Last week, Facebook launched a major initiative geared towards getting users to share more information more openly. In the few days since, many people have criticized Facebook’s move as misleading, though it’s too early to tell if a significant number of users will be upset enough by the changes to complain or change their actual behavior. More broadly, however, the move reflects deeper changes in Facebook’s longer term product strategy. What were Facebook’s motivations for this “privacy” initiative, and what’s likely to happen as a result? Let’s take a look.
Facebook’s Privacy Foundations
Facebook’s privacy model has always been foundational to the trust users put in the company. While other services like MySpace have encouraged a more “open” way of sharing information and building an online identity, Facebook’s default information-sharing settings have always been relatively private. As a result, hundreds of millions of users around the world today routinely do things on Facebook that only a few years ago would be unthinkable, like parents commenting on their children’s status updates, and millions of people uploading thousands of personal photos to the Internet. Without Facebook’s historically strict privacy settings, much of what has happened over the last few years would not have been possible.
Privacy is not only foundational to the trust users put in Facebook, it’s a fundamental part of users’ conceptual models of how Facebook works. It’s what makes people feel safe sharing personal information. Many people just don’t feel comfortable putting their status updates or sonogram photos in the public Internet archive forever.
The Limitations of a Default-Private Model
However, while many people don’t want to share much information publicly online today, some do. For those people, Facebook’s historical default privacy settings did not make it the right product for them. As a result, Facebook recognized that its default-private model made it vulnerable to other services with default-public models, like Twitter. Even though only a relatively few people may want to share in a predominantly public way, many people may want to share some (arguably increasing) subset of things more openly, and many people are interested in consuming a variety of different types of public information.
In addition, the default-private model might actually slow down the spread of memes compared to more open systems (though we don’t have access to data necessary to back this up). The fundamental nature of a News Feed comprised of mostly private content makes “resharing” on Facebook a less common/normal behavior than “retweeting” on Twitter. While this dynamic keeps the content in the stream more pure, it also means that there may be fundamental limits to the amount of reshared public content that might ever come through the stream.
In other words, there are several use cases in which some users – and Facebook – would get more value out of a more open system.
This put Facebook in a tough position: if it believed that a more open system would create more overall value in the end, how could it move from a default-private model to a more open one? There was no painless way to enable even the people who would want to be more open to change their privacy settings en masse quickly. Facebook had to choose whether to let the historically private settings ride, or to make a push to get people to open up – even at the risk of losing some users’ trust.
Facebook’s Calculated Move
While Facebook’s decision to launch this openness initiative has been called a lot of things, one thing it can’t be called is thoughtless. Facebook was well aware of the implications of making this push, but ultimately felt that it was vital to its future to shift its default privacy model more toward open sharing. Facebook initially announced its intentions to put people through this “privacy transition” in July, even so much as showing a mockup at that time that looks nearly exactly like the “privacy transition wizard” that people saw last week – though with most of the options set to “Old Settings” instead of “Everyone,” as most people ended up getting.
That being said, Facebook’s decision to make the recommended privacy options for profile data like “Family and Relationships” and “Posts I Create” be set to “Everyone” – as well as its move to remove privacy controls for Gender, Current City, and Friends – were pretty aggressive by almost anyone’s standards. In particular, its decision to present users with a binary choice between “Everyone” and “Old Settings” for some privacy preferences was especially confusingly executed. Nevertheless, Facebook decided to bite the bullets of potentially significant user confusion and potentially severe loss of user trust in order to take this risk.
It’s interesting to observe how Facebook ultimately chose to delineate between fields it defaulted to “Everyone,” “Friends of Friends,” and “Friends” in the transition wizard. Those settings reflect the intended use cases Facebook’s product leadership has for the future of the service.
The Challenges of a Hybrid Public/Private Model
As Facebook has grown over the years, so has its ambitions. Today, Facebook is by far the largest social networking platform in the world, and has enabled new forms of efficient communication, advertising, and software distribution. As Facebook has mapped an increasingly larger portion of the “social graph” of human connections, it has expanded its definition of the “social graph” to include the businesses, products, brands, and services that we communicate with every day. That change fundamentally added a degree of complexity to the Facebook ecosystem, adding the new concept of asymmetric “fan” relationships between users and public profiles to the traditional concept of symmetric “friend” relationships that have existed between Facebook users since the beginning.
However, now that Facebook has chosen to push further toward the public end of the public/private hybrid system that Mark Zuckerberg envisions, it must face a more challenging and complex problem: letting users apply per-item privacy rules to each and every profile field and piece of content shared. Facebook isn’t satisfied with a mostly-private platform: it wants to be the single place where both sensitive personal information is shared and public memes spread.
While the “transition tool” that Facebook chose to roll out will effectively enable those who want to share more openly to do so en masse quickly, it also comes with some built-in problems. First, it will inevitably lead some people to inadvertently grant public access to content they intended to be private. That could lead to losses in user trust. Second, it creates an intrinsically more complicated privacy model, putting the burden on users both to construct a robust framework for sharing different types of information on Facebook, and to remember who they’ve allowed to see what. That could lead to user confusion, and fear of the unknown.
The transition tool Facebook implemented will likely lead to more users understanding their privacy settings better, and choosing the settings that are right for them. But both of these problems could cost Facebook in terms of its own engagement and virality numbers. The question is, if so, how much?
Grappling with Facebook’s Motivations
A fair question to ask at this point is: Was this openness initiative by Facebook either 1) evil, or 2) stupid? Most of the arguments out there that say Facebook’s move was one or the other (or both) conclude that Facebook’s ultimate motivation behind this initiative was to get more traffic through search. These arguments usually assume that Facebook is misleading a sufficient number of users to make the wrong choices about their privacy settings through the clever design of the “privacy transition wizard.”
Ultimately, those that believe that this move was evil or stupid must also believe either one of two things: 1) That Facebook has more to gain through increased openness than it has to lose through decreased user trust, and the company is intentionally and aggressively choosing to pursue this new open strategy despite the problems it might produce for users, or 2) That Facebook doesn’t understand its users and has made a short-sighted product decision. Let’s look more at each of these.
We would be cautious before running with the claim that Facebook is ready to abandon privacy wholesale. First, any material losses in user trust as a result of these changes will have a greater and longer lasting negative impact on Facebook than any gains it can realize through accelerated openness in the short term. Facebook’s privacy model is fundamental to users’ conceptual models of how Facebook works, and if enough users lose trust in Facebook’s privacy, that could cripple it forever. So much of the deep identity that lives within Facebook is verified through social interactions between real friends and family. Without privacy, many of these interactions would never happen. Over time, this would drastically weaken users’ ties to Facebook as a reliable place to share information, the signal-to-noise ratio in the News Feed would plummet, and the only people left sharing on Facebook would eventually be those comfortable sharing without privacy (i.e. like those who user Twitter today). That would open the door for another service which provided a privacy model more like what Facebook started with.
The contrarian view to this argument is more or less based around the thesis that people are on the whole becoming more open with what used to generally be regarded as “private” information – as has generally been the case over the past few years – and that Facebook is ahead of the curve on these changing patterns of global culture and values. This is a debate worth having, and Facebook has commented on the question as much as to say that it believes this is happening. Many smart entrepreneurs believe this to varying degrees, though no service has been able to accomplish as much as Facebook has (with its relatively private settings) to date.
If this does indeed become the case, it is possible that Facebook may be willing to increasingly sacrifice its legacy of privacy, at least partially, for a future of openness. But at the end of the day, we fundamentally believe that there will always be some information that most people just don’t want to share publicly, that Facebook believes that too, and that Facebook will take the steps to preserve that trust if this move ends up causing bigger problems than the company thought. Ultimately, Facebook’s willingness to sacrifice some users’ trust for a future of more openness may not only be due to the fundamental challenges of virality of public content in a mostly-private system, but also due to how well Facebook thinks it can monetize a stream of generally-private data compared to a stream that contains more types of public information, like marketing and news.
The possibility that Facebook has made an over-aggressive product decision due to over-infatuation with more open services, however, is a more plausible option. Facebook has shown, as recently as a few months ago with its launch of the “real-time” stream as the default News Feed, followed by its decision a few months later to go back to the algorithmic News Feed, that it is capable of making suboptimal product decisions due to intense feelings about services like Twitter, yet is also willing to correct them relatively soon afterward when sufficient feedback is in. Despite its investment in data-mining resources and user studies, Facebook’s largest product decisions are generally driven by the visions of those at the top. This had led to a variety of bold initiatives over the years, from the News Feed to Beacon to the Facebook Platform, with a wide range of results. If it turns out that Facebook is wrong about this move, I’d expect them to make similar course corrections in this case as well – even if it means reincarnating its fundamental vision in a new form, like the way Facebook Connect launched a year after Beacon.
How will all of this play out over the next couple of months, and the next couple of years?
While we may see an uprising of protest movements in the coming weeks, it’s also entirely possible that Facebook’s approach won’t actually cause widespread problems, and that most users will be able to navigate their privacy settings as they intend to. In addition, it’s even possible that not as many users will choose to make their data public as Facebook may have hoped – if that is the case, Facebook could take further steps to encourage or force more user data to be public in the near future.
The most likely result of all of this is somewhere in the middle: some users will open up their privacy settings willingly, and others will do so inadvertently. We assume that Facebook optimized these consequences across a variety of product designs it considered and tested before choosing to go down this path. If Facebook has made a significant miscalculation here, then we’re likely to see substantial numbers of users discovering these problems over the upcoming days, weeks, and months. If that happens, those users will likely change the way they use Facebook – either by publishing less information there, or using it less altogether.
At the end of the day, users will vote with their feet. This was a relatively aggressive move by Facebook to encourage people to open up, and while it will cause some problems for Facebook, Facebook is also acutely aware of the possible outcomes, and is monitoring the results of these changes with a fine toothed comb. If they don’t like what they see, they’ll change course. We’ll of course be monitoring all of the available gauges on Facebook’s traffic and engagement as well.
Ultimately, Facebook is trying to achieve something very difficult: create a system for sharing information that works for both private communication and public publishing, all in one place. While Facebook is in position to serve a variety of users’ needs, it will not be able to be all things to all people.