This article was originally published in Interactions Magazine.
Most weekday mornings are fairly predictable: I make a pot of coffee; I walk the dogs with my wife, Eliza; I have a second cup of coffee while Eliza gets ready. This probably sounds familiar, as we all have our routines. But this is not where the predictability in my day ends.
I check email on my phone to find a daily handful of mass mail from various research firms and business publications. Many of the articles within these emails (especially those targeted toward marketers) will be on the topic of social media. Perhaps this, too, is a normal part of your morning. If that’s the case, perhaps you have noticed that the content of these emails is also a bit predictable.
You tend to see these words: engagement, metrics, conversation, ROI, community, sharing, measurement, dialogue. Ultimately, these emails regress toward some variation on “Social media is about engagement! Companies need to join the conversation!” Starting your day in this way can make you feel a little bit like Bill Murray in the movie “Groundhog Day.”
I suppose there’s good money to be made in periodically browbeating companies into “joining the conversation.” But there are problems with this advice (that go beyond the sheer banality of it all). The biggest problem with these discussions is that they tend to substitute peer pressure for insight. You need to do this because everyone else is. No wonder so many brands feel panicked about social media. They feel like they need to be there, but they don’t know what to do or why to do it. And if they dare to question any of the conventional wisdom on social media, they’re accused of “not getting it.”
The more significant problem I see is the tendency to lump all social networks into one. If any differences are discussed, the conversation tends to be framed as a horse race. “Twitter is hot!” “Is Facebook getting old?” “Does YouTube make money?” But these various networks have some important structural differences. The better we understand what these differences are and how they affect behavior, the better equipped we’ll be to use these networks, design for these networks, and advise clients what to do with these networks.
Architects have long understood that the structures we inhabit can influence not only the way we feel, but also the way we behave. This turns out to be true in digital environments like social networks, too. Subtle differences in the underlying structures of these networks give rise to distinct patterns of behavior.
Some of the ways in which social networks differ affect the central metaphor that we often use to describe them: “the conversation.” In fact, the more time I spend in the social web, the more I’m convinced that the conversation metaphor isn’t quite right. Just as one’s “friends” aren’t necessarily one’s actual friends, “the conversation” isn’t always a conversation.
Take Twitter, for example. Despite consistently being described as a medium for conversation (its tagline is even “Join the Conversation”), there are some tangible, structural reasons why Twitter can actually be quite hostile to real conversation.
Reason No. 1: Public by default
In a post on the PopMatters blog Marginal Utility, Rob Horning wrote:
Because the online space is devoid of conflict—everyone is ‘friends’—it is anodyne; ‘the Tyranny of Positive Energy’ assures that politics is screened out of online social behavior.
Horning also points out that social networks often “prompt us to replace the tussle of genuine connectedness with further self-display. Instead of arguing with one another, we preen.” Horning equates this preening to a lack of conflict, but Twitter’s preening can actually become quite aggressive. Newt Gingrich infamously labeled Sonia Sotomayor a racist on Twitter: “New racism is no better than old racism,” he wrote, one presumes with a measure of self-satisfaction. It may be true that social networks don’t represent “genuine connectedness,” but that lack of connection often seems to escalate conflict. Just like when we’re behind the wheel of a car, being somewhat disconnected from others can bring out a whole new side of ourselves.
We act differently in public spaces than we do in private. Sometimes we’re funnier, sometimes we’re angrier, and sometimes we’re quieter. But in any case, we’re very much aware of our audience. Twitter’s public nature yields conversations that are often for bystanders as much as the participants, which changes both the way we behave and the content of our conversations.
Reason No. 2: Following, not friending
Connections in Twitter are non-reciprocal (one-way) by default. I can follow you even if you don’t follow me back. There is typically no approval involved—I don’t request to follow you; I simply follow you. It’s much more akin to “being a fan.” (In fact, “Become a fan” was actually the phrase Facebook previously used to describe the one-way relationships it offered to brands.) In Facebook, on the other hand, users have reciprocal friendships (except for brands with “Pages,” which allow the aforementioned “fan” relationship). I can’t be friends with you if you’re not friends with me. Friend requests are sent, and then the recipient approves or ignores them.
Non-reciprocal ties create a unique dynamic in Twitter. The first implication of this structure is that an individual’s network on Twitter is far less likely to look like their real-world social network. In other words, followers are not the same as friends. This effect is amplified by the fact that Twitter is simply less ubiquitous than Facebook. This often means the people within one’s Twitter network are less likely to know one another. Sociologists would call this a less transitive network. Low transitivity of connections may be another inhibitor to conversation in Twitter because you are less likely to have a shared context with the people reading your tweets.
The second implication of having non-reciprocal ties is that Twitter tends to take on the dynamics of a popularity contest. Followers can be amassed more easily than friends. And the more followers you have, the more influential you appear to be (whether or not that’s actually the case could be debated, but surely there’s a correlation). In fact, one’s followers are listed right under one’s name in a Twitter profile. So, almost naturally, Twitter becomes a game: Whoever gets the most followers wins.
What do people do in a public environment where they are evaluated based on the number of fans they have? They don’t talk about what they had for breakfast (unless they’re a celebrity). Rather, they try to be fascinating. This tendency has made Twitter a great venue for content sharing, and often a great filter for the news. But again, it doesn’t mean that people are having conversations. Even if there are occasional two-way exchanges, they are quite limited.
Reason No. 3: The 140 character limit
Twitter’s 140-character limit is surely the most talked about facet of its structure. And it isn’t difficult to see why. Having a conversation 140 characters at a time is like trying to swim laps in one of those tiny hotel swimming pools. As soon as you get started, you hit a wall.
South Carolina Representative Joe Wilson infamously screamed “YOU LIE!” at Barack Obama during his healthcare speech to Congress in September 2009. Whatever else you might think about this remark, it seemed to me a sign of the times, representing the new low our public discourse has reached. And yet Wilson’s outburst is exactly the type of “dialogue” one typically encounters on Twitter, particularly when one person is responding to another.
Consider Twitter’s ubiquitous interjection, “FAIL.” These days it seems that every one of life’s letdowns gets the #fail treatment. Here’s an example from @iprash on the recent Conan O’Brien “Tonight Show” debacle: “Not sure what Leno’s angle is — my feeling is he is being pushed around too. I think this is NBC FAIL pure and simple.” There’s just as little intellectual content in saying that something “is fail” as there is in screaming at the president “YOU LIE!” And how is one supposed to respond to such one-dimensional criticisms? Is this even a conversation worth having?
Perhaps this is simply the byproduct of what Clay Shirky recently pointed out on edge.org, that the “shock of inclusion, where professional media gives way to participation by two billion amateurs, means that average quality of public thought has collapsed.” Beyond simply giving the thoughtless a bullhorn, there is certainly also a structural reason behind Twitter’s remedial discourse. Even when you have otherwise intelligent people participating, they often end up sounding one-dimensional. There’s only so much you can say in 140 characters.
None of this is intended as a criticism of Twitter. I am a frequent (though not influential) user of Twitter and find the service to be consistently delightful, if a bit distracting. Twitter’s virtues are several, and foremost among them is a refrain commonly repeated in the world of social media: It makes relationships possible that never would have occurred otherwise. Nonetheless, its virtues don’t make it a business panacea, even though you might get that impression from its champions.
Companies and people who decide to give Twitter a try are often baffled by what they find. I’ve had several colleagues tell me over the past year, “I gave Twitter a try, but I just didn’t get it.” Sure enough, Twitter’s usage numbers tell a similar story. (According to a study cited in The Wall Street Journal, over a third of registered users have never even Tweeted .) I can’t help but wonder if those silent users of Twitter were disappointed because they were expecting to find a conversation but found something entirely different.
As products become more complex, designers will need to act more like policy makers or economists, who are concerned with the effects of rules on complex systems. In the world of social networks, default settings and network structure are more than just details. The structure of a network has a profound impact on user behavior. Understanding how these structures drive behavior can help us understand and design systems more effectively. You only need to look at the growing outrage over Facebook’s default settings to understand what’s at stake with something as seemingly mundane as network structure.
The structure of Twitter makes it a great platform for discovering content, finding people who share your interests, and getting connected with people to whom you wouldn’t otherwise have access. It isn’t, however, a great place to have a conversation. But the more we use social networks and develop a shared vocabulary of how structural considerations affect behavior, the better we’ll be able to design products, services, and networks that create whatever kind of conversation space we desire. Christopher Alexander gave us a pattern language of physical, architectural patterns. Now it’s time to create the digital version.
In the meantime, I’ll be clearing the spam out of my inbox, waiting for tomorrow’s mass emails, and hoping they have something more to say.
© ACM, 2010. This is the author’s version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published in Interactions Magazine, VOL 17, ISS 5, September + October 2010 http://doi.acm.org/10.1145/1836216.1836221