Month: February 2017

Breitbart explained

What can data tell us about the new voice of the White House

The White House is narrowing the list of media sources it wants near to it and chose a select few to attend last Friday’s press briefing.

The digital newcomers Breitbart, One America News and The Washington Times were all invited. Journalists from ABC, CBS, WSJ, Bloomberg, and Fox News were there, too. Some of the traditional outlets such as The New York Times, The Washington Post and CNN were excluded from the event.

Given the President’s support of this new group of media sources we must understand what makes up that lens through which the public will see its messages. Breitbart is a good place to start.

The most successful stories published last week by Breitbart that Kaleida is tracking include:

  1. Spicer to Reporter: ‘We’re Going to Raise Our Hands Like Big Boys and Girls’ — 28,840 shares on Facebook
  2. Steve Bannon Details Trump Agenda: Deconstruction of the Administrative State — 12,370 shares
  3. Witches Unite to Cast ‘Binding Spell’ on Trump and Followers — 9,493 shares
  4. Report: NBC News Behind ‘Access Hollywood’ Video Leak to Hurt Trump — 10,595 shares
  5. Caitlyn Jenner calls Trump transgender decision ‘a disaster’ — 11,703 shares


Looking at those headlines it’s easy to see why many in the media business dismiss Breitbart as a news source. It doesn’t seem to take itself very seriously based on what resonates with their readers, or maybe it’s the readers that consider Breitbart lightly. But the President of the United States is a serious supporter of Breitbart and those numbers are very strong.

It would be foolish to dismiss Breitbart. Instead, let’s see what it’s all about.

Kaleida data shows that Breitbart articles generally outperform the market averaging nearly 4,500 shares on Facebook. That’s three times the overall average of about 1,500 shares per article for the publishers we track. CNN is bigger averaging 6,500 shares per article. The New York Times is just over 5,000. And The Washington Post is lower averaging under 3,000 shares per article.

“Fake news” is a big topic for Breitbart. There are over 130,000 references to the term across their site compared with 42,000 for Fox News, 11,000 for CNN, and 162,000 for NYT.

They accuse competitors of producing fake new while producing a lot of it themselves. In fact, Breitbart’s most successful article clearly falls on the fake news spectrum with an intentionally misleading headline. This piece, “President Obama Awards Himself Distinguished Public Service Medal” earned a remarkable 200,000 shares on Facebook.

Yet they don’t support the core assertion. They provide quotes, a link to an AP story and several tweets which create the appearance of evidence, but those references do not backup what is stated as fact in the headline.

Secretary of Defense Robert M. Gates awards the medal to President George W. Bush. Photo by Tech. Sgt. Jerry Morrison, released by the US Navy.

Breitbart tapped into a strong feeling by many Trump supporters that Obama was weak on military leadership with this article. The feeling represented by the assertion may be a real feeling, but feelings are not news. It discredits Breitbart as a reliable source of information. Even the belief that the story could be real is hard to defend when records show Presidents on both the left and right received the same award in the past.

There are many examples of media on both the left and the right fueling agendas through their reporting and use of language. This example is a particularly flagrant use of a factually incorrect headline that intentionally misleads Breitbart’s own readers.

There can be no doubt the headline helped Breitbart’s advertising model given the scale of the response. By comparison NYT had only 15 stories with more shares during the same period. CNN had 5. Fox News’ biggest piece earned 166,000 shares.

Other Breitbart stories have broken the 100k shares ceiling such as this story about Ivanka Trump’s perfume (106,245 shares) and this one about Madonna’s comments about Trump (116,275 shares). Of the top 25 stories all but one are about Trump either in the form of a Trump conflict with a celebrity or group, a Trump family member’s successes, or the end of Obama’s presidency.

By contrast, of the top 25 stories for Fox News during the same period only 10 are explicitly about Trump. An equal number are news articles about military issues or sporting events. Less than 5 are opinion pieces.

Breitbart’s followers on Facebook are eager to spread the news, too. Many Breitbart articles sit on their web site mostly unseen until they promote them on their Facebook brand page.



Their biggest story right now lived on the site for 13 hours without getting noticed. Then it soared immediately after a Breitbart Facebook post, presumably driving considerable traffic to their web site, too.

According to SimilarWeb their web site traffic is experiencing some impressive numbers.

  • Total Visits 108M, up 26.26%
  • Avg. Visit Duration 00:03:28
  • Pages per Visit 2.31
  • Bounce Rate 57.86%

There’s still a lot we don’t know about Breitbart, yet.

We don’t have figures tracking more subjective things like editorial standards. We aren’t aware of an ombudsman at Breitbart. We know little about their policies around corrections, disputes, fact-checking, image verification, or many of the things professional news organizations invest in and think about.

We don’t know much about their policies around advertorial, conflicts of interest or their social media policies and their approach to comment moderation. Though Breitbart states clearly in their terms of service that they don’t permit content that is “false, misleading, libelous, slanderous, defamatory, obscene, abusive, hateful, or sexually-explicit”.

Breitbart’s Senior Editor voiced support for paedophilia on an Internet talk show recently. He has said many things that would fall into one or more of those categories and has now left the organization.

Being so close to the inner circle of the President of the United States Breitbart may require more rigorous standards for their staff as well as their commenters.

Breitbart is still very new.

In fact, it wasn’t long ago that Pew’s Research on Media Habits showed that only 15% of the news-reading population had even heard of Breitbart, and only 25% of conservatives who knew of them considered Breitbart trustworthy. That’s compared with 88% of conservatives who said Fox News was trustworthy.

Those numbers have surely changed now, though it’s unclear how much the center-right is interested in the type of coverage produced by Breitbart. Fox News is strong with that audience and has more resources to dedicate to their journalism and the challenges of providing reliable information.

Breitbart’s team appears to be very small. LinkedIn shows only 25 people employed at the company which includes mostly reporters and editors. None of the people listed are copyeditors, researchers, lawyers, marketing or sales people. The business appears to be designed for one thing — publishing content quickly.

According to Crunchbase’s startup database Breitbart raised $10M from investors in 2011, though Crunchbase does not indicate who the funders are. Funding since then is unknown.

There was a time not long ago when Buzzfeed was making similar waves across the media industry, including questions about its reliability as a news source. LinkedIn shows over 1,500 people on staff. They’ve raised about $450M from investors. And they break major political stories with serious impact.

While it would make sense to compare Breitbart with Fox News given the conservative agenda, there might be more in common with Buzzfeed. Serious coverage happens, but their bread and butter is entertainment.

Based on the data we have available Breitbart appears to be a sort of early stage, political gossip version of Buzzfeed focused exclusively on serving Trump supporters. It is evolving quickly from unknown Los Angeles blog to White House messenger.

Their story will have many lessons to offer the rest of the industry, though it might be smart to stand back and watch how they cope when flying so close to the sun.

Trump’s declining influence on public discourse

Opponents are using his own tactics against him as his domination of the wider conversation fades.

I was reluctant to write another analysis piece on Donald Trump this week. There’s a lot of other fascinating stuff to talk about, but it’s very hard to look anywhere else at the moment.

And then I saw a Trump trend that actually seemed genuinely interesting — his falling presence in public discourse.

Source: Kaleida, Feb 2017

We’ve tracked over 25,000 articles about Donald Trump over the last three months from about 20 leading publishers in the US and UK. Trump articles average about 4,500 shares on Facebook compared with 1,700 shares on average for all other articles. That rate peaked at 6,700 shares per Trump article the last week in January.

Now that his ratings have fallen performance is running at a normal-ish level, though he will surely climb again. When they do he will have a long way to go to reach Obama’s levels. Obama coverage peaked at 9,500 shares per article and averaged 5,500 over the same 3-month period we’re looking at here. Some of that was clearly related to Trump, but Trump performance figures surely benefitted from Obama’s draw, as well.

This is where things can get really interesting.

Kaleida can identify both the subjects with high correlation and those with a noticeably significant relationship to Trump. For example, the White House, Russia and Muslim all have high correlation scores. They get mentioned together a lot. This week Shinzō Abe, Nordstrom, Saturday Night Live and now Michael Flynn are showing a significant relationship to Trump. Those subjects are outperforming the norm.

Trump seems to account for over a third of public discourse most of the time. Then by including related subjects his numbers get closer to 50%. The last week in January Trump and his network accounted for nearly two thirds of all coverage shared on Facebook.

We discussed this idea of the Trump lift before. He uses it as a weapon. Trump’s tweet about Lockheed-Martin in December cut $4bn in value off their stock.

But Twitter is a conversation, and those opposing Trump’s agenda are learning how to play Trump at his own game.


US retailer Nordstrom recently cancelled their contracts with Ivanka Trump and her fashion line, apparently due to poor sales. They may have used the opportunity to troll the President knowing that his influence might help them. And, in fact, Trump’s aggressive response to the company inspired a massive surge in Nordstrom’s stock price.

Trump knows a thing or two about trolling powerful people and using their position to garner support for his own agenda. Perhaps more people were paying attention to his antics than he realized. Kmart and Sears have both joined in the game this week.

If Trump is able to build more strength across a network it could be harder to beat him at the attention game. He is in a very powerful position. But the same tool he is using to build that position could be the one that strengthens his opponents.

The name of the game at this point is trolling. It may not last. But, then again, it may have just started.

Is terrorism underreported? Let’s look at the numbers

We were planning to launch Kaleida today with a press release. Conveniently, Trump’s latest claim gave us a reason to demonstrate the power of our data and research tools instead.

This week the President of the United States accused mainstream media of insufficiently covering terrorist attacks. Kaleida’s new analytics and insight tools which we are launching today show clearly that this is not true.

How do we know the media isn’t ignoring terrorism? Let’s look at the numbers.

The most obvious example in the list of underreported attacks provided by the White House was the December attack in Berlin at the Christmas Market. There were hundreds of stories from across all major media outlets about that event. We estimate nearly 500 articles were written about it amongst the 20 sources we track at Kaleida over the course of 2 weeks, and countless others from sources we haven’t yet begun working with.

On the day of the event the New York Times wrote at least 12 stories, The Washington Post did about 18, we counted 19 from CNN, Reuters posted 21, and Fox News published 31. Stories continued rolling out for several days afterwards.

Many sources posted liveblogs and reports as information became available. Many were talking to victims. Some were following the hunt for the suspect. And others were writing about the context of the event and the state of terrorism in the world.

Facebook provides proof that people saw all this news. There were well over 1 million shares of those stories across their platform which then appeared in news feeds for the friends of all those people who shared. Most coverage averaged about 1,500 shares per article.


CNN had the biggest story on the day with 45,530 shares, and Fox News’ story about the suspect being killed in Italy earned 41,202 shares. Our records show that Fox promoted it in their top editorial slot on their home page for over 31 hours. Everyone was leading with that story.


Trump’s comments about the attack were covered, too. The Washington Post noted that the German authorities were asking people not to spread rumors in the early moments after the event. The President seemed to know who was behind the attack before the media did and made a statement saying, “ISIS and other Islamist terrorists…regional and worldwide networks must be eradicated from the face of the earth.”

That story was promoted from the top editorial position on washingtonpost.com for an hour. CNN ran a story quoting Trump, too, “President-elect Donald Trump said Wednesday the rampage in Berlin was an ‘attack on humanity.’”

Despite all that Trump’s voice in the coverage of the Berlin attack was not very loud. His voice was included in the coverage, but it’s likely that his fans didn’t hear him.

Our research shows that middle-aged men who don’t trust the news, who have right-leaning politics and lower education levels were not very interested in coverage about Trump and the Berlin attack.


We have more data about media coverage of the terror-related events on Trump’s list — the Minnesota Mall stabbings, the Ohio State shootings, the ISIS attack in Jordan, and the Tunisian beach attack.

Many of the events on Trump’s list happened before Kaleida began tracking what’s happening in the world and who is interested in what and when. However, there’s nothing in our data to suggest that the leading publishers on both the left and right sides of the political spectrum only began covering terror events in September 2016. So it’s very likely all the mainstream news orgs have been covering terror attacks in earnest during the time period in Trump’s claim.

News orgs often cover terrorism at great cost with little or no obvious commercial benefit.

Curiously, the media’s coverage of Quebec City’s terror attack by a white supremacist wasn’t on the President’s list. We tracked nearly 100 articles about that attack. And the December attack on the Russian Ambassador to Turkey in Ankara is not on the list, either. There were hundreds of articles about that attack, too.

It’s not clear why Trump believes those attacks were sufficiently covered versus the others that weren’t covered well enough. Did he want the media to look away from those attacks? Or were they omitted by mistake?

The President of the United States may have spoken from the heart before doing any thoughtful analysis.

Our data says the opposite of his claim that terrorism and terrorist attacks in the United States and Europe have “gotten to a point where it’s not even being reported.”

Politifact rates Trump’s statement “Pants on Fire”:

“We found no support for the idea that the media is hushing up terrorist attacks on U.S. or European soil. The media may sometimes be cautious about assigning religious motivation to a terrorist attack when the facts are unclear or still being investigated. But that’s not the same as covering them up through lack of coverage. There is plenty of coverage of in the American media of terrorist attacks.”

We would agree with that.

Many of the assertions Trump makes about the media seem more like feelings than thoughts or insights. After seeing the data a reasonable person would come to a different conclusion.

Equally, we appreciate that people see the world through different perspectives. It’s not wrong to feel that media is failing to do the best it can do. We feel that way, too, which is why we founded Kaleida.

Our hope is that Kaleida can make our differences and our similarities much easier to see and to help us all create a healthier media ecosystem that fuels diverse perspectives. Today’s launch is a big step toward that goal.