.
For the fullest account of the founders' Facebook years, including the fights summarized in Part I, see Sarah Frier's No Filter: The Inside Story of Instagram, winner of the 2020 Financial Times and McKinsey Business Book of the Year Award.
Maxims
- Read the logs, not the pitch deck. The product your users actually love is usually hiding inside the one you built.
- Make the flaw look intentional. If you can't fix a limitation, give it a style people want to show off.
- Borrow the network, then build your own. Distribution from a bigger platform gets you started, and it belongs to someone else.
- Pester the gatekeeper. A record-store job and an internship at Twitter's predecessor both came from the third or fourth email.
- Call the smartest person you know. Relationships built over red cups are what keep you running at two in the morning.
- Autonomy has a half-life. Independence granted by a buyer shrinks as your success grows.
- No constraint is sacred. The square mattered until the mission needed more than a square.
- Copy openly, improve honestly. Admitting where an idea came from shifts attention from the theft to the execution.
- Hire for obsessions. A candidate drying salumi in the basement probably thinks unusually about other things too.
- Leave while it's in orbit. End the chapter yourself, because if you don't, someone else will decide when it ends.
