Today is Facebook’s 10th anniversary. On February 4th, 2004, Harvard undergraduate Mark Zuckerberg launched “Thefacebook.” A story in the Harvard Crimson a few days later was headlined “Hundreds Register for the New Facebook Website.” The opening sentence was “When Mark E. Zuckerberg ’06 grew impatient with the creation of an official universal Harvard facebook, he decided to take matters into his own hands.” Some 650 people had already joined, and thus began the company’s wild ride toward becoming a social networking site with over a billion users, thousands of employees, and a market capitalization well north of $100 billion.
As it happens, today is also the official publication date for Scaling Up Excellence by Huggy Rao and me. The book is all about the challenge of finding something that works really well in some corner of your world, and getting it to spread much further (or, for shorthand, “the Problem of More”). We were tickled when we realized the coincidence, because it wasn’t long at all after we started thinking about scaling that we had our first conversations with someone from Facebook. In the years since, we’ve learned a lot from people at that crazy place.
In early 2006, Diego Rodriguez and I started teaching a class we called “Creating Infectious Action” at the Stanford d. school. Sitting in the back was a woman named Katie Geminder, who said that she was head of product at a company called Facebook, that she had a 21-year-old boss, and that they were growing so fast they couldn’t lease buildings in downtown Palo Alto fast enough. They were desperately looking for great engineers, designers, and business people – but were very picky about who they hired. It was, she said, a demanding and rather quirky place to work. Katie left Facebook a couple years later, but not before she and her team sponsored a challenge for the Creating Infectious Action class. (The point of it: get using Facebook to catch on in some demographic group that wasn’t using it much in those days, such as people over 40, small businesses owners, or people who lived in rural areas.)
Of course, Facebook’s organization kept growing, so we kept watching it for lessons that might apply to other situations. Here’s the main lesson we took away from its success: Effective scaling isn’t just about establishing the biggest footprint you can, and as fast as possible. It’s even more a challenge of spreading the mindset your great new solution requires.
Facebook’s focus on spreading the right mindset became clear to us when, in late 2007 and early 2008, researcher and consultant Beth Benjamin and I had a series of conversations with Chris Cox, then Facebook’s 25 year-old head of Human Resources. Cox was employee number 30 and had joined Facebook as programmer; among other accomplishments, he helped create the “newsfeed” feature. But Mark Zuckerberg had asked him to take on the HR role (after trying a traditional HR head or two), because Cox embodied the priorities, skills, and beliefs he wanted to see take hold as the company grew. Watching Cox in new employee orientation meetings, we saw the wisdom of the move: given the scale that Facebook had the potential to attain, it was crucial that these early waves of hires share a set of beliefs and behaviors, and no outsider or non-technical person could instill them like he could. Cox’s HR leadership emphasized something that we later recognized elsewhere: It was essential to slow way down and find the right people, and to make sure that they learned to live the right mindset, in order to set the stage for speedy and effective scaling down the road.
Facebook’s method for spreading a shared mindset in new engineers and technical employees has become more systematic and exacting as the company has expanded. As Cox explained to Huggy and me in 2011, the company had figured out it needed to add an engineer for every 500,000 new users, so the pressure to make hires was relentless. Later, when we talked to Shona Brown, a former Google SVP who played a key role in growing Google from 2,000 to 30,000 people, we heard the exact same point. It came up again when we talked with Claudia Kotchka, who in her tenure at Procter & Gamble as vice president of design innovation and strategy, started with a tiny team and one innovation project and ended with over 300 innovation experts embedded in dozens of businesses
Facebook takes extreme measures to assure that newcomers live, understand, and commit to its distinct and shared beliefs about what is sacred and taboo. The people it hires, after rounds of grueling interviews, go through a six-week Bootcamp, which is led almost entirely by engineers (not HR). Bootcamp is designed so that newcomers immediately start living Facebook’s most sacred belief: “Move fast and break things.” Cox explained that the newbies start by working on actual Facebook software code knowing that their changes will be pushed live right away: “We tell them, put your hand on it. Grab it. Now bend it.” Cox said that a good definition of a successful first week at Facebook is that you’ve made a change to the site you can show your family and friends. For example, one new engineer reported that his dad called to say, “There’s a problem with this drop-down menu.” That engineer called back the next day: “I fixed it, Dad. Did you see that?”
Meanwhile, hesitating to act for fear of screwing something up is taboo. As engineer Sanjeev Singh explained, if you keep waiting for people to tell you what to do, and don’t ask for help when you get stuck, you aren’t going to last long at the company. Each newcomer is assigned a mentor—usually an engineer who isn’t a manager—to help him or her navigate Bootcamp. Pause to think of the investment being made here: in 2011, a new “class” of twenty to thirty hires was started roughly every two weeks—and seventy or eighty engineers at a time were pulled away from their jobs to be mentors. Obviously in a talent-hungry organization, those were brains that could have been deployed to pressing work. But leaders at Facebook are convinced that the time is better spent spreading the mindset that is essential to the company’s enduring success.
In short, the scaling lesson from Facebook and many other successful scaling efforts is that, to scale fastest and most effectively over the long haul, you can’t always be charging ahead at breakneck speed, grabbing up resources or territories and slapping your logo on them. There are times when you’ve got to slow things way down. You have to get the right people on board and make sure they share your mindset. This is especially true in early stages of scaling. The current struggles of Groupon, the once highflying “deal of the day” website, seem to prove the point in the negative. Our conversations with Silicon Valley insiders, as well as published reports and the former CEO’s letter to employees after he was fired, convince us that the mindset of focusing on customer needs failed to take hold throughout the company after it went public.
But it’s also true that, even in much later stages of scaling, you sometimes need to pause and regroup. Starbuck’s founder Howard Schultz lamented in a 2007 internal memo: “Over the past ten years, in order to achieve the growth, development, and scale necessary to go from less than 1,000 stores to 13,000 stores and beyond, we have had to make a series of decisions that, in retrospect, have led to the watering down of the Starbucks experience.” Schultz’s forthright book Onward details the steps his team took to “get their groove back” once he returned as CEO in 2008 and acknowledges how tough it was to reclaim and reinvent a once great mindset.
Finally, as much as we admire Facebook’s success in scaling, we don’t recommend that other leaders and organizations copy the particular mindset its leadership worked to spread. No one set of attitudes is right for every organization, or every part of the same organization. What is sacred in one place might best be considered taboo elsewhere. A VMware executive laughed when we asked if they lived by the philosophy of “move fast and break things.” His organization lives the opposite mindset in most parts of the company, especially the part that develops software for nuclear submarines!