Intelligent Machines
Information's Social Highways
A startup studies the paths taken by viral messages
The thicker a line, the more retweets that connection generated. The larger the node, the more retweets that user's participation generated.
On the night last spring when Osama bin Laden was killed, the chief of staff to a former U.S. secretary of defense broke the news to the world—more than an hour before President Barack Obama’s announcement. Keith Urbahn (aka @keithurbahn) wrote to his 1,016 Twitter followers that he’d heard the news from a “reputable person.” Within a minute, 80 people had reposted the message. One of them was New York Times reporter Brian Stelter, whose retweet led to another large burst of responses. Urbahn’s tweet was on its way to going viral.
There is no recipe for virality, says Gilad Lotan, head of R&D for a startup called SocialFlow, which aims to help clients from the Economist to Pepsi more effectively capture attention on Twitter. But the deluges of data that viral tweets generate hold potentially valuable insights into how and why certain things spread beyond their author’s network of regular contacts. After the bin Laden event, Lotan took advantage of SocialFlow’s access to the Twitter “fire hose,” a real-time stream of every tweet, to analyze—and visualize—the responses to Urbahn’s post. The results are seen on this page.
Each colored circle, or node, represents a Twitter user who repeated the original message (or posted something similar) and mentioned the author’s Twitter handle. The color gradient conveys how long it took for any given message to join the conversation; for instance, bluer circles represent people who took up Urbahn’s message within minutes.
Where circles are connected by a line, Lotan is representing the likely pathways along which the message passed. He determined them by analyzing, among other things, when each message was published and the relationships between users (who follows whom).
The larger the node, the more retweets that user's participation generated.
Lotan made the graph on this page using the same methodology. It shows responses to a tweet, posted by engineering professor Deb Chachra (@debcha), that resonated especially widely during last summer’s riots in Britain. This message spread much more slowly than the bin Laden news, and it spread without the involvement of a widely followed journalist. These differences are reflected in the diffuse shape and smaller clusters of the graph.
Being heard isn’t always easy in an age when anyone can become a broadcaster. But analyzing and visualizing such data helps SocialFlow guide customers about how, when, and what they should tweet to have the best chance of disseminating their messages widely.
Note: The orientation of the nodes was determined by a force-directed algorithm, a tool for organizing network graphs to aid visual understanding.
