Source Code: What matters in tech, in your inbox every morning

×
×

Sign up for Protocol's newsletters to get insights on the people, power and politics of tech.

Not today, thank you!

Will be used in accordance with our Privacy Policy

YouTube gives advertisers lots of control about where their ads show up. Other social platforms don't.

Image: David Pierce/Protocol
Power

Why the social ad boycott didn't come for YouTube

Research shows that YouTube faces a hate-speech problem as rampant as any other big social media company, and it's often included in debates over platforms' handling of controversial content. But not this one.

Last week, in a Facebook group for far-right conspiracy supporters called "Q-Anon Patriots," a user posted a video that drew on hateful and antisemitic rhetoric. It's titled "FEMA TRAINS, REX-84, GULLITINES, MARTIAL LAW, FEMA COFFINS In The USA." Next to that video: an ad for Verizon Wireless, selling the Galaxy S10.

The Anti-Defamation League spotted the odd pairing of ad and video and included it in an open letter that also showed this was not the only group, or the only post, or the only placement that advertisers might find problematic.

Sign up for Protocol newsletters

Verizon announced that it would halt all of its advertising on Facebook and Instagram hours later. "We have strict content policies in place and have zero tolerance when they are breached, we take action," Verizon's chief media officer, John Nitti, told CNBC in a statement. "We're pausing our advertising until Facebook can create an acceptable solution that makes us comfortable and is consistent with what we've done with YouTube and other partners."

In doing so, Verizon joined a growing list of companies that have announced plans to pause advertising on social media platforms in response to untended hate speech online. So far, the ad boycott has largely targeted Facebook, Instagram and Twitter.

Noticeably absent from the list? YouTube.

Research shows that YouTube faces a hate-speech problem as rampant as any other big social media company, and it's often included in debates over platforms' handling of controversial content. But not this one. Verizon is among only a handful of companies to even mention YouTube in their announcement. And it named YouTube not because of the current boycott, but because of a past one.

In 2017, dozens of companies worldwide, including Verizon, Pepsi and Walmart, halted their ads on YouTube following an investigation by The Times that showed paid advertisements had run alongside content that promoted hate, including terrorism, racism and anti-Semitism. To quiet the boycott, YouTube made significant changes to its ad policies, which gave advertisers a greater sense of understanding and control over exactly where their ads go. The policies that resulted from the boycott against YouTube offer a glimpse into the changes that could lie ahead for Facebook, Twitter and other social platforms currently under fire.

Today, YouTube advertisers have the option to adjust "content exclusions" within the dashboard of their Google Ads account, at both the account and campaign level, in order to steer clear of certain types of content such as violence, profanity or even animal mating. "The intention of these policies is to prohibit ads from appearing on pages or videos with hate speech, gory or offensive content," Ronan Harris, a managing director at Google U.K, wrote in the company's public response to the 2017 ban.

The three groupings of content exclusions offered to advertisers. The three types of content exclusions offered to advertisers.Image: YouTube

Advertisers can adjust "Inventory Type," which offers broad exclusions in a Goldilocks-and-the-three-bears style: Expanded, the least-selective set that only keeps an advertiser away from extremely sensitive content; Standard, the common-sense option recommended for most brands; and Limited, the smallest set with the most content filtered out.

The categories of content (accessible through a Google Ads account) that advertisers can exclude on YouTube.Image: YouTube

Advertisers running Display campaigns can get more specific from there, using a simple checkbox interface to opt out of additional "excluded content" categories: tragedy and conflict; sensitive social issues; sexually suggestive content; sensation and shocking; and profanity and rough language.

But there's more: Similar to movie ratings like G and PG, Google says it grades content based on its suitability for various audiences and gives advertisers the ability to exclude by rating. Advertisers can also exclude content by topic, site URL and keyword. They can also choose not to be next to livestreams, which are typically harder to police in real time. After all of that, advertisers can run placement reports to see which YouTube channels and videos their ads have been shown on.

YouTube relies on a combination of machine learning and human input to review content for advertising purposes, and Google has found its category exclusion to be 90% accurate. In April, YouTube also rolled out a self-certification program for creators to self-report on the content of their videos to reduce errors from automated reviews. "It's a reinforcing process: The more accurate you are in your self-reporting, the more our system trusts you," YouTube CEO Susan Wojcicki wrote in a blog post to creators in December, when the feature was still being tested.

At Facebook, on the other hand, placement is focused on real estate and audience, not specific kinds of content. Advertisers are given the option to show ads in Feeds or Stories. Within Feeds, advertisers can choose placement in Facebook News Feed, Instagram Feed, Facebook Marketplace, Facebook Video Feeds, Facebook Right Column, Instagram Explore and/or Messenger Inbox. There's no way to exclude the type of content next to which the ad will appear within those placements. Anything that meets Facebook's platform rules — or slips past them — is fair game for all advertisers. "While we apply brand safety controls as effectively as possible," Facebook's guidelines say, "we can't guarantee that all content and publishers will be compliant or aligned with your unique brand safety standards." (The Facebook Right Column placement seems to be the highest-risk option, as that opens the gate for placement alongside a Facebook group — hot beds of hate speech and misinformation on the platform.)

Another change Google made in response to the 2017 boycott was to tighten requirements for its YouTube Partner Program. It now keeps creators from monetization until they've reached 1,000 subscribers and 4,000 watch hours in a year's time, and passed an activity review against YouTube's policies and guidelines.

Content creators at the time lashed out at YouTube's expanded control over which videos can make money from ads. Popular YouTuber PewDiePie called it "the adpocalypse." PewDiePie, having joked about anti-Semitism and Nazis months prior, was one of the creators to see revenue drop as a result of the changes.

But YouTube didn't drop PewDiePie for his questionable statements. In fact, it has only since expanded his influence. Earlier this year, YouTube signed its first exclusive livestreaming deal with the creator. It underscored how well YouTube has achieved its intended goal with policy reform: A range of content, including sensitive content, can continue to exist (or even thrive) on the platform, but cautious advertisers need not worry about being placed next to content misaligned with their brand.

Advertisers and social media platforms will surely look to YouTube's 2017 changes for guidance about where to go next. Already, Mark Zuckerberg announced changes to Facebook in response to the current boycott, including labeling problematic content and cracking down on hate speech in paid ads.

Just as Verizon returned to YouTube following its boycott in 2017, it (and others) will in all likelihood return to Facebook after the platform institutes changes. Many brands are more interested in looking good in the current news cycle than in forcing social media platforms to make sweeping changes related to the content they allow overall, said Adam Kleinberg, head of San Francisco ad agency Traction Corp. "Simply having their brands on Facebook while there's this boycott in the news, it's negative content adjacency inherently. Many brands are looking at the current economic climate, and making cuts to budgets is something they want to do anyway."

Case in point: The post that led Verizon to pull advertising from Facebook last week was an embedded YouTube video. While Verizon didn't want its advertisements to sit next to content deemed harmful and anti-Semitic in the Q-Anon Patriots group, the company didn't call for the video itself to be removed from the platform. Social media companies may continue to make changes to their policies in response to boycotts. What changes, exactly, will depend on the outcomes advertisers really care about.

Latest Stories