Watch Microsoft Visualize Viral Content

What does “viral” look like? That’s precisely the question Microsoft Research is trying to answer.

We know when something goes viral from YouTube pageviews and the related emails from our moms. But what does this concept of viral really look like, at the user level? That’s a question Microsoft Research is working to show us in ViralSearch.


The engine, which isn’t available to the public, breaks down story sharing like a branching tree. Its chief goal is to clearly demonstrate the difference between typical “broadcast” content and viral content. A piece of broadcast content consists of one entity firing off a story to countless individuals (just like we always did through TV or radio). A slight fraction of those viewers see it and share it, so within a few generations, the story funnels away. It’s a missile-barrage approach that represents how every media company (including us) generally tweets.

In viral content, ViralSearch demonstrates that information really does spread like the plague. That initial tweet is still a missile barrage (because that’s just how Twitter works), but then at every level, there’s a multiplication effect–a sort of self-sustaining information-based pyramid scheme in which one person convinces 10 people to each share this story with 10 more people. Not everyone shares the story, of course, but enough do so effectively that the trend can build upon itself.

This branching view is fascinating, but it’s just one perspective to understand viral content. In another view, we see the sources of these branching narratives, a forest of what Microsoft calls independent “trees.” Here we see all of the discrete, original shares of a story that spawned their own viral growth. Within each circle, a dark center represents how many people shared the content. Then around it, you’ll see a series of concentric rings. These rings age how many generations this share has been around, just like rings date a tree, signifying who promoted a story first. That’s particularly useful from an attribution standpoint to separate the original thinkers from the opportunists who try to lay claim to an idea late.

It’s a shame that ViralSearch isn’t available for public consumption. Luckily, these viral visualizations are on trend, so it’s only a matter of time before the Chartbeats of the world feature them as a standard. There’s just so much more we can see in a picture than the sheer number of RTs.

See more here.

[Hat tip: The Verge]

About the author

Mark Wilson is a senior writer at Fast Company. He started, a simple way to give back every day.