Facebook just had modern history's worst IPO and it's down again today by some percentage that will be quoted endlessly.
Yet Facebook is still the world's largest social media site with more than 900 million users. And most importantly: Many of those people still visit the site each and every day, regardless of grumbling from cynics or short-term price fluctuations.
We've been hard on Facebook at times through the last year, particularly around issues of privacy. We will continue to report on them with the same rigor that we've maintained over the years. But as the company's stock has fallen since its IPO, so many people are piling on with glee that I've become skeptical of my own skepticism about the company. It's not so much that nearly a billion people can't be wrong, but more that I know that I can be. (Especially when media sentiment is running so strongly in one direction.)
So, consider this the skeptic's guide to the bull case for Facebook. Because, although you might not know it from the recent commentary, there is still a case for Facebook's success and it is strong.
A Natural Monopoly
Let's start with the most fundamental argument for Facebook: That its unique market position is profitable and will remain profitable.
"The case for Facebook is that it is a natural (tech) monopoly, and natural monopolies tend to be stable, profitable and rare, thus good investments," investor Paul Kedrosky wrote in an email to The Atlantic. "Further, such companies tend to be underestimated by the markets, with most people initially skeptical -- That can't ever get to the scale it needs! -- and then skeptical later -- That can't continue to grow! -- as the company matures. Confusing the trainwreck IPO with the company's fortunes is classic market myopia, getting caught up in short-term drama and forgetting what was interesting in the first place."
This is also a big one. Take a look at the most-recent Comscore numbers for how many minutes per month visitors to the most popular social media sites spend. Let's add up Facebook's competition: tumblr (89), Pinterest (89), Twitter.com only (21), LinkedIn (21), MySpace (8), Google Plus (3). That's 227 minutes a month. Facebook is pulling in 405 minutes. And it has the largest user base at almost 170 million monthly active users in the US.
Stop and think about that. Facebook is bigger than tumblr, Pinterest, Twitter.com, LinkedIn, MySpace, and Google Plus ... times two. Put all these numbers together and what you get is: One in every five page views on the Internet is a Facebook page. If the Internet is valuable, Facebook is valuable.
Let's put Facebook's 900 million global users in perspective. The largest international broadcast network is the BBC World Service. It reaches 188 million people a week. Facebook, by contrast, reaches so many more and in so many more places. It has become the most global and largest media channel the world has ever known.
Whatever hipster backlash to Facebook that exists in the United States, I didn't see any sign of it in Korea earlier this month. That country had a very strong entrenched social network in CyWorld. But every youngish person with a phone I could collar told me that Facebook has crushed CyWorld. Many still maintain accounts on both networks, but only really use Facebook.
Point being: Our media obsessions are not the world's media obsessions. Our backlashes are not the world's backlashes. And the fact is that 80 percent of the service's users are outside North America. Four out of every five Facebook users comes from somewhere else.
Facebook is poised for big user growth in just about every developing country except China. Facebook is already over 51 million users in India, for example, and Gartner predicts that India will overtake the US as Facebook's largest market within the next three years. Facebook's also very strong in Brazil and Indonesia; it has more than 40 million users in each country.
Even civil rights activists like the Electronic Frontier Foundation's Jillian York have grudgingly accepted that Facebook is useful for the people they are trying to help. (Though I can't help noting that Facebook can do a lot more to make itself a better platform for public action.)
Facebook's design was amazing when it launched. The competition was MySpace, which was obsessed with "personalization." MySpace's theory of design seemed to be that the more offensively cluttered the page was, the more its users would care about the service. Facebook launched with the cleanest, whitest interface imaginable, setting a new standard for How Things Should Look.
Fast forward some years and Facebook's design may not be what it once was. But watch out. Last year, they acquired the design firm Sofa. Then, of course, there was Instagram and its design team. And then Lightbox, makers of a nice Android photosharing app. The list goes on and on. Wilson Miner, who helped redesign Apple.com, just left Rdio for a new gig at Facebook. Lastly, they just acqui-hired the design usability firm Bolt | Peters.
Add it up and we have to agree with Flipboard designer Dider Hilhorst's declaration, "Facebook design is the team to beat." And notice how many of those hires had to do with A) mobile B) photos or C) both.
Here, I will defer to mediamaker (and former Twitter employee) Robin Sloan and his newly published piece on Facebook and photos:
[Photos are] completely central photos to Facebook's product, and by extension its whole business. The company's S1 filing reports that, in the last three months of 2011, users uploaded around 250 million photos per day. For context, around 480 million people used the service on any given day in that span. That's like... quite a ratio. A whole lot of people sign up for Facebook because they wanted to see a friend or family members's photos, and a whole lot of people return to the service to see new ones. (And I mean: does the core Facebook behavior of stalking really provide any satisfaction without photos? No, it does not.) Really, Facebook is the world's largest photo sharing site -- that also happens to be a social network and an identity system.
And it's not just *your* photos that matter on Facebook. It's all those photos other people have taken of you and your friends. That means you can't simply take your ball and go home; all the other memories captured by friends that you have easy access to through the system? You can't have them without everyone sitting on the same system.
Google has a ton of data about people's search behavior. They used it to improve the search engine, counting your clicks as votes, essentially. But for a long time, we didn't necessarily see the impact of that information accumulation. Then, in the past year, Google started to unleash a variety of services that show how much data they really have. Start typing any query and Google can tell you what you're most likely to be looking for. That saves you time and opens up new mental pathways for thinking about searches you're doing.
I think Facebook is in a similar position as Google was a couple of years ago. They've been using the data gathered by all those clicks and +1s and using it to make Facebook more addictive. Somehow, the things you're most likely to be interested in show in your news feed. That's the data, right?
But that's really only the beginning. They've amassed the data, which is the hard part. As The Believer's Max Fenton put it, "The pulled off the big heist of making a database of all the humans that isn't explicitly evil. It's a babel, and it's great." But the majority of the data that Facebook has isn't all as easy to parse as counting the +1s and comments to posts.
It's much more difficult to parse the real language that people use to communicate with their friends to determine what's going on in their lives and what that means they want to purchase. But people are flocking to the fields of natural language processing and sentiment analysis, among others, and Facebook will benefit from advances in those and related AI fields.
Perhaps the most highly publicized problem with Facebook's business came when General Motors pulled its advertising from the service, saying Facebook's ads didn't work. Today, AdAge revealed why that happened. GM wanted to take over pages as many companies do on media sites. Facebook wanted to keep the user experience it had built. GM walked and Facebook's users were protected from intrusive advertising. If you're an advertiser, maybe that gives you pause, but if you're a user: HEY, Zuck's looking out for you!
More broadly, we know when Facebook has erred with its users (remember Beacon?). But we don't know how many times they've avoided making mistakes that would have annoyed users. Can you imagine how many pitches they've gotten to increase revenue by placing more invasive ads? GM certainly wasn't the first and definitely won't be the last.
Facebook is not exactly known as a sanctuary of privacy. Their business, after all, is selling you ads based on what you say and do. Insofar as you life on Facebook, you live a perfectly observed and recorded life. So I was shocked when I took a look at a recent paper by two scholars in the North Carolina Law Review arguing that Facebook should actually be seen as a model for privacy. Hear this out, though. It's interesting.
University of North Carolina legal scholars Andrew Chin and Anne Klinefelter look at the problem of "reidentification" and how Facebook appears to have solved it. The basic problem is that we say that data has been anonymized if you remove someone's name, but in reality, that's not true. Many different studies have shown that by combining outside data with the output of a database, it's not that difficult to reidentify someone. For example:
Latanya Sweeney, then a graduate student at MIT, merged presumably anonymized Massachusetts state worker hospital records with voter registration records and was able to identify rather quickly the health records of then-Governor William Weld.30 Sweeney later published a broader study finding that 87% of the 1990 U.S. Census population could be indentified using only gender, zip code, and full date of birth, and others reproduced this work in the 2000 Census with 63% success in identifying individuals.
This is a big problem. Big enough that University of Colorado privacy and legal scholar Paul Ohm declared anonymization the core problem with our current world's "broken promises of privacy."
So, one can imagine within this context, that Facebook would be a big problem. They've got the world's biggest database of individuals, right? And they have a way for advertisers to access groups of those people. So one could imagine that Facebook's advertising system might be what Ohm calls "a database of ruin." Fascinatingly, this does not appear to be the case, though. The North Carolina researchers argue that Facebook has achieved "differential privacy." They provide the best possible data to advertisers but manage to protect users' identities as well.
They do so with some clever techniques that Chin and Klinefelter tried to reverse engineer. The researchers hypothesize that Facebook is cleverly changing the rounding on its advertising reach numbers as well as intentionally injecting noise into the system to disrupt people who may be trying to mine the database. They are executing both in conjunction with short-term caching, so that queriers can't fine-tune their searches in order to probe the database and reveal more data than Facebook allows. Their conclusion:
Facebook's apparent implementation of a commercially successful, differentially private database mechanism provides hope that in at least some contexts, the law can recognize best practices that go beyond traditional anonymization techniques to better protect privacy while maintaining utility of data.
That's right! Facebook and privacy and best practices all in the same sentence. And they've created this system within their monetization wing, no less.
The MySpace Red Herring
It has long been trendy to compare Facebook to MySpace and Friendster, two social networks that were once dominant. But let's get real here. There's dominant and there's DOMINANT. No social network has ever commanded a greater share of Internet users, their time, or their shared media. And it is not even close. MySpace got passed by Facebook when they had something like 120 million worldwide visitors a month. That makes Facebook 7.5 times larger than MySpace ever got. Friendster? They were in the single-digit millions.
The story is the same with engagement. MySpace peaked at something like 240 minutes a month. Facebook's over 400 minutes. Repeat this same analysis for media sharing or any other metric of your choosing. Facebook is not the same kettle of fish.
We're in such uncharted territory that the historical precedents don't really make sense to apply. Or if we are going to apply them, it certainly can't be that every dominant social network is inexorably going to collapse.
Media companies online do not control their distribution channels. They can't directly deliver you content and ads. You must come to us and increasingly that means that you come to us through a social network. We used to put guys in trucks and send salesmen to retail stores; now, the social networks control the channel. Just to be clear here: many sites depend on Facebook for a substantial percentage of their overall traffic.
So, if Facebook decides that it wants users to see more video applications and less social readers, BOOM, it happens. In late April, Facebook made some changes to Timeline and next thing you know, Viddy and Socialcam are taking off like crazy and the Washington Post and Guardian social reader apps tank. If you're in the media, that should scare you. If you're Facebook, it's just another sign that you're in the driver's seat.
Note, too, that Facebook now knows precisely which publications and applications are doing well. Think that gives them an advantage in the public relations and acquisitions games?
Let's return to where we started with Paul Kedrosky. Here he discusses monetization with refreshing simplicity:
This is a two-dial company: You can grow users, or you can grow revenue per user. (Or both, of course.) Growing users is doable, but getting harder given increased saturation in existing markets. Further, it's obvious Facebook is having some ad inventory issues with respect to putting enough quality ad content in front of people, especially on mobile, so that's causing some of the per-user revenue slow down we're seeing.
Is the current dollar per user figure at Facebook a likely ceiling, floor, or average? It's almost certainly a floor, given that Facebook hasn't done much to monetize its traffic, beyond slapping ads against it. That means the current $2-ish /user figure that gets bandied about almost certainly goes higher, and if it, say, doubles in four years, that's an extra 18% juice in the revenue model on top of whatever subscriber growth the company gets.
How do you grow revenues? My guess is it's mostly ads -- more in mobile, smarter everywhere else -- but it's also getting into transaction curation. Look at Pinterest, Svpply, etc., to get a sense of how community, or at least a sub-community, is becoming very active in helping one another make purchases. That is easily fee-linked, and then you have another dial to turn in growth per-user revenues.
To recap: A natural monopoly is still an invaluable advantage. The largest social media site is still the largest social media site. The site is still growing quickly across the globe. One in every five page views on the Internet is a Facebook page -- and vast portions of the traffic that flows to other sites around the Internet is channeled through Facebook. If you think the Internet is valuable, then you implicitly think Facebook is valuable, too.
Other systems may be possible -- social networking may decentralize or transform in an unforeseen way -- but let's not let 10 days of share price fluctuation blind us to Facebook's unprecedented accomplishments. Not least because an accurate representation of their market power is necessary to hold them to the right standards of privacy, responsibility, and fair dealing.
Department of Omissions: We did not go into Facebook's possible role as a payment method. There's much to be said on this topic as an alternative monetization strategy as investor Eghosa Omoigui pointed out to me on Twitter. Also note that it will be the subject of an upcoming story.