Inside Mashable's Tool That Maps How Branded Content Is Shared

Knowledge Graph Pairs With Site's Tool to Predict Virality

By Published on .

Reprints Reprints

Example of Mashable's Knowledge Graph mapping the shares for a story about Stevie Wonder doing karaoke with James Corden.
Example of Mashable's Knowledge Graph mapping the shares for a story about Stevie Wonder doing karaoke with James Corden.

Mashable is making a science out of predicting when a story is about to pop around the web, and now it's built a tool to map how that virality ripples across the social sphere.

Mashable's Knowledge Graph tool tracks how brands' content posted to its site can pinball across various social networks, including difficult-to-track ones like email and text messaging. Mashable has been testing the tool with MasterCard -- which has been running advertorials on Mashable tied to the publication's Social Good Summit, of which MasterCard was a sponsor -- and plans to preview it this week during a party corresponding with the Internet Advertising Bureau UK's Digital Upfronts in London.

Mashable's Knowledge Graph deconstructs the traditional share count by visualizing how a tweet can lead to a number of retweets, which can lead someone to share the content on Facebook, which can lead someone else to share it via email or text message. Then it displays those shares in a web-like chart documenting how a given piece of content populates on one social network and then another, which can help a brand's social marketers decide the progression of their posts. The tool's data is anonymized, so brands won't be able to identify the individuals that are able to ignite sharing among a wider net of people. But Mashable is talking with some outside companies about incorporating demographic data, which could inform brands' media buys.

"It's early days for this kind of data. There's just a lot we need to learn and look at before we come up with super concrete steps" on what actions brands should take based on the Knowledge Graph's charts, said Mashable's CTO Robyn Peterson. "At this point the really interesting thing is it tells you what's driving success in general and then what's driving your success on the networks and then where there's opportunity."

However, Mashable isn't the only one trying to expose that opportunity. Its Knowledge Graph is similar to BuzzFeed's Pound, which was unveiled in April and does roughly the same thing. The two products are so similar, in fact, that Mashable internally code-named its version "Kilogram." "It was just kind of a fun thing, how it's a little over two pounds," quipped Mr. Peterson.

But Mashable's Knowledge Graph isn't exactly a fast-follow attempt to keep pace with BuzzFeed. The site had first tested the tool in December 2013, then decided in the second half of this year to develop it into a product it can license to brands. Though its origins really stem back to Mashable's other analytics tool, Velocity.

Officially launched in December 2012, Velocity forecasts when an article is about to go viral. By crawling more than 3 million links and directly churning through Facebook's and Twitter's data, the algorithm behind Velocity collects information on a story's subject, how many times it has been shared and how it's being shared, including how quickly. Then it calculates how many shares the story still has in the tank and how long it will take to hit that mark.

The first time Mr. Peterson showed an early version of Velocity to Mashable founder Pete Cashmore, he thought it was broken. Velocity was supposed to reveal which articles or memes were about to shoot up in popularity, but instead of a link to a scoop on The New York Times or a GIF on Tumblr, the top result Mr. Cashmore saw was a link to the Facebook page of a Walmart store in Kodiak, Alaska.

"We shut down the crawlers. We started crawling through the data and were like, why did the Facebook page of the Walmart location in Kodiak pop. We didn't even tell it to crawl that, and how did it even get there," Mr. Peterson said.

But Velocity wasn't broken. Walmart had been hosting a contest asking people to choose which one of its stores nationwide rapper Pitbull should visit. The store whose Facebook page had the most likes would win. So two guys behind comedy site Something Awful campaigned to send Pitfall the most out-of-the-way Walmart in America, not just continental America. That campaign to #ExilePitbull began to gain attention around the web, and Velocity noticed it was only getting started. The Kodiak Walmart's Facebook page gained tens of thousands of likes, winning the contest and proving Velocity right.

"What that told us was this system could actually work," Mr. Peterson said. "So from that point on, we really started investing in it and utilizing it across the business." First Mashable took the tool to its editorial team to get an inside edge of what stories to jump on, and then Velocity took over Mashable's home page.

Since the site's December 2012 redesign, Velocity has decided which articles to sort into the home page's three sections -- what's new, what's rising and what's hot -- and is also used by Mashable's social team to pick which stories to post to its Facebook, Twitter, Pinterest, etc. accounts. Mr. Peterson's team is currently developing Velocity to sort through videos embedded on the various pages it analyzes in order to identify which videos are about to go viral.

Mashable's Velocity tool can now predict videos' virality.
Mashable's Velocity tool can now predict videos' virality.

Recognizing its tool's utility for any content producer, last year Mashable began licensing it to agencies, starting with 360i and then MEC, as well as brands, which are able to incorporate Nielsen data to filter what content is about to go viral for which specific audience demographic. Charging companies a yearly fee to use its tool opened up a new revenue stream for Mashable, which also makes money from advertising and events. And now with Knowledge Graph, the site is looking to grow that revenue stream by pairing the two tools for brands to license together.

"If you say that Velocity is our weather prediction system -- tells you which hill to stand on if you want to see the lightning strike -- what KG does is it helps you understand thoroughly how that lightning is striking and then ideally it helps you bottle it so you can let it out whenever you want," Mr. Peterson said.

Most Popular