View profile

Fixing Facebook with a labelmaker

Is misinformation a problem we can label our way out of? Two initiatives announced today promise to m
March 5 · Issue #96 · View online
The Interface
Is misinformation a problem we can label our way out of? Two initiatives announced today promise to move us in that direction. After doing some interviews, though, I worry cosmetic touches like labels may prove to be distracting at best — and counterproductive at worst.
Let’s start inside Facebook, where the company will begin to let a small group of publishers add a “breaking” tag to stories they publish on the platform. They’re allowed to use the tag once per day, with five bonus “breaking” indicators to be used each month at their discretion. Once it’s added to a web link, and Instant Article, or a Facebook Live, the indicator can remain live for up to six hours. 
Here’s product manager Joey Rhyu:
Getting clear signal on what stories are breaking news can help. We’ve been running a test in the US that lets a small group of local and national publishers identify and label breaking news. Starting this week, we will expand the test so that more than 50 additional publishers in North America, Latin America, Europe and Australia will be able to label their stories as breaking news on Facebook. If the expansion is successful, we may add more publishers.
Rhyu says that the “breaking” tag leads to more engagement, with a 4 percent increase in clicks, a 7 percent increase in likes, and an 11 percent increase in shares. But the data he cites is from December 8 to January 14 — the period just before Facebook adjusted the News Feed to reduce the volume of news. And even if engagement holds up in the new era, no publisher is going to get rich off their bonus “breaking” traffic.
The more positive development here is that Facebook made an editorial choice in selecting 50 trusted publishers to get access to the tool, which will maybe drive marginally more attention to more important stories shared on Facebook. If publishers can earn trust based on their accurate reporting, and get access to a larger share of Facebook’s audience as a result, it could be a positive development.
The second set of labels comes from outside the company. Steven Brill, the polymath journalist and entrepreneur, announced today the launch of NewsGuard, a company that plans to hire dozens of journalists to evaluate the quality of 7,500 news organizations and then sell that information to the tech platforms. (The basic conceit of NewsGuard, which is that Facebook should pay journalists to tell them what is journalism, is my favorite aspect of NewsGuard.)
The good thing about social media is that it dramatically increases the number of sources from which a person might learn about the world. The bad thing is that platforms’ one-size-fits-all design means that a fake site like the Denver Guardian looks identical to a real one like the Denver Post when links are shared on Facebook or Twitter. Brill’s gambit is to try to return some of that missing context to our feeds, via data products that are purchased by the tech companies. 
NewsGuard plans to produce two main widgets for Facebook and other platforms. The first is a label of “red,” “yellow,” or “green,” which corresponds to the general accuracy of the site. (Green and red are fairly self-explanatory — think Denver Post vs. Denver Guardian. Yellow is for advocacy-oriented sites that don’t disclose their financial backers, such as the industry-sponsored What Is Fracking? site.)
The second product the company calls a “nutrition label.” Its job is to provide context about the publisher: how old it is, who owns it, and what it covers. The label will also explain the site’s green, yellow, or red ranking. Brill plans to have labels for all 7,500 sites, representing 98 percent of Facebook publishers, in place before this year’s midterm elections.
I spoke today with Brill and his co-founder, former Wall Street Journal publisher Gordon Crovitz. Crovitz said NewsGuard would fill a niche that was both necessary to promote a healthy information economy and impossible for platforms to do themselves. 
“The message we got from them was, they do not want to be in the position of putting their finger on the scale with regard to specific brands,” Crovitz said. “That’s completely understandable — they want to be neutral.”
But can labeling help? Interface readers know that Facebook’s own research showed that labeling stories as false help them spread more widely. Won’t NewsGuard simply reinforce that behavior?
Brill said that labeling at the publisher level, rather than the story level, would prove to be more effective. “It’s one thing to have several dozen headlines in a News Feed and one of them has a red dot, and you’re sort of curious about it, so you open it up,” he said. “It’s quite another thing to have 10 headlines, and nine of them are green, and one is red, which instantly puts the red in context.”
Brill’s larger point is that it’s easier to police publishers than it is to police individual stories. The next time a Denver Guardian pops up during a news cycle, Brill’s team plans to swarm it like a SWAT team. “We will have a team that will get alerted to any suddenly trending site, and we will be working on a 24/7 basis — and we will wrestle those down in an hour.”
It’s a worthy goal, but I can’t help feeling like the whole thing feels a little flimsy. No platform has yet signed on to license NewsGuard’s data, and even if they did, it’s unclear what practical effect it would have in reducing the spread of misinformation. 
Labels are easy to create and deploy, which helps explain their appeal to Facebook and startups alike. My worry is that they are too easy to create and deploy — and what looks like real action we may decide in retrospect was only dithering.

State Dept. Was Granted $120 Million to Fight Russian Meddling. It Has Spent $0.
China Presses Its Internet Censorship Efforts Across the Globe
Russian Facebook Accounts Had Millions More Followers Than First Reported, New Research Shows
Conservative publishers hit hardest by Facebook News Feed change
Reddit says Russian propaganda was shared by ‘thousands’ ahead of the 2016 election
Here Come the Fake Videos, Too
Brussels’ move on digital taxes raises transatlantic stakes
Facebook says survey about condoning pedophilia ‘was a mistake’
The man who stole Frances McDormand’s Oscar streamed it to Facebook
Facebook pitches brand-safe video ad buys for $750,000, but lack of control irks buyers
Snapchat’s redesign is baffling publishers
YouTube is adding a new AR green screen effect to its stories feature
The Case Against Retweets
The last joke is Tide Pods
The platform is the message [PDF link]
The Supreme Court Case That Could Give Tech Giants More Power
And finally ...
Marriott Employee Roy Jones Hit ‘Like.’ Then China Got Mad
Talk to me
Would you label this newsletter red, green, yellow, or BREAKING?
Did you enjoy this issue?
If you don't want these updates anymore, please unsubscribe here
If you were forwarded this newsletter and you like it, you can subscribe here
Powered by Revue