At BuzzFeed, even the data scientists are masters of memes.
Last spring, we first wrote about BuzzFeed Pound, a remarkable technology that uses network diffusion to reveal exactly how a piece of content spreads online, from sharer to sharer and platform to platform. A few days later, a tweet arrived in response to the story—a GIF of the headshots of red-headed identical twins flipping back and forth at mesmerizing velocity—thanking us for writing about the technology that BuzzFeed’s data-whiz Kelleher twins had been working on for years.
So we went to see them. Inside a conference room at BuzzFeed’s office near Madison Square Park in Manhattan, Adam Kelleher scribbled excitedly on a whiteboard, showing exactly how Pound works. While every other publisher just sees shares in buckets (100 Facebook, 50 Twitter, 30 LinkedIn, etc.), BuzzFeed can see the exact process that makes content go viral. Using an oscillating, anonymous hash in a sharer’s URL as a UTM code, Pound can track its content’s journey like an overprotective parent tracking his teenage son’s phone. Pound can see how a piece of content starts the night on Twitter, then gets picked up by a cluster of sharers and goes viral on Facebook, and even spreads to “dark social” platforms like Gmail and chat apps.
A classic example is #TheDress, an infamous Tumblr post about a dress that looked either blue and black or white and gold, depending on how your brain is wired. BuzzFeed was the first news site to cover #TheDress, which sparked an Internet-wide debate. BuzzFeed’s six tweets about it triggered a million views in downstream traffic. But only 25 percent of those clicks actually occurred on Twitter. The tweet’s biggest impact came from people subsequently sharing the story on Facebook or reporters writing about it. By traditional metrics, the impact of BuzzFeed’s original tweet would have been undervalued by a factor of four. But thanks to Pound, BuzzFeed can see value and opportunities that no one else can.
While Pound is impressive in its own right, it’s just one part of Hive, BuzzFeed’s ambitious data initiative. As the name suggests, Hive’s goal is to make sharing information about which content works instantaneous. It requires a massive investment in engineering—hundreds of top engineers organized into “scrums.” Hive is an unusual thing to see from a media company, but then again, it’d be a mistake to think of BuzzFeed as a media company at all.
Critics like New York Times fellow Greg Howard have called BuzzFeed “an advertising shop with a journalism wing.” While that characterization seems like a dismissal, it’s not so different from how BuzzFeed views itself. In financial documents leaked last year, BuzzFeed described itself as a company “that is redefining online advertising with its social, content-driven publishing technology.”
As an advertising venture, BuzzFeed is a success, although not a big of one as the company might have hoped. According to leaked financials, the privately-held company was growing at a 200 percent clip and was on pace to break $100 million in revenue in 2014. Per a recent report from The Financial Times, 2015 fell short of expectations, with the company setting a revenue target of $250 million but raking in less than $170 million. Still, nearly all of that revenue comes from creating and distributing content on behalf of brands. The company is aggressively expanding abroad, from the U.K. to Japan to Brazil, in hopes of being one of the world’s most powerful global advertising companies.
“Where this is all headed on the business side is global and cross-platform advertising,” Jonah Peretti, BuzzFeed’s founder, told Fast Company. “That means brands who work with us, they’re getting branded video, they’re getting lists, quizzes, they’re getting things on BuzzFeed’s site, they’re getting distribution off BuzzFeed’s site. That model will be the same in the U.K. as it is in the U.S.”
Put more simply, BuzzFeed is building the global ad agency of the future.
Swarm the industry
Twenty-five years ago, ad agencies only had to deal with a few formats to gain user attention: the TV spot, the print ad, the billboard. Then the Internet presented a couple more wrinkles, like display ads, pop-ups, and YouTube pre-rolls. But on a fundamental level, not a lot changed. You made an ad, and you blasted it out there.
The social web, however, has changed the game. User attention has increasingly clustered inside individual social apps, each with unique formats and content-related behaviors. While these apps have ad formats, they all mimic in-feed content, which means ads needs to be good enough to win user attention. It also has to be custom. What works on Facebook won’t necessarily work on Twitter, and what works on Instagram probably won’t work on Snapchat. For most brands and agencies, this is a logistical nightmare, and they’re just starting to grapple with the implications.
BuzzFeed, meanwhile, has already productized this problem. At SXSW Interactive in March, CMO Frank Cooper introduced Swarm, a new ad product that lets advertisers run campaigns across BuzzFeed’s website, mobile app, and six top social platforms: Snapchat Discover, Vine, YouTube, Facebook, Instagram, and Tumblr, with unique content crafted for each platform.
“Instead of forcing all your traffic and spending all your resources to drive people to your owned and operated platforms, why not go to where they are?” Cooper said. “We’ve done it, and in doing that we’ve built this power of an audience that crosses these different platforms.”
BuzzFeed has already been executing this strategy, both for itself and beta advertisers. Seventy-five percent of BuzzFeed’s content is now published outside its own website, driving almost 6 billion monthly content views, according to the company. What all of these views provide, ultimately, is data that BuzzFeed can plug into Hive to figure out exactly what content works on every platform and at most every major company. In the recent Fast Company profile, Peretti likened Hive to a fleet of self-driving cars that learn from each other and get exponentially smarter over time. BuzzFeed publisher Dao Nguyen preferred a Voltron metaphor, with data from each successful piece of content helping BuzzFeed build “an even more powerful robot that no one can defeat.”
This is all possible thanks to BuzzFeed Motion Pictures, the 52,000-square-foot studio on Sunset Boulevard in Los Angeles that houses 200 creatives cranking out video content at a torrid pace. Teams create videos on behalf of a brand around a specific message until they find a strategy that works; then, they double down on the most successful content.
“Their ability to conceive, create, test, define, and then either abandon or rinse and repeat is phenomenal,” Rob Norman, chief digital officer of GroupM, told Fast Company. In August 2015, GroupM signed a partnership with BuzzFeed to collaborate on projects with over a dozen brands, like Target and Nike, and gain access to Pound’s technology. “We’re spending more money with them than either of us expected,” he said.
Essentially, BuzzFeed is trying to recreate advertising in its own image: a robot that uses a combination of data and creative firepower to create brand messaging that users are powerless to resist.
But for as many reasons as there are to think that BuzzFeed is unstoppable, there are also reasons to think it might fail.
One of the biggest challenges facing BuzzFeed’s platform-agnostic approach to content and advertising is measurement. Social media platforms are still lagging behind when it comes to the kind of deep metrics advertisers have come to expect, and even with BuzzFeed’s technology, it’s difficult to track campaigns across platforms.
In fact, comScore, one of the leading measurement platforms for publications, only covers one-fifth of BuzzFeed’s reported 6 billion monthly views, according to Cooper; the rest are measured by an amalgamation of other third-party tools and platform tools.
“It is the toughest thing that we’re facing,” Cooper admitted in his SXSW keynote.
There’s also the fact that Swarm will distribute advertisements only on BuzzFeed’s owned channels. Besides limiting reach, it also puts BuzzFeed in the crosshairs of the FTC, which released much more restrictive native ad rules in December.
Traditionally, BuzzFeed has labeled native advertisements with the brand’s logo as the byline, along with the words “Brand Publisher.” BuzzFeed’s plans don’t mesh with the FTC’s desire for publishers to use terms such as “ad” and “sponsored.” Changing those labels may affect people’s willingness to read them. Labeling them according to FTC guidelines on social media will be even tougher.
One rule, for instance, states that publishers “cannot use ‘deceptive door openers’ to induce consumers to view advertising content. Thus, advertisers are responsible for ensuring that native ads are identifiable as advertising before consumers arrive at the main advertising page.”
Todd Krizelman, CEO and co-founder of ad sales platform MediaRadar, believes that some publishers may even resist the rule. “Some publishers have a good working model, so they’re not eager to disrupt that in a negative way,” he said. So far, as Krizelman predicted, the rule has largely been ignored, and it remains to be seen whether native ads distributed on social will be affected by its enforcement.
And despite positive feedback from GroupM and BuzzFeed’s considerable promise, there are reasons beyond the missed revenue targets to believe that BuzzFeed may still struggle to deliver results that will keep brands coming back. In a 2015 Wall Street Journal article, an ad buyer claimed that “only 15 percent of its clients that advertised on BuzzFeed in 2013 returned in 2014, largely because some marketers had a tough time figuring out how the ads helped their businesses.”
BuzzFeed has evolved rapidly since then, and as a data initiative, Hive will play a huge role in showing circumspect media buyers that products like Swarm are worth the investment. That will require taking Pound from a technology that excites data nerds to a technology that any media buyer can grasp.
According to Fast Company’s profile, BuzzFeed’s data-content machine is not yet ready for primetime. Internal teams still rely on Slack and Google Spreadsheets to track its content and distribution, and no one knows exactly how much content BuzzFeed creates or where it all goes. BuzzFeed’s biggest challenge will be taking the genius talent of engineers like the Kelleher twins into something that can be easily used by the thousands of internal and external people in its growing empire.
Still, BuzzFeed’s potential to transform the way the media and advertising industries work is tremendous. This is what makes it a character of endless fascination. BuzzFeed can see the future, but only if it can overcome a barrage of challenges and make its own envisioned future come true.
Ultimately, BuzzFeed’s biggest impact may come from the way it changes the philosophies and visions of everyone watching just outside its walls, mesmerized by flashing GIFs of data-science twins, trying to figure out what happens next.