Last week, Parse.ly published an article arguing that the media’s near-obsession with Donald Trump isn’t just a matter of chasing page views.
The article, written by Parse.ly data analyst Conrad Lee and based on more than 100,000 news articles from 300 newsrooms, seemed to go against the narrative that Trump has been a singular financial boon for the media. It’s a narrative the news media itself has helped to create. All the way back in February, for example, CBS chief executive Les Moonves said as much, stating of Trump’s campaign that it “may not be good for America, but it’s damn good for CBS, that’s all I got to say.”
Lee found a more nuanced story in the Parse.ly data. For instance, as many articles focused on Trump as every other major candidate combined, Lee wrote, but a typical Clinton article generated six percent more page views than a piece on Trump. Traffic to articles on Ted Cruz and Bernie Sanders was nearly as high. Would a journalist looking for traffic be well-served by writing more stories about Clinton? With social media referrals peaking for articles on Sanders, should journalists producing content for social media pay him more attention?
These are helpful questions for newsroom analytics and metrics-savvy journalists to be ask they continue to report on Clinton, Trump, and the election between now and November.
With the story trending in many news circles, we wanted to know a little bit more. Here are the questions we asked Lee, and his answers, which he emailed to us.
This is data for online media, but we also know, for example, that Republicans drew twice as many TV viewers per debate as Democrats. What do you draw from that comparison?
Conrad Lee: That’s an interesting statistic. To put it into context, I’d also like to know the average number of minutes each day that Republicans and Democrats spend watching conventional television (excluding on-demand content such as comes through Netflix). Perhaps this number is explained by Republicans simply watching more television, and Democrats consuming media through other channels.
With regard to the data that Parse.ly analyzed, I can say that page views were pretty equally sourced from left-leaning and right-leaning publishers.
Do the data suggest anything about how the competitiveness or calendar of the Republican campaign contributed to a higher number of Trump stories? If so, what?
Lee: The data suggest that Trump didn’t captivate the media because articles about him are simply more popular among readers; they’re not. That’s certainly contrary to the common wisdom that Trump is some kind of ‘crack’ that brings in page views, and journalists just can’t help themselves when faced with such temptation. So there has to be some other reason that journalists are focusing so intently on Trump, and I believe it’s that he’s easy to write articles on. Trump’s PR machine has mastered the art of juicy soundbites that are easy to commentate on, and make journalists’ jobs easier. That allows Trump to dominate the calendar, because regardless of what’s going on, he’ll produce the easiest-to-cover blurbs.
Page views per article for the Democratic slate are almost exactly equal to the PVs per article for the Republican slate (9,638 vs 9,623). Is that just a quirk of the data or does that tell us something about our audience?
I think that’s largely because articles on each of the major candidates are being consumed by all audiences. Note this is not the same as saying that Democrats and Republicans read the same articles; but rather, they read different articles on the same topics. It would be incorrect to assume that Republicans only read about Republican candidates, and Democrats only read about Democratic candidates.
What patterns did you find among the 300 news organizations? For example, do news organizations that get above-average social traffic publish a different mix of candidate stories?
The main pattern was that Trump received the vast majority of coverage. That was true for social-oriented publishers, conventional publishers, left-leaning publishers, and right-leaning publishers.
For the analysts and editors who read MetricShift, what is one thing that you would ask them to share with their newsroom about future Election 2016 coverage and one thing they should pay attention to in their site metrics between now and November?
Editors and journalists should be aware that Trump is not ‘pageview crack.’ On average, articles written on all major candidates bring in about the same number of page views. So you shouldn’t justify writing on the easiest topic to cover (which is usually Trump) by thinking it will bring in more page views.
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.