Facebook and Social Networks

“I think that network effects shouldn’t be underestimated with what we do.” These were the words of Mark Zuckerberg, Facebook founder, in an interview in 2007.14 Facebook is one of the most successful network-effect-builders in history.

Zuckerberg created Facebook in 2003 as a sophomore at Harvard. Zuckerberg, a computer science major, enjoyed producing small software programs based around the notion of social relationships, such as “Course Match” a software program for selecting courses based on which students had already enrolled in them. Facebook started as one of these small experiments. The idea was to produce an online version of the popular paper face books that universities produced with the pictures of freshmen students as they arrived at campus. The first version of the software, “Facemash,” invited users to compare student pictures to vote on who was sexier. The software was an instant hit at Harvard dorms: in a matter of hours 22,000 pairs of pictures were voted. However, the experiment was short-lived, as the university sanctioned Zuckerberg for a violation of the college’s code of conduct for breach of security, copyright, and privacy, given that the pictures had been hacked from the university’s system. Facebook had committed its first privacy breach even before it was incorporated!

On January 11, 2004, Zuckerberg registered the web address “thefacebook.com,” and a few weeks later, on February 4, the website went live. Facebook invited users to upload their own content (without any more hacking), like pictures and basic information about themselves. It also facilitated communication between registered users, empowering them to invite their friends to join what was already named as the “social network.” Access was limited to individuals with a Harvard email address. In four days, 650 Harvard students had registered. After three weeks, more than 6,000 students - three-quarters of Harvard undergraduates - were members.15 By the end of the month similar services were launched at Columbia, Stanford, and Yale, with similar success.

The biggest source of new users was a program that imported people’s contacts from Microsoft’s Hotmail (see an analysis of Hotmail in Chapter 7). Microsoft objected: “They were [...] trying to build their social network on the backs of others.”16 Facebook ended 2004 with 1 million registered users. It was not only numbers that ramped up very quickly; even more relevantly, users would spend hours browsing through the site and improving their own profiles.

Network effects were key. The more registered users there were, the more value the network would have for each user. The growth strategy was perfectly designed.17 Instead of opening Facebook to any user, they opened universities one by one, to ensure that a significant proportion of each community would enroll; this made it attractive for users as they would be able to connect with a substantial number of their friends. Had Facebook opened the network to any user, new members would have found out that none of their friends were connected and would probably never return to the platform, making the network useless. By focusing on elite and closed environments like Ivy League universities, one by one, Facebook cracked the “chicken and egg” challenge of users expecting other users to join the network. Thiel at PayPal had used the same strategy.

Discussions about network effects were common in the early days of Facebook, just as McLuhan’s Understanding Media was favored reading. After all, the founders of Facebook were not just college dropouts, they were Harvard dropouts. Facebook exhausted the US university market, then moved to universities in other countries, and then to high schools (September 2005); it was only in September 2006 that Facebook opened itself up to the general public. At that point in time, it had already around ten million active users. It was a perfect lesson in how to grow a community.

Facebook attracted the attention of the early champions of network effects in the internet. The company appointed Sean Parker as president in its first months. Parker had started another digital platform in Napster, the famous file-sharing service (discussed in the next chapter). In September 2004, Parker invited his Silicon Valley contacts to invest the first $500,000 in seed capital, in exchange for a 10 percent of the corporation. The leading investor was Peter Thiel, who had founded and then sold off PayPal and had also invested in other platforms such as Linkedln and Spotify (see Chapter 8). The other investors were Reif Hoffman, co-founder of PayPal and subsequently of Linkedln, and Mark Pincus, who was also an early investor in Napster, Twitter, and Snapchat. These men all had a deep understanding of the power of network effects, which is precisely why they invested in Facebook.

Over the years, Facebook has built the largest community in the world, with 2.7 billion active monthly users as of 2020. Facebook acquired Instagram, a popular alternative social network, as well as WhatsApp, the largest instant messaging app in the world (see Chapter 8). “[M]ake no mistake, growth tactics are how we got here.” “The best products don’t win. The ones everyone uses win.” These statements in an internal memo under the title The Ugly encapsulate Facebook’s strategy.18 Facebook creates massive direct network effects, as the largest pool of users exchange their content, comment, give likes, and so on.

Having the largest customer base provides a large advantage to Facebook. Once customers have opened an account, created the network of connections, spent hours curating their profile, and years and years uploading their pictures and videos, they are locked in. “[T]here are network effects around social products and a finite number of different social mechanics to invent. Once someone wins at a specific mechanic, it’s difficult for others to supplant them without doing something different”. These words from Zuckerberg are repeatedly quoted in the antitrust suit filed by the US Federal Trade Commission against Facebook in December 2020.

In 2007 Facebook launched Facebook Platform. It encouraged developers to build apps and tools (games, page-design tools, e-marleting apps and so on) on top of Facebook. Developers would be allowed to use Facebook’s Application Programming Interfaces (APIs) to extract content from Facebook, including the Find Friends critical API. Facebook successfully built an ecosystem of apps, which grew on top of Facebook’s social graph. Interoperability was encouraged when it strengthened the platform (more content and apps), but it was restricted when it would allow the development of competing products or multi-homing with other social networks (individuals in small social networks cannot be reached from Facebook, content in Facebook cannot be accessed from other networks, and exporting it to other networks is certainly not facilitated). The US FTC found these restrictions anticompetitive and they were included in the antitrust suit filed in December 2020.

From the start, advertising was Facebook’s business model. Facebook included so-called flyers or small ads for a low cost ($ 10—$40), mostly around university activities, such as school material and parties. Facebook reached $400,000 in revenue in its first year of operations. This is how Facebook builds indirect network effects, as the platform connects 2.7 billion viewers with millions of advertisers, including small companies and individuals that were previously excluded from the advertising market. Over time, Facebook has built a non-transactional platform, just as traditional media have, but at a much larger scale of operations and therefore with much larger network effects. The 2.7 billion active users multiply the audience of any newspaper, radio station, or TV network in the world. This is one reason why Facebook makes more revenue that all the newspapers in the world, combined.

Nonetheless, data is as relevant for Facebook’s success as direct and indirect network effects. Data feed algorithms optimize the matching of users and advertisers, thus empowering the network effects. Algorithms target the ads to be displayed to users. Not only does Facebook allow the largest audience to be reached, but it also has such a deep knowledge of its audience that ads can be personalized. It is well known that that ads can double in price when they are personalized.19 The more data you have about your audience, the better you can personalize the ads, creating so-called algorithmic network effects (see below).

Creating algorithmic network effects poses obvious risks in terms of privacy protection. It is not by chance that the largest fine due to privacy breaches, $5 billion, was imposed on Facebook by the US Federal Trade Commission in July 2019, due to the Cambridge Analytica scandal around voter profiling with data extracted from Facebook.20

The value created by Facebook in terms of direct, indirect, and algorithmic network effects multiply the benefits of traditional media. In 2019, Facebook had revenue of more than $70 billion. It has reached a market share beyond 50 percent in digital display advertising, for instance in the UK, where the second-largest player’s Google’s YouTube market share is below 10 percent.21 Even more extraordinary is Facebook’s net income, which has reached 40 percent of its revenue. The margins enjoyed by Facebook are exorbitant, proving that the company’s network effects are creating a lot of value, but also that the company has succeeded in capturing a substantial amount of it, not distributing it across its ecosystem.

< Prev   CONTENTS   Source   Next >