Online news outlets are interested in driving traffic to their websites. One way to do that is to get people to disseminate news stories through social media. A recent study attempts to outline which features in a news story make people more likely to share it.
Getting people to share news stories online is important to online news companies, particularly those whose revenue models rely on online visitors. A study from Columbia University and the French National Institute found that Twitter users were more likely to click on news links (and to share them) when the links were shared online by friends (as opposed to links shared by official news outlet accounts). That study highlights the importance of online news sharing (the full study is available here), but that’s not the study I want to focus on.
The study I want to focus on addresses the question of what makes a news story “shareworthy.”
The study, “From Newsworthiness to Shareworthiness: How to Predict News Sharing Based on Article Characteristics,” was published online June 20 in Journalism & Mass Communication Quarterly. The study was authored by Damian Trilling and Björn Burscher of the University of Amsterdam, and Petro Tolochko of the University of Vienna. I’ll be discussing the paper in broad terms below, but encourage you to read the full paper.
The researchers analyzed 132,682 news articles from six Dutch news sites, as well as social media activity (via Twitter and Facebook) for each story. The Netherlands has exceptionally high rates of internet access and social media use, which the study authors posit makes it a good subject for a case study like this one.
For Twitter, the researchers looked at the number of times a link to each story was shared. For Facebook, the researchers looked at the number of shares, the number of “likes,” and the number of comments on Facebook for each story. The researchers argue that likes and comments on Facebook “indirectly contribute to sharing, as an article that enjoys more activity whatsoever is more likely to be displayed in users’ news feeds.”
The researchers found that only 8.1 percent of the articles received no shares on Twitter, with 72.6 percent of stories receiving 10 shares or fewer. Only 1.1 percent of stories garnered more than 100 shares. The most shared stories on Twitter received more than 4,000 shares.
Meanwhile, a whopping 40.4 percent of stories had zero interactions on Facebook. But the most popular items had tens of thousands of interactions.
In other words, Twitter and Facebook showed very different patterns in terms of how people shared or interacted with news stories.
Now we get into the characteristics of stories that were shared (or not) on social media. I’ll tackle these in very concise terms.
Articles about Dutch (or domestic) issues got 1.29 times as many shares on Twitter as stories about non-domestic issues. That jumps to 1.8 times as many interactions on Facebook.
Similarly, stories about non-Western countries fared poorly – receiving 0.83 times as many shares on Twitter and 0.69 times as many interactions on Facebook, relative to stories about Western countries. As the authors note, “Whatever measure we use, a topic that is closer to home (because it is inherently domestic, because of geographical distance, or, in a more abstract sense, because it involves another Western and thus culturally similar country) is shared more often.”
Stories involving conflict got a small boost on both social media platforms (1.11 times for Twitter, 1.9 times for Facebook), but there was a significant difference between the two platforms when it came to “human interest” (or entertainment) stories. Human interest stories got 1.33 times more interactions on Facebook than stories about other subjects. But for Twitter, human interest stories fared no better (and no worse) than anything else.
The study also found that stories with a positive tone fared slightly better than stories with a negative tone, and that exclusive stories (i.e., those written by the news outlet) fared better than non-exclusive stories (e.g., stories picked up from AP or other wire services, which appear in multiple outlets).
In addition, the researchers found that stories about topics that were already “very present in the media” received fewer shares on Twitter than stories “that did not belong to the top issues.” Facebook was just the opposite – stories about the day’s hot topics fared better than average there. As the authors note “one interpretation would be that sharing on Facebook centers more around few dominant issues, whereas on Twitter there is more variation.”
In summary, I’ll quote the authors: “The most shares will be received by an article about the [news outlet’s] own country (or at least another Western country) and not written by a news agency. Of less, but still considerable importance, is the presence of conflict, while human interest works only on Facebook (where it has a strong influence). Regarding tone, positivity works better than negativity, especially on Facebook. The results regarding the popularity of a topic are inconclusive, as the effect on Twitter and Facebook shares is opposite.”
The study looks at news stories, and what makes them more likely to be shared. One of the things I find interesting here is that the authors did this in a systematic way. They didn’t just look at the 10 or 20 or 50 most-shared stories and try to figure out what they had in common. Their n was truly impressive (more than 130,000 stories!), drawn from six different news sites over the course of eight months in 2014.
While this is the first study of this type that I’ve seen, my guess is that it won’t be the last. I’d be interested in seeing a similar study done of large news outlets in the United States. And I wonder to what extent large news companies are not already conducting similar research on their own sites. Further, I wonder to what extent their findings are informing (or will inform) their editorial decision-making.
In other words, will studies like this one – done by academics or by news companies – be used to guide decisions about what news stories to cover and how to cover them?
The thought makes me sad. I think good journalism is essential to our well-being as a society, and that news decisions should be made based on what’s important. But news companies have to be able to pay their reporters and keep the lights on, so it’s probably naïve to assume that this sort of research (or variations on it) won’t be used in newsrooms (assuming they’re not being used already).
But what does this have to do with science communication?
That’s simple. Where does science news fit in to the “shareability” puzzle? This study didn’t look at research stories specifically. One wonders where they would land on the spectrum of shared stories. Or, at least, I wonder where science stories would land on that spectrum – and what that may mean for the future of science reporting in large news outlets.