5

How Facebook Grows . . . and Grows

Ben Mezrich’s Accidental Billionaires painted the early days of Facebook as a story of looking for one thing—acceptance in Harvard’s rarefied Final Clubs—and bungling into another—an ascendant global social network.

Even as the book was coming out in 2009, however, behind the closed doors in Palo Alto, Facebook was anything but accidental. They were deep into becoming one of the best deliberate growth engines of all time. Facebook’s success was not manifest destiny. Not “build it and they will come.” It was, from the beginning, a conscious pursuit of constant growth. (See Figure 5-1.)

Figure 5-1. Growth of Facebook’s monthly and daily active users globally (millions)¹

And that pursuit wouldn’t get difficult at 1.5 billion people, or 1 billion, or even 500 million. It would get difficult much, much earlier than that.

Hitting a Wall

Everything began, of course, at Harvard on February 4, 2004. Using the three-decade-old medium of e-mail to bootstrap their brand-new medium, Facebook’s founders invited friends and sent messages to campus mailing lists—the predominant way to form digital groups at that time—to join their new creation, The Facebook, to connect with their friends, to see who else was in their classes and to stay in touch. After new users had signed up for Facebook, they would then be encouraged to invite their friends. In the first week, about half of Harvard’s undergrads had registered.

Opening to Harvard was followed carefully by opening to Columbia, Stanford and Yale later that February. The three were chosen not only to carefully manage the process of increasing the size of Facebook without breaking the team’s nascent infrastructure but also because the three campuses had some of the most established local social networks at the time. If they could win there, the fledgling crew really might be onto something.

Cornell, Dartmouth, University of Pennsylvania, MIT, Boston University and New York University followed in March, and then Brown, Princeton, UC Berkeley, Duke, Georgetown, University of Virginia, Boston College, Tufts, Northeastern, University of Illinois, University of Florida, Wellesley, University of Michigan, Michigan State, Northwestern, UCLA, Emory, University of North Carolina, Tulane and University of Chicago came on board in April.

At that time, cross-school friend connections were implemented—someone from school A could invite someone from school B—which were a major factor in disrupting any established but highly local social networks. We’ll see the importance of this “exogenous” mechanic repeat itself at a much larger scale a little later in the story.

Three months after launch, Facebook had a total of 30 schools and 100,000 members.

The rest of U.S. college campuses followed throughout 2004, and the year ended with Facebook having a million users.

In September 2005, Facebook opened to U.S. high schools and to 20 universities in the UK, and by the end of 2005, it was available at 2,000 colleges and 25,000 high schools in the United States, as well as universities in Canada, Mexico, the UK, Australia, New Zealand and Ireland.

By September 2006, Facebook had grown to about 10 million users and had made the momentous decision that could have ended it all—my mom is on Facebook?—to open Facebook to everyone.

With a great sigh of relief, a year later Facebook had grown to 50 million users.

And then they hit a wall.

Getting the Band Together

With growth stalled in late 2007, some very large questions faced Zuckerberg and the team. Why are we stalling? Why had even MySpace (remember, in 2007, Facebook was not yet the leader) been stalled around 100 million users for the better part of six months?

Was there a chance social networking just isn’t interesting to more than 100 million people?

There were enough questions that everyone agreed a team had to be put in place to redouble the deliberateness of Facebook’s growth.

The team’s leader would be brash Sri Lankan–born, Canadian-raised, recovering investment banker, product manager, venture capitalist and self-declared merchant of progress Chamath Palihapitiya. The 31-year-old had just joined Facebook with relevant experience working on small teams for products with large user bases, including music application WinAmp and AOL instant messaging products AIM and ICQ. Palihapitiya’s leadership style blends an often-wrong-but-never-in-doubt confidence that hovers just this side of arrogance with an intense do-the-work pragmatism that proved just the right cocktail for Facebook at the time. His team would look-try-measure-repeat their way to success, and Palihapitiya had direct lines to Zuckerberg and Facebook’s engineering, infrastructure and product leadership to get the necessary help.

The team’s international conscience came from Spanish surfer and recently minted Stanford MBA Javier Olivan. Brought in by Palihapitiya in late 2007, the 30-year-old Olivan had founded a small social network in Spain and came with an engineering and product management background from the European wireless telephone industry, which was growing faster than the American mobile industry at the time and gave him important background for the mobile wave ahead.

Their analytical mind would be the intense, Cambridge-educated Internet marketing savant Alex Schultz. Arriving at Facebook in November 2007, Schultz at age 25 already had 10 years of experience in Internet marketing, having gotten his start with search engine optimization in the pre-Google days for his personal website paperairplanes.co.uk, which he maintains to this day as his personal workbench for the kinds of marketing that he and his team do at a global scale for Facebook. After Internet marketing stints for eBay in the UK and United States, Schultz had the disciplined foundation of picking a target and optimizing for it, from which an even more sophisticated growth engine could evolve. It didn’t hurt that he combined passion and urgency in a way perhaps best summed up by one his favorite quotes—from General George S. Patton no less:

A good plan, violently executed now is better than a perfect plan next week.

Finally, the team’s product and people sensibility—a perfect yin to Schultz’s analytical yang—would be provided by the quiet, almost elfin, Naomi Gleit. At 22, Gleit had been so convinced of Facebook’s eventual success before they even had a million users that in her final year in Stanford’s Science Technology and Society program she wrote a paper on how Facebook had beaten local social network Club Nexus and then proceeded to hound Facebook for a job until she got one, starting in 2005 in one of the first nonengineering roles at the company. Having volunteered in Botswana, taught on a Navajo reservation and lived in a Buddhist monastery in Thailand, she was deeply connected to her team’s mission to “remove barriers and make Facebook available to everyone in the world.”

With Palihapitiya, Olivan, Schultz and Gleit at the core, the team—which would grow to well over 100 people over the years—would go about attacking the wall.

The Math of Growth

Attacking walls is not glamorous work. Not work that is often visible. Not work that even works much of the time. Schultz describes the daily, never ending effort this way:

If you can run more experiments than the next guy, if you can be hungry for growth, if you can fight and die for every extra user and you stay up late at night to get those extra users, to run those experiments, to get the data, and do it over and over and over again, you will grow faster.

But before we crack open the doors on the specifics of what the team has done over the years, what is growth?

You don’t need to know calculus to understand the math of growth. It is as simple as it is unforgiving. For a service like Facebook, the number of active users in any particular period t—a month, for example—is the following:

Active_t = New_t + Retained_t−1 + Resurrected_t−[n] − Churned_t−1

Although your users make up one large group, they are composed of four very different cohorts:

New: People who joined for the first time in the current period via acquisition and activation techniques. They have the least experience with the product and need to be nurtured carefully.

Retained: People who used the product in the current period and the prior period, a strong sign of engagement and perceived product value. By far the most important—and hopefully largest—cohort.

Resurrected: People who used the product at some time in the past—but not in the prior period—and have returned in the current period through reacquisition techniques. An audience that hangs in the balance as they may be as likely to depart for the long term as they are to stay. What can be done to keep them?

Churned: People who used the product in the prior period but not in the current one, signaling a lack of perceived value or loss of access. (On a related note, it’s a little known fact that Facebook routinely purges what it deems to be inauthentic accounts from Facebook, which is a significant user base price to pay—it has accounted for tens of millions of removed users over time—in order to maintain authentic identity across the service.)

Although the math may be simple, the moving pieces of cause and effect can quickly drown companies, so the Facebook team broke growth into three simple factors:

The North Star Metric: It’s one thing to have data, it’s another to be overwhelmed by it. To avoid endless navel-gazing or too many people chasing too many goals, Facebook selected a single metric that would be the subject of all their growth attention. The lingua franca for all teams—analytics, product, engineering and marketing—chipping away at The Wall.

For a company like AirBnB, this might be the number of nights booked on the service, for Uber the number of rides, for WhatsApp the number of messages sent, for eBay the amount of gross merchandise volume. It is the number that is to your service as gravity is to bowling. More inescapable and fundamental to success in the long term than any other.

For Facebook, that number would not be a Web 1.0 metric like page views or registered members. It would be engagement. The number of people who found enough value to use Facebook on a regular basis.

Over the years, Facebook would become increasingly sophisticated—and exacting—about the nature of engagement as they went from monthly active users (MAU) to daily active users (DAU), to the ratio of daily active users to monthly active users (DAU/MAU, a ludicrous 65% globally even to this day), to understanding the distribution of the frequency of people’s usage over the past 28 days, to the number of people who used Facebook seven out of the past eight days (L7/8). They meticulously segmented their users into vintages (how long they had been on Facebook), engagement cohorts (how frequently they used it), demographics (their gender, age and location), methods and speed of access (desktop, smartphone, tablet, feature phone, 2G, 3G, LTE, broadband) and even psychographics (their actions and interests). They then carefully studied the differences and similarities in engagement among these cohorts, putting a special concentration on the “marginal” user who was not—yet—consistently engaged with Facebook.

When we talk about big data, this is what we mean.

The Magic Moment: It’s good to have clarity on the outcome you’re hoping to achieve, but what is the difference-maker that gets you there? The experience that gets people hooked on a product? The single most important thing that, if done well, causes you to grow and, if done poorly, destines you to irrelevance?

For Uber, the magic moment is the first time you push that button and a car shows up. For WhatsApp, it’s the first message you send—often internationally—that doesn’t incur an SMS or roaming charge. For AirBnB, it’s that surprisingly lovely place with character in another city that doesn’t feel like a hotel.

For Facebook, it was seeing your friends in your News Feed, so all efforts were aimed at getting you to that moment as fast as possible. Once it was clear how crucial this was to Facebook, it was all hands on deck to get you to seven friends within ten days of joining Facebook, then to ten friends in two weeks and to fifty friends as fast as possible after that. Nothing was more predictive of long-term engagement with Facebook—independent of your demographics or psychographics—than helping you make these connections as quickly as possible.

The Core Product Value: After getting people hooked via your Magic Moment, you have to deliver the day-in-and-day-out value that earns loyalty from your users. For Facebook, that is the feeling of connectedness people so consistently point to about the service, which is delivered most importantly by a great experience in the Facebook News Feed. Thank you, Chris Cox.

The rest was just blood, sweat and tears.

Inside Facebook’s Growth Engine

Armed with their North Star Metric, Magic Moment and Core Product Value, the team got down to business. Under Palihapitiya’s direction, there would be no ego and no lore. Just hard work and data: (1) get people to the front door, (2) have the Magic Moment as quickly as possible and (3) offer core product value as often as possible. Schultz and his team would focus on all things related to the front door—acquisition and resurrection and the data behind conversion and retention, Gleit on activation and the speed-to-Magic-Moment, and Olivan on one of the most significant product value evolutions in Facebook’s history.

When it came to acquisition, Schultz’s team wouldn’t just use best-in-class search engine optimization and search engine marketing with significant budgets; they would broker a landmark deal with Google to allow the search engine giant to use its automated tools to read, store and make searchable the public parts of Facebook profiles—basics like your name and profile picture, which were not behind the walls of Facebook’s privacy controls. This ensured that whenever you typed the name of a friend into Google, you’d be made aware—the vast majority of the time in the top search result—if they were on Facebook, driving increased interest in joining yourself.

Activating you on Facebook entails everything from the moment you first arrive at the home page to the process of signing up and your early moments on the site. The New User Experience—known affectionately as NUX inside Facebook—included the painstaking work of nailing the first few crucial moments of sign-up by building simple pages that depended on the careful testing—including eye-tracking studies—of every word, button, color and even page load speed. Every confusing direction, extra click and fraction of a second wasted threatened to destroy the work of getting a new user to this point.

Then came an important moment of truth on the way to the Magic Moment: importing your contacts from other services, especially e-mail providers like Hotmail, Yahoo, AOL and Gmail, to make connecting with friends of yours already on Facebook—and inviting those that were not yet—much more effortless. Contact importing came with a heaping of political intrigue as services like the e-mail providers had to allow Facebook access to retrieve users’ contact lists through tools like application programmer interfaces (APIs), a classic example of co-opetition. In 2010, Facebook even acquired contact-importing technology provider Octazen who had a reputation both for building useful tools to help import contacts for the over 30% of Facebook’s users who were not on one of the big e-mail platforms, as well as operating in gray areas obtaining contact information from public web pages of other services without using provided—and controlled—access methods, known as “scraping.”

To understand the importance of contact importing to Facebook’s growth, take a look at the math involved in creating virality in a business with so-called network effects like Facebook. To grow exponentially, you have to be able to translate one new user into a little more than one additional new user. If U is the original user and UF are the user’s friends, it looks like this:

Invites_U × E-mail click rate_UF × Facebook sign-up rate_UF > 1

Here’s an example: a user sends 100 invites to friends, those friends click on the invite 30% of the time, and those clickers complete the Facebook sign-up 5% of the time:

100 × 30% × 5% = 1.5

Presto, that first user is responsible for bringing 1.5 new users to Facebook.

Contact importers ensured that the first number in this equation was as high as possible. The way invite e-mails were composed controlled the second number, and the simplicity of the Facebook sign-up flow controlled the third. Even small changes in the individual components have a large compounding effect on the eventual outcome. Taking the same example a few steps further, if it went on with similar numbers for three “generations” of new users, it would look like this:

(100 × 30% × 5%) × (100 × 30% × 5%) × (100 × 30% × 5%) = 3.4

The first user would actually be responsible for bringing 3.4 people onto the platform. But if the number of invites in each case had been just 80 instead of 100 (or Facebook’s sign-up conversion had been just 4% instead of 5%), it would have been just 1.7 people. Half as many. Powerful stuff!

For resurrection, Facebook relied on an old medium—e-mail—and a much older human motivation: curiosity. Resurrection targets were already users of Facebook with existing connections to friends, so sending the occasional e-mail to them about what their friends were doing—that new wall post or photo tag—created reasons to swing by Facebook if you hadn’t in a while.

By using e-mail for friend invites and resurrection, Facebook became a master of using the medium they were replacing to feed the new medium they were creating.

More important than all of that, of course, was retention. Getting you beyond the first Magic Moment quickly and fortifying your engagement by delivering Core Product Value. To extend the contact importers’ role in getting you to seven friends in 10 days, 10 friends in two weeks and 50 friends as soon as possible, in early 2008 Gleit and the team introduced the People You May Know feature—known internally as PYMK—that resided to the right of your growing News Feed. Pulled from a sophisticated combination of information from your imported contacts, your friends’ friends with whom you share the most friends and commonality of work, college or interests, Facebook could determine a “friend distance” that allowed them to rank people you were not yet friends with but likely would want to be, and surface the most likely candidates for you. Conversely, as new friends joined Facebook and connected with you, Facebook would reach out to you to encourage you to suggest additional friends for the new user.

At the same time, Olivan and a team had been working on another big opportunity to deliver Core Product Value: translating Facebook into languages beyond English to increase the site’s usability and appeal internationally. This wasn’t as simple as translating the site into half a dozen common languages. It wasn’t just that in countries like Germany, there were only 600,000 Facebook users even though there were over 40 million Internet users. There were hundreds of millions of Internet users in countries beyond the top 10 global languages.

Instead of moving a team of dozens into each prospective country and translating the site themselves, as MySpace was doing, or using just professional translations services, the Facebook team built an application on top of Facebook’s own platform that allowed the site to be broken into its 300,000 component words and phrases and localized in a crowd-sourced way by letting target groups of Facebook users suggest—and vote on—local translations that would then be overseen by professional translators.

Using the technique, Facebook was able to launch its first translation—in Spain—in just two weeks, its German translation in two days, and its French translation in just 24 hours.

By the middle of 2008, they had 16 translations that would eventually grow to 20, overseen by professional translators and an additional 80 just crowd-sourced, creating loyalty from traditionally underrepresented communities such as Basque, Cherokee and Hausa.

The translations were a major contributor to a growth phenomenon Facebook had seen since its earliest days: early adopters of Facebook in a particular country would make the majority of their initial friend connections to other countries that had a more established Facebook user base. These were known as “exogenous” connections and implied a stronger comfort with non-native languages on the part of the early adopters. As translations became available in-country, the site became more inviting to primarily native speakers—the majority of that country’s population—and as friend connections made between people continued to grow, the majority of those connections became “endogenous”—between people in the same country. In country after country, when that crossover from exogenous to endogenous connections came, Facebook user growth would kick into a new and sustained gear, signaling an acceptance beyond early adopters that cemented country-level engagement and loyalty over the long term.

The outcome of all the team’s efforts in its first six months was much more than just interesting sociological phenomena: according to com-Score, by June 2008—on the back of strong international results—Facebook had grown to 124 million users, accelerating past MySpace’s global user base for the first time. A year later, in June 2009, Facebook finally exceeded MySpace’s user base in that company’s final stronghold, the United States, where Facebook had doubled in size to 70 million users in just one year and MySpace had declined by 5%.

Eighteen months of intense focus by the growth team had put MySpace behind them and proven that social networking is interesting to way more than 100 million people.

The Importance of Being Mobile

But if Facebook’s sine qua non is retention and engagement, there would be no more important enabler—or inhibitor—of growth than the emergence of mobile as the world’s favorite medium, which happened practically in its entirety within the lifetime of Facebook’s growth team as less than 1.5 million iPhones had been sold by the time of the team’s formation in late 2007.

In a world dominated by mobile, Facebook would have to become a daily—and preferably hourly—habit since, according to Forrester, people spend a giant 84% of all their time with apps in just five apps.

A simple first step of offering a mobile-optimized website at m.facebook.com came in 2007 and was followed throughout 2008 with native Facebook apps for Apple iOS, Blackberry, Windows Mobile and Nokia’s Symbian, and finally in September 2009 for Google’s Android, which had proven a more complex effort given the wide range of hardware and operating system versions for which an Android app has to work.

By February 2010, 25% of Facebook’s 400 million monthly active users were accessing it over mobile, making it a good time to turn a much keener eye to how people engaged with Facebook in emerging markets where oftentimes broadband landline infrastructure had been skipped entirely in favor of a mobile one but where purchasing power lagged, making low-end phones and limited cellular speeds and data consumption prevalent.

In May of 2010, Palihapitiya announced the launch of Facebook Zero, a minimal text-only version of Facebook’s service that would not accrue any cellular data charges on the networks of the 50 operators in 45 countries the service launched with. The intention was to extend the feeling of the basics of being connected via Facebook to many more people, some of whom would—having tasted some the value of Facebook and the Internet—steer more of their limited discretionary income toward cellular data plans with the providers offering Facebook Zero. Uptake of the service was rapid in certain regions—it is often credited with doubling the number of Facebook users in Africa over the course of only 18 months. Over time, however, the “zero rated” pricing of the service began to run into regulatory concerns around so-called net neutrality—the principle that all Internet traffic must be treated equally by operators—causing the offering to either be tightly restricted in duration or ended altogether. We’ll see more about this complex issue in the story of Internet.org in Chapter 14.

About a year after the launch of Zero, Facebook addressed a different element of lower-end mobile infrastructures when Gleit traveled to Tel Aviv to acquire Israeli company Snaptu in March 2011. The company had developed a way for so-called feature phones—phones that had some data and messaging capabilities but lacked the sophistication of smartphones to run web browsers or apps—to access a basic version of Facebook. With the added functionality of Snaptu, Facebook was able to reduce the fraction of phones they could not address down to just 20%. After renaming the service Facebook for Every Phone, they grew it to 100 million monthly active users—nearly 10% of all Facebook users at the time—in the 18 months following the acquisition.

In June 2015, five years after launching their first emerging market mobile offering, Facebook combined all of its accumulated knowledge and technology into what will likely be the most important long-term offering for these markets, Facebook Lite.

Aimed at so-called typical Android phones that, according to analytics firm IDC, account for about two-thirds of the global smartphone market but have less performance and are connected to slower networks than your phone and mine, Lite uses a sophisticated combination of a lightweight app on the phone—a mere 1 MB, which is one-hundredth the size of Facebook’s iOS app and downloads in about a minute even on 2G networks—coupled with a persistent and compressed connection to Facebook servers that do most of the heavy work of fetching content and formatting it for display on the phone in order to deliver a low-bandwidth experience that still feels very much like Facebook. By March 2016, just nine months after its launch, Facebook Lite became Facebook’s fastest growing app ever as it reached 100 million users and counted critical markets like Brazil, India, Mexico, Indonesia and the Philippines among its top five heaviest users.

All the mobile effort has led to an amazing 90% of Facebook’s users now accessing it via mobile, a complete inversion compared to desktop in just seven years. (See Figure 5-2.)

Figure 5-2. Percentage of monthly global Facebook users accessing via mobile²

The Outcomes of Wearing Overalls and Looking Like Hard Work

Thomas Edison is often credited with saying, “The reason a lot of people do not recognize opportunity is because it usually goes around wearing overalls looking like hard work.” Edison would have liked Facebook.

When you walk through the areas belonging to the growth and internationalization teams in Facebook’s cavernous offices today, you notice three things: dozens and dozens of flags representing countries all across the world, many different languages being spoken, and hard work in the form of monitors.

Lots of monitors.

Like some unholy union of Bloomberg terminals and hospital heart rate monitors built just for Facebook, they hang everywhere, and the teams meticulously observe country by country—and practically in real-time—the impact on engagement of every obscure local holiday, every broken contact importer, every periodic communist government shutdown in Vietnam, every launch—or end—of a mobile operator promotion, every rise or drop in the average review scores of its mobile apps, every little change in Facebook’s product, even the impact of local weather. They do it because even a 1% difference in either direction is a big day.

What, then, has all this effort added up to? Good news, that’s what.

In early 2016, eight years after the birth of the growth team, Facebook had grown to 1.59 billion monthly users—and 1 billion daily users, making it the leading social network in 129 countries (with Facebook’s own Instagram in second place in 41 of those countries).³

Only four countries with Internet infrastructures had eluded Facebook’s concerted winning pursuits:

China: Facebook is currently blocked by the government.

Russia: Local social network VKontakte—a blatantly stolen and localized version of Facebook—illegally offers premium TV, movie and music, keeping Facebook from picking up more than 15 million of the country’s 105 million Internet users.

Japan: Local messenger Line has roughly double Facebook’s 30 million users.

South Korea: Local messenger KakaoTalk has 15% higher penetration.

And what of Chamath, Javier, Alex and Naomi eight years later? Fear not. They are still plying their trade.

Palihapitiya left Facebook in 2011 to become part owner of the Golden State Warriors NBA franchise, occasional professional poker player and outspoken founder and CEO of the ascendant venture capital firm Social Capital Partnership, which has made growth marketing—with partners formerly from Facebook’s growth team—one of the primary services it offers to its portfolio companies. Palihapitiya has become so influential that his is perhaps the only voice that could displace that of Marc Andreessen as the seer of things and speaker of truths among the new generation of venture capitalists. Or at least that’s what Palihapitiya would have you believe.

Following Palihapitiya’s departure, Olivan became the leader of all of Facebook’s growth efforts, reporting directly to Zuckerberg and overseeing all growth-related products, marketing, analytics, data science, internationalization, Internet.org and Facebook’s Social Good efforts.

Schultz now leads growth marketing and data science for not just Facebook’s users but also its long tail of advertisers—as of September 2016 there were more than 4 million—and other Facebook products that can benefit from his team’s expertise, including Messenger, which crossed its own 1 billion user milestone in July 2016. He has also become a leading light in the ranks of senior business leaders around the world active in support of the LGBT community.

And Gleit is now the second longest tenured Facebook employee after Zuckerberg himself and has ascended to lead all growth and engagement products after having been involved in everything from Facebook’s New User Experience to People You May Know, privacy simplifications, tools to extend Facebook’s experience to very low-end phones, the Follow feature and the Social Good capabilities including nonprofit donations tools that have facilitated millions of dollars in giving between people and organizations all over the world, and the Safety Check feature that Facebook turns on locally during disasters such as earthquakes to allow Facebook users to check themselves in as safe and to stay connected with family and friends around the world.

Looking Ahead: The First Billion Were Not the Hardest

If the 1.59 billion people using Facebook every month represent the first 12 years of Facebook’s mission to connect the world, what has to happen in the next 12?

Before Facebook looks to grow further, they have to retain all their current active users, as Alex and Chamath taught us in Growth Math 101. That means keeping a very close eye not just on the data of their own services but also on that of others in order to understand what it is that people most want to use to feel more connected. This is so important that in 2013, Facebook spent $180 million to acquire little-known Israeli mobile analytics company Onavo and proceeded to shutter their public-facing insights, retaining just for themselves the ongoing data on which mobile apps were being used on a daily basis—and how they were used—on millions of phones around the world. While it was Facebook’s own platform data that would help Zuckerberg understand Instagram’s growth and daily engagement and determine the value of having it be a part of the Facebook mission, it would be Onavo’s data that allowed him to understand WhatsApp’s growth and intense engagement prior to that acquisition. And it will be Onavo’s data that allows him to keep a close eye on other international players like Line, KakaoTalk, QQ and WeChat, and those closer to home like Snapchat.

With the flanks covered, it’s time to get on with growth. To get a clearer picture of Facebook’s process for future growth, Table 5-1 shows an overview of their current global presence across regions overall and a few of the largest countries specifically. The left side of the table captures the current state of the three most important variables in Facebook’s fate: country population, the degree to which the Internet is accessible to that population, and the degree to which those Internet users use Facebook. I’ve shaded the regions and countries where Internet or Facebook penetration is particularly high (darker shade) or low (lighter shade).

The right side of the table is a hypothetical growth-hacking “worksheet” that we’ll get to a little later.

Getting to two billion monthly active users—a trend line using data from the last few years suggests this could happen as early as the end of 2017—and far beyond can be simplified into two simultaneous long-term pursuits.

1. Increasing Facebook penetration among existing Internet users: While there are 3.4 billion global Internet users,⁴ there are only 1.59 billion monthly Facebook users. Not everyone, of course, will use Facebook every month, but many countries have not yet reached the nearly 70% usage rate Facebook enjoys in countries like the United States (190 million Facebook users out of 280 million Internet users).

Table 5-1. Population, Internet penetration and Facebook penetration (millions, Q4 2015)

As we can see in the right part of the growth hacking worksheet, the biggest opportunity here would be for Facebook to reach an agreement in the longer term with the Chinese government to provide some sort of connectivity product to the 675 million people on the Internet in China. A “winning” Facebook penetration of around 56% would mean another 375 million users. Zuckerberg is prepared for this to be a patient, long-term engagement, and his learning of Mandarin and becoming a lecturer at Tsinghua University are just a part of this effort.

The next biggest opportunity after China is clearly India, where Facebook is the social networking leader but has only 136 million of 376 million Internet users as monthly active users, an upside of between 74 and 119 million additional users depending on penetration.

The next closest opportunity would be the three big sub-Saharan countries of Kenya, Nigeria and South Africa, where Facebook counts only 33 million active monthly users out of 150 million Internet users, leaving an opportunity for 50 to 70 million additional users.

Japan and Germany will need to continue to be a focus for increased penetration as Facebook lags there with only 22% and 40% penetration of Internet users, respectively. The two countries are 3rd and 4th in advertising spend in the world—after the United States and China—and Facebook will need to make sure they are increasingly the choice for being connected to people to show that it is a relevant personal influence in these markets.

While Russia also presents a significant population with healthy Internet connectivity, it has been hard for Facebook to grow beyond a small 11% share of that audience. Instagram is actually more popular in Russia, but the country promises to continue to stymie Facebook’s overall efforts here.

2. Increasing Internet connectivity of the world’s population: Only about 46% of the world’s 7.3 billion people are on the Internet today, even though the fraction of people over age six in the world who have a mobile phone is projected to reach 90% by 2020.⁵ The state of the infrastructure, cost of Internet access relative to discretionary income and even the awareness of and desire for the Internet are very real barriers in many places in the world.

In Asia (expected to be the region with 56% of all new smartphone subscriptions globally between 2016 and 2021 at 1.43 billion⁶), the Indian subcontinent (India, Pakistan and Bangladesh) and the Southeast Asian countries of Indonesia, Philippines and Vietnam present significant connectivity upside.

As can be seen in Table 5-1, infrastructure projects for India, Pakistan and Bangladesh that would lift Internet penetration anywhere from the global average of 46% all the way to 71% (similar to countries like Russia or Kenya) and Facebook penetration to between 56% and a high of 79% (in Pakistan) could add between 250 to 600 million users.

In Indonesia and the Philippines, where practically everyone with Internet access is on Facebook, and in Vietnam, where Facebook has 74% penetration, improved Internet penetration could add between 40 and 150 million users.

Another big connectivity opportunity are the nearly 50 countries in sub-Saharan Africa, which have a combined population of about 1 billion people but by far the world’s weakest Internet penetration at less than 28%. It’s obvious that these numbers would not escape a team as focused as Facebook’s growth team. They’re so focused on it, in fact, that together with Facebook’s Connectivity Lab, they have partnered with European agency Eutelsat, Israeli satellite manufacturer Spacecom and U.S. satellite launch vehicle provider SpaceX—the first collaboration between Silicon Valley titans Zuckerberg and Elon Musk—to launch broadband satellite AMOS-6 in the second half of 2016 to provide coverage to 14 countries in that very region: Nigeria, Ethiopia, Democratic Republic of Congo, South Africa, Kenya, Tanzania, Uganda, Ghana, Cameroon, Cote d’Ivoire, Angola, Senegal, South Sudan and Gabon.

Together, the countries to be covered by the satellite represent a population of 670 million people, of which 202 million are on the Internet but only 54 million on Facebook.

If Internet connectivity could be improved to the world average or beyond, and assuming winning or better penetration for Facebook, it could mean an additional roughly 150 million to 300 million people connected via Facebook.

Table 5-2 shows the detailed numbers behind this sophisticated growth effort.

Table 5-2. Intended AMOS-6 satellite coverage country population, Internet penetration and Facebook penetration (millions, Q4 2015)

As if to illustrate both the very real risks of increasing connectivity around the world and Zuckerberg’s imperviousness to them, as this book went into production in September 2016 the first instantiation of the AMOS-6 satellite was the victim of a catastrophic explosion of a SpaceX Falcon rocket on its launch pad at Cape Canaveral Air Force Station while being fueled ahead of a test firing the day before the satellite’s launch. Zuckerberg, who was in Africa at the time to celebrate the satellite’s anticipated arrival in space, was undaunted in his update on Facebook following the incident:

As I’m here in Africa, I’m deeply disappointed to hear that SpaceX’s launch failure destroyed our satellite that would have provided connectivity to so many entrepreneurs and everyone else across the continent. Fortunately, we have developed other technologies like Aquila that will connect people as well. We remain committed to our mission of connecting everyone, and we will keep working until everyone has the opportunities this satellite would have provided.

Perhaps even more telling was Zuckerberg’s reply to a commenter who asked what the insurance costs of a satellite were: “The problem isn’t the money; it’s that now it may take longer to connect people.”

Certainly Zuckerberg is committed, but the satellite’s demise makes abundantly clear that none of this will be easy. Contrary to popular opinion, the pursuit of connecting emerging markets is not a Machiavellian revenue grab by Zuckerberg. Connecting these markets is both more expensive and less lucrative than established markets like the United States—they monetize at less than one-tenth the level per user today.

If Zuckerberg were obsessed only with the return of every dollar he spent today, he would not be putting satellites in orbit. It’s just the continuation of the world’s most deliberate growth engine in the service of a mission to connect the world, and in about five years, just the examples I’ve covered could represent an additional billion users for Facebook.