The Science of Buzz

I'm off in sunny San Diego for a couple days at a Predictive Analytics conference, where I'll be giving a short talk tomorrow on pricing analytics.

Today, however, I'm just a listener. The interesting thing about the conference being focused on analytics is that while many of us use similar methods, we're working in very different industries. Two of the more interesting talks I heard this morning were by data scientists from Mashable and BuzzFeed, each talking about how his company uses data analysis to decide what to write about and which stories they write to promote.

I'll admit, I regard these kind of sites with disdain, since I'm a relatively snobby reader most of the time, so I'd never really thought about how it is that these traffic driving sites (who are happily sucking the oxygen out of the old media world) decide what to write about. Apparently they both actually use pretty sophisticated analysis of what stories are trending and what type of stories trend how much to look for stories that are in the early stages of breaking, and then they jump on the bandwagon.

The fellow from Mashable gave a fascinating example in which one of their Australian writers noticed the other day, using the company's dashboards of stories that are in the early stages of trending on social media, that there was an Australian story picking up steam about how a 109-year-old man knits sweaters for penguins. Their data tools suggested that this was a story which not only was trending well in Australia, but that if it was picked up by a global site like Mashable it could become a world-wide trending story. So she wrote a Mashable story re-telling what local news had already covered and providing some pictures and links, and it immediately took off. Now if you search for the story on Google, you'll see that everyone else is going the same.

Part of what drives this whole approaching to producing and circulating content is that most of the clicks to content sites these days are driven by shares on social media: Facebook, Twitter, Pinterest, etc. People may still go to some single sites they like, whether it's The Atlantic or DarwinCatholic, and read down to see what's new, but significantly more visits come from shares. And visits is what allows sites to show advertising, which in turn is what makes the sites money.

Mashable and BuzzFeed may have a pretty sketchy reputation as "news" venues, but they are pretty good at driving lots and lots of clicks, which means that in the new media world they are growing, and they're by following the mission which the BuzzFeed speaker showed: "We accelerate the promotion of content that is taking off"

In other words, if it seems like these sites often aren't producing quality, thoughtful content, that's not big surprise. Their goal isn't necessarily to do that, it's to drive clicks. And the most efficient way to do that is to jump on an already accelerating bandwagon.

Another example provided was the Lord of the Rings/The Hobbit themed airline safety commercial that went viral a while back. The airline, of course, had produced it, the video was available on YouTube, but what Mashable, Buzzfeed, and even a lot of "respectable" news outlets was simply to throw up stories which embedded the YouTube video, so that people would share their story instead of a direct link to YouTube, thus netting them a lock of impressions simply by riding the coattails of something that was taking off anyway.