No Filter: The Inside Story of Instagram

by Sarah Frier

Scooter Braun, Bieber’s manager, had seen this movie before. All the celebrities, Bieber the most internet-famous among them, had been posting their content on social media sites and getting nothing in return. Braun had discovered Bieber in 2006 when he was barely a teenager, singing on YouTube, but he hadn’t been as famous during the early rise of Facebook and Twitter. Maybe, Braun thought, he could get something out of Instagram. The star music industry negotiator called Systrom while the latter was in a station wagon full of friends, passing by Davis, California, on the way to Lake Tahoe. “Kevin, I’ve got Justin on the line,” he said. The two of them made a pitch: let Bieber invest, or pay him for his content. Or else he’ll stop using Instagram. Systrom had already decided that Instagram wouldn’t be paying anyone for their content, since he wanted everyone to be spending time on Instagram because it was fun and useful, not for commercial reasons. He said no to paying Bieber or taking his investment. Bieber followed through on Braun’s threat. But his on-and-off girlfriend, the Disney actress and singer Selena Gomez, loved to use Instagram, and their relationship was all the gossip blogs wanted to write about. Soon Bieber was back on the app, continuing to overload Instagram’s infrastructure, to the point where the company had to devote half a server just to his account’s activity.

“Don’t do that!” Zollman said. “If we start proactively reviewing content, we are legally liable for all of it. If anyone found out, we’d have to personally review every piece of content before it goes up, which is impossible.” She was right. According to Section 230 of the Communications Decency Act, nobody who provided an “interactive computer service” was considered the “publisher or speaker” of the information, legally speaking, unless they exerted editorial control before that content was posted.

Kutcher was nothing like the dopey characters he played. He absorbed everything he could about the technology industry, resolving to use his trend-spotting skills in a more lucrative way. He worked with Guy Oseary, Madonna’s manager, to sort through all the opportunities, and ended up giving money to dozens of companies—not just in social media—including Uber, Airbnb, Spotify, and Instagram competitor Path. “Whenever there was a new type of experience for consumers, there would be like three companies doing the same thing,” he remembers. There were several versions of Instagram, Pinterest, and Uber. “Who would get traction first? And then the network effect would take all.”

Kutcher was put off by Systrom’s lack of flexibility in the face of a good idea. But he was still intrigued enough to invite him, with their mutual friend Joshua Kushner, on a ski trip in Utah with other technology founders. A half dozen men stayed there overnight in a large cabin in the snow.

The number the deals team had in mind for Instagram was around $80 million. Costolo said that $80 million was far too expensive for such a young company that was just a better way to share photos within Twitter. It was unlikely to compete with Twitter on being a resource for news and updates from public figures. Others in the room guessed there was another element to his thinking: if the deal came to pass, Dorsey would get credit for it. The discussion was tabled.

That same month, Krieger was invited to be Michelle Obama’s guest at the State of the Union, to explain that he wouldn’t have been able to help start Instagram without an immigration visa.

The technical tools to manage accounts were still scrappy—or nonexistent. One day that winter Scooter Braun sounded an alarm to Instagram: Bieber was locked out of his app. But Instagram didn’t have a reliable system for resetting passwords. They told Braun they could do it over the phone, but Bieber would need to verify his identity. “All right,” Braun said, “Justin’s going to call you guys.” Richardson picked up. “Hey, this is Justin,” he said. There were no security questions prepared, so that declaration would have to be enough proof of identity. She reset his password over the phone.

Only Zuckerberg understood what would appeal to Systrom: independence.

Armed with those experiences, Zuckerberg thought he knew how to talk to Systrom, founder to founder. Systrom didn’t want to run a Facebook product, just like he didn’t want to run a Twitter product. He wanted to keep his company, and to keep being the Instagram visionary, just with none of the risks of independence.

Instagram didn’t make money yet, but Zoufonoun surmised that because the Instagram product provided its users with the ability to endlessly scroll through posts, just like the Facebook news feed, they could eventually develop the same kind of advertising capabilities. They could use Facebook’s infrastructure to grow faster, the way YouTube did at Google.

The discussions continued at Zuckerberg’s sparsely furnished $7 million home that Saturday.

If he valued Instagram so highly, with its tiny team and no revenue, he would start a bubble in Silicon Valley, raising the price of every related company he might want to buy in the future. (He was partially right. In 2013, venture capitalist Aileen Lee came up with a name for startups with billion-dollar values: “unicorns.” At the time, there were 37. When she wrote an update on the rare breeds in 2015, there were 84. By 2019, there were hundreds. But if it’s a bubble, it hasn’t burst.)

Zoufonoun stood next to Zuckerberg as he tended the meat, smoke billowing out of the grill. Beast looked on intently, then started growling. Suddenly the dog ran 20 feet, his corded coat flopping in the wind, lunging to attack Zoufonoun’s leg. “Shit!” Zoufonoun exclaimed. “Did he break skin?” Zuckerberg wanted to know. “Because if he did, you know, we have to record it. And they might take him away.”

Systrom gave four reasons. First, he reiterated Zuckerberg’s argument: that Facebook’s stock value was likely to go up, so the value of the acquisition would grow over time. Second, he’d take a large competitor out of the picture. If Facebook took measures to copy Instagram or target the app directly, that would make it a lot more difficult to grow. Third, Instagram would benefit from Facebook’s entire operations infrastructure, not just data centers but also people who already knew how to do all the things Instagram would need to learn in the future. Fourth, and most importantly, he and Krieger would have independence.

The structure of the Instagram acquisition—a company purchased not to be integrated—would become an important precedent in technology M&A, especially as giant companies got even gianter, and small companies like Instagram wanted to find some alternative to competing with them or dying.

Google would buy Nest, keeping it separate. Amazon would buy Whole Foods, keeping it separate. And so many corporate development teams would court startups, promising to “do it like Instagram,” only to change their minds about granting independence once everyone was in the building.

Instagram’s perceived independence at Facebook would help Zuckerberg win some otherwise impossible deals with headstrong founders, especially in 2014, with the chat app WhatsApp and the virtual reality company Oculus VR.

Imagine an alternate reality, in which Microsoft buys Apple while Apple is still small. That would have been tremendous for Microsoft. And that’s what Facebook got with Instagram.

“If we don’t create the thing that kills Facebook, someone else will. The internet is not a friendly place. Things that don’t stay relevant don’t even get the luxury of leaving ruins. They disappear.”

The morning after Systrom signed the contract, Dorsey was on his way to work at Square, the payments company he’d cofounded. Despite his wealth, he always enjoyed taking public transportation to absorb the culture of San Francisco. That morning he noticed that he’d have the whole Route 1 Muni to himself. “A simple morning pleasure: an empty bus,” he posted on Instagram, with a picture of the brown-and-tan seats of the car, unfilled, and not even the driver visible. He’d been posting once a day, sometimes twice a day when inspiration struck, about Square and sunsets, coffee, and air travel. Despite Instagram’s recent rejection of Twitter’s overtures, Dorsey had only become more invested in the app’s success after helping a couple investor friends get in on their latest funding round.

In theory Dorsey would be richer too, as one of Instagram’s earliest investors, but all he felt was sadness. He couldn’t stop thinking about Systrom. After all his advice and support, he’d thought they were friends. Why hadn’t he called, even just for business reasons?

As time passed without any explanation from Systrom, Dorsey stopped feeling hurt and started feeling angry. He realized Systrom had never wanted to sell to Twitter. Twitter had been played. Dorsey deleted the Instagram app and stopped posting altogether.

The headlines imagined what their day might be like, just based on the numbers. “The 13 employees of photo-sharing service Instagram are celebrating today after learning they are set to become multi-millionaires,” the Daily Mail wrote. “Instagram is now worth $77 million per employee,” The Atlantic reported. Business Insider published a list of all the employees they could find, complete with their photos and information scraped off the internet about what schools they went to and where they worked before. Team members were fielding calls and Facebook comments from their friends and family, congratulating them on having made it in life. But had they made it? They wouldn’t have any answers to their money questions for another couple of weeks.

One of Systrom’s closest friends, the venture capitalist Joshua Kushner, was able to get his firm, Thrive Capital, invested in the latest funding round. Like everyone else, Kushner doubled his money in record time and made a name for himself in the process. So Kushner asked his sister-in-law, Ivanka Trump, to make sure the employees enjoyed themselves. Everyone got to stay in suites in the gold-windowed Trump International Hotel, where they received little personal notes from the heiress, congratulating them.

On the team’s return from Vegas, the first news was that Facebook would not be able to help them with resources or infrastructure until the deal was actually approved by regulators. That could take many months, according to Facebook’s lawyers. The governments in the U.S. and Europe were investigating whether buying Instagram would give Facebook monopoly powers. Until then, Instagram couldn’t be at Facebook’s headquarters and wouldn’t be able to do much hiring, so they would remain just as overworked.

Silicon Valley employees often decide to take lower salaries in order to work at a startup like Instagram that offers stock options—the option to purchase shares cheaply at a later date. Those options are restricted based on time. Usually a quarter of the total grant in the job offer becomes available after each year that they continue working, giving an incentive to stick around. For an employee who picks a winning company, the small slice of ownership yields life-changing wealth, like winning the lottery.

Frustrated, Zollman and some of the others took a trip to Santa Monica together without the cofounders to get their minds off it, but found themselves in a sort of group therapy session instead. If Kevin had just given each of us one of his millions, they’d say to each other, we wouldn’t have to rent apartments anymore. We could invest in startups, or start our own. They asked their friends and learned that it wasn’t uncommon for startup founders to distribute life-changing sums to their employees after a big deal.

Antitrust law was not written for modern acquisitions like Instagram. A traditional monopoly was a company with such a hold on its industry that it harmed others by fixing prices or controlling a supply chain. Facebook and Instagram presented no obvious consumer harm because their products were free to use, as long as people were willing to give up their data to the network. Facebook’s advertising business was relatively new, especially on mobile phones; Instagram didn’t have a business model at all. Something was a monopoly if it undermined its rivals; Instagram had many rivals. Instagram wasn’t even the first company to make a mobile photo app with filters.

First, regulators needed a clear picture of what Instagram thought of Facebook and vice versa, based on internal emails and text messages. Oddly, the FTC would not be gathering this documentation itself. The lawyers for Facebook and Instagram—the same ones who had worked on the deal—were now tasked with finding any evidence showing that the deal shouldn’t go through. They were paid by the companies to investigate the companies.

The real value of Facebook and Instagram was in their network effects—the momentum they gained as more people joined. Even if someone enjoyed using an Instagram competitor like Path more, if their friends weren’t on it, they wouldn’t stay. (Path shut down in 2018 after selling to a South Korean company, Daum Kakao, three years before.) Zuckerberg understood that the hardest part of creating a business would be creating a new habit for users and a group they all wanted to spend time with. Instagram was easier to buy than to build because once a network takes off, there are few reasons to join a smaller one. It becomes part of the infrastructure of society.

Even Chris Hughes, one of the cofounders of Facebook, would in 2019 call for the deal to be undone. “Mark’s power is unprecedented and un-American,” he wrote in the New York Times.

Instagram needed to sell to Facebook because Systrom and Krieger had been slow in hiring. They were so particular about picking employees who would be a perfect fit, despite being so frenzied with keeping the site alive.

They were exhausted. Selling was the simplest way to solve the problem. Facebook had more than 3,000 employees—some of the smartest engineers in the world. Once Instagram joined them, if the FTC allowed, they’d be able to recruit from within. They would just have no relief in the meantime. Their investments in employees and infrastructure were on hold, pending the deal close, at a time sign-ups were accelerating. As usual, Krieger’s erratic sleep schedule was the surest sign of Instagram’s unending expansion.

Jack Dorsey, meanwhile, was grappling with the fact that he’d essentially been the first Instagram influencer, selling Instagram itself.

It was called a “fairness hearing”—a rarely used option California offers to companies with uncomplicated deals, so that they can issue stock with the approval of state regulators instead of going through the longer federal process.

If Facebook’s employees were overindulging in the middle of the week, it was because they needed the boost. Employee morale tracked closely with the stock price.

Threats to and opportunities for Facebook’s product were evaluated with the same depth of analysis as everything else. Facebook had access to data that tracked how often people were using different apps on their smartphones. The data acted as an early warning system for a potential competitor’s rise. If there was any chance Facebook could build its own version of the app that might ultimately reach more people, they would try, immediately. If it didn’t work, that’s where acquisitions like Instagram came in handy.

Facebook’s strategy for giving people what they wanted would be accused of addicting the world to the digital equivalent of junk food. Its data collection would spark further panic over privacy. But for now, with the stock down in an era before the public reckoning, Facebook was singularly focused on demonstrating that it could create a viable long-term business, even on mobile phones, proving all the haters wrong.

Instagram wanted things to be carefully considered and designed before they were released to people. Humans, not numbers. Artists, photographers, and designers, not DAUs, the Facebook term for “daily active users.” They didn’t want to limit people to their likes and dislikes; they wanted to introduce them to things they’d never seen before.

Over the years, Facebook had learned that people would get mad about breaches of privacy and then forget about them because they actually enjoyed what they were seeing—after all, users were getting exactly what Facebook thought they wanted, based on their previous behavior. Usually, people calmed down.

The only real consequence, so far, had been a settlement with the U.S. Federal Trade Commission, which said the company had to get users’ express consent before siphoning off a new kind of data.

“This shirt costs more than your car,” Systrom retorted, ready to fight in defense of fashion as art, before onlookers dragged the men into the meeting, Boz rolling his eyes, thinking Systrom was either arrogant or insecure or both. The shirt was from Gant, a men’s boutique for yuppies. Boz drove the 10-year-old Honda Accord parked outside.

Exactly what kind of authority Systrom and Krieger had at Facebook was unclear. They came in with regular employee ranks, as a product manager and an engineer, respectively. Systrom reported to Mike Schroepfer, recently promoted to chief technology officer, while his transition was managed by Dan Rose, Facebook’s head of business development.

At Facebook, product requests were ranked by priority number, with ones and zeroes being top priority. The only thing above that priority level, superseding anything else on the road map, was unofficially called a “ZuckPri,” which meant that Zuckerberg was tracking the progress. Photo tagging on Instagram was a ZuckPri. It had been such a boost for Facebook in its early days, he was sure that it would work for Instagram.

Facebook saw no reason for Instagram photos to continue to display in Twitter posts. The feature that helped the app catch on, through filtered photos displayed by Jack Dorsey, Snoop Dogg, Justin Bieber, and others, was also creating posts, for free, that Twitter could advertise around—not Facebook.

There were no lawyers among the early employees, so when the startup created its first “terms of service,” they simply copied and pasted some boilerplate language from the internet, and then edited it to be Instagram-specific until it looked right. As a public company, Facebook had standards that were a tad higher. In December, Instagram accepted their counsel’s edits adjusting the language for the new era, and for a future that might include making money and sharing information with Facebook.

Mostly, they had completely underestimated how much their users would mistrust—and even hate—Facebook. The angry tweets made it clear the Instagram community was looking for signs that the acquisition had ruined the app forever.

But the same hacker mentality that was rewarded at Facebook caused clashes among the expanding Instagram team. Former Facebookers would give obvious ideas to boost activity, like adding a re-gram button, and then original Instagrammers would reject them, saying, “That’s not how we do it here.” Instagrammers would explain the charm of InstaMeets or discuss a plan for highlighting the Albuquerque International Balloon Fiesta on the @instagram account, and some former Facebookers would roll their eyes.

The biggest was “community first,” meaning all their decisions should be centered around preserving a good feeling when using Instagram, not necessarily a more fast-growing business. Too many notifications would violate that principle.

Then there was “simplicity matters,” meaning that before any new products could roll out, engineers had to think about whether they were solving a specific user problem, and whether making a change was even necessary, or might overcomplicate the app.