Do fake news sites help companies make money?
Advertisers aren't so sure, which is why they're clamping down on how their ads are spread around the Internet through marketing and media agencies.
Fake news and "alt-right" -- or racially charged or hate speech -- websites are publishing stories that seem semifactual, but really are just what's known as online "clickbait." That attracts eyeballs, which advertisers are keen to reach.
But companies are growing more careful now about where their advertisements appear -- especially since they buy "programmatic" advertising, based on computer algorithms. Those instructions send out ads across the web to reach certain demographics, but often without regard to where the ads end up.
Since the election cycle, "we have clients who now don't want to be associated with the fake news sites," said Jon Seitz, managing partner with the Mayo Seitz marketing agency in Blue Bell.
Bankers are keen to be involved in web-based advertising, said Tim Davis, head of marketing at Royal Bank America. But "there are regulations in banking that cover advertising. That's the difference: We have rules to follow."
Readers can often discern real from fake, if they want, though it takes effort. (More on that later).
Meanwhile, fake sites generate clicks by the millions from readers.
Fake news sites "take trending controversial topics and leveraging specific keywords in the headlines for eye-candy. The people sharing these types of pieces are rarely clicking and reading through the entire article, which means people are reacting to both true and false powerful headlines that play into their emotions," said Evan Urbania, CEO of the Philadelphia PR firm Chatterblast.
There's no way to bypass human nature. And computer algorithms that track ads rely on that.
"Fake news sites are writing content that ... is hitting the two strongest algorithmic needs: relevancy and engagement," Urbania said. "Fake news is absorbed because it looks and feels emotionally real," especially compared with "the more sterile tone of legacy media."
Online clicks are the currency of the internet, and they amount to big dollars.
The trade publication eMarketer expects U.S. programmatic display ad spending to reach $25.2 billion in 2016, accounting for 73 percent of all display ad spending. By 2018, that could rise to $37.9 billion, or 82 percent.
Fake news sites "are getting very good at watching human behaviors and adjusting quickly, while real news sites' goal should be to create emotional reactions based on the facts," said Urbania. "We all know from the results of the election that the American people are more emotional and more easily fired up than we realized."
Digital marketers are good at delivering a specific group of people -- "they can say, 'You want to get this ad in front of someone who makes $50,000 and owns a Volkswagen,' " said Greg Matusky, founder of Gregory FCA, a Philadelphia PR firm. "Then they auction off those ads and deliver a cost-per-click, or cost-per-view. The rate varies tremendously."
Blacklist, whitelist: Philadelphia media and marketing agencies are getting more requests from clients to crack down on "programmatic" ad-buying by using "whitelists" and "blacklists" of websites.
For instance, the American foods maker Kellogg recently pulled its ads from the blacklisted Breitbart website, after finding that Kellogg ads showed up as part of a computer algorithm that spread the cereal brand logo all over the internet.
Marketing firms recommend clients specify "whitelists" and "blacklists" -- meaning websites where advertisers want their products to show up and where they don't.
"During the election, we were being bombarded by negative ads," Seitz said. "For instance, you would see an anti-Trump ad, and right after that, Rita's Water Ice. It's a [bad] environment to be in."
"The same applies to fake news sites." It's not good for clients to be associated with "that world."
Tools to fight fake: Social media are playing a greater role in spreading news. The share of voters who tracked election returns on TV this year was similar to those who did so during the last presidential election (92 percent in 2012, 88 percent this year), according to a Pew Research Center.
On the other hand, digital sources gained ground. Voters who followed returns online rose by 14 percentage points since 2012 (from 34 percent to 48 percent), while the share who tracked results using social media more than doubled (from 8 percent to 21 percent).
Social-media sites are tweaking their tools to allow the public to change what is seen: Twitter now lets users mute entire conversation threads, as well as mute conversations based on keyword filters. Twitter users can also report offensive tweets based on hate speech.
Facebook, which has taken flak for its lack of response, announced a new partnership Thursday. If enough users report a story as fake, Facebook will pass the story to its pool of five fact-checking groups, including University of Pennsylvania-based FactCheck.org, to verify.
Stories that flunk the fact check will not be removed from Facebook. But they will be publicly flagged as "disputed" and will be downgraded in users' news feeds. Even participants say this won't fully solve the problem. And some observers have already criticized the effort for making a small group of vetters the arbiters of "truth."
Philadelphia PR firm founder Matusky of Gregory FCA has some tips on how to discern real from fake news.
"Look at the byline of the reporter. Try to find out what their background is, before I start following them. Read the comments. That's the fastest way to tell if something is left or right and it can give you good insight."
In some ways, "the whole fake-news phenomenon is not new," he says.