Back in June, I talked with Parse.ly data analyst Conrad Lee about why stories about Donald Trump weren’t attracting the same audience online as they were on television. At the time, based on data from more than 100,000 news articles from 300 newsrooms, it looked as if Trump stories were actually less popular than stories about other candidates, including Hillary Clinton. This was before the convention and the formal nomination of both candidates of course.
Lee took an another look at the data last week and found that things have shifted.
Today the data show that Trump stories get more page views than stories about Clinton do. Both the overall number of stories and page views per post peaked around the conventions in late July, before dipping again in August.
Lee makes no distinction between positive and negative stories. Both candidates have been the subject of intense reporting in recent weeks. But this is an interesting set of data nonetheless as we think about the proper role of journalists going into the first debate next week, when 75 percent of voters plan to tune in.
Here are the questions we asked Lee this time around, along with the answers he sent back to us.
You are using headlines to figure out who the story is about. Now that we’re in the main election, with more stories about both candidates, is there something to be said here about the traffic potential of naming Trump in the headline versus Clinton?
Conrad Lee: The dashboard certainly suggests that for the last couple months articles with ‘Trump,’ and not Clinton, get more views than articles with Clinton, and not Trump. This wasn’t true back in May and June.
What kind of engagement metrics do Clinton stories and Trump stories get? How do they compare to each other?
Lee: I ran the engagement numbers for the week starting September 4th, and suprisingly articles on Trump got an average of 67.2 seconds of engaged time per user, whereas articles on Clinton got much less: an average of 40.2 seconds of engaged time per user. These number might vary a lot depending on which topics are trending in a given week, but I haven’t done a formal study to see whether that’s true.
In June, you told me Trump was getting a majority of coverage across the board, including “social-oriented publishers, conventional publishers, left-leaning publishers, and right-leaning publishers.” Have any of those patterns shifted, or have you seen others among types of publishers?
Lee: Nothing big has changed here: Trump still receives the lion’s share of articles, even from publishers that are left-leaning.
We’re also watching Facebook’s role in the election with interest. Do you see any change in Facebook traffic to political news in general, or to conservative and liberal news outlets in particular?
Lee: I haven’t studied the relationship between Facebook referrals and liberal/conservative publishers in detail. There’s not a whole lot to say here — in the dashboard you can see that there’s an especially pronounced spike of social referrals around the conventions — the vast majority of these come from Facebook. There’s a spike for all referral sources, but it’s most pronounced for social.
Finally, you also told me in June that you thought news outlets were covering Trump that because he’s easy to write articles on. Has that changed? What’s your takeaway now?
Lee: My takeaway is that before Trump secured the nomination and he was just one of many candidates, publishers were writing so many articles on him because he was easy to write on. Once he actually secured the nomination though, I think things changed. There are more legitimate reasons to focus attention on Trump now–he’s a real threat.
Jason Alcorn (@jasonalcorn) is the Metrics Editor for MediaShift. In addition to his work with MediaShift, he is an analyst with the Media Impact Project and works as a consultant with nonprofit newsrooms.