YouTube is Asking Creators to Self-Regulate for Brand Safety
With brand safety a top concern amongst YouTube advertisers, the video platform is asking its creators to self-monitor for brand-unsafe content.
YouTube has tested this practice with 15 creators, giving them questionnaires to fill out while uploading a video. These questionnaires give creators the opportunity to declare any potentially inappropriate content before the video hosts any advertisements. This experiment will determine whether forms like these can reinforce the ability of both human reviewers and computer systems to classify inappropriate content. If successful, YouTube will use them to help identify which videos should not carry ads, as well as mitigating cases of videos being mistakenly demonetized. The video platform stated plans to expand the test to a few hundred more creators in the near future.
The questionnaire, which is currently eight questions long, tries to identify not only undesirable content but also its context. Profanity, for example, is allowed under the circumstances that it is “non-hateful, comedic, or artistic”. It can also be used in the context of a documentary or news clip. Creators can also describe on the form whether profanity is used “repeatedly in a vulgar or hateful context.”
YouTube still reviews each video after the questionnaire has been completed to ensure creator honesty, but so far, the creator-provided classifications have been a close match to YouTube’s own. “The majority of those self-certifications were very consistent with the determinations we would have made internally with our human reviewers,” said Tom Leung, director of product management at YouTube, in a video published earlier this month detailing the test.
Unfortunately, the tests come too late for some advertisers.
300+ brands were recently found by CNN to be running against videos promoting things like racism and pedophilia.
Many advertisers already don’t trust YouTube’s review system, and choose to remain within the safety of their whitelists of approved channels. Even so, they may still rely on YouTube’s discretionary system to decide if the individual videos on those channels should carry their ads. And some choose to avoid the risk entirely, like Chase Bank, which instituted its own system to avoid controversial videos, or like Proctor & Gamble, which ended its year-long boycott of YouTube only last week.
As of now, YouTube has demonetized a relatively small percentage of videos. Rising pressure for brand safety may push them to raise the bar, but that could risk shrinking the supply of content, increasing ad rates and potentially scaring off ad buyers. Using this questionnaire, Youtube is attempting to strike a balance for videos that may fall in a grey area of brand safety concerns.
“They’ve demonetized a ton of content that shouldn’t be getting ads. But what’s left is still highly nuanced and subjective in the eyes of any particular brand. What’s left is largely OK to monetize broadly, but maybe only 10 percent of YouTube is OK for anyone [to advertise against],” said Mike Henry, CEO of video analytics firm OpenSlate.
Using this questionnaire, YouTube may be able to fill some of the gaps in its existing content review process. As more creators use the form, they can establish a track record of trust, which the review system could consider when making monetization calls. The information gathered from the questionnaires can also be plugged into YouTube’s automated review system, which employs a machine-learning model that can use the data to improve its own ability to review controversial content.
“It reveals some of the potential limitations, or at least the need to approach the issue from all angles. It can’t just be an algorithm. It can’t just be human moderators,” said Mike Dossett, vp of digital strategy at RPA. If the video platform were reliant solely on self-certification by content creators, it would be a concern, he said. “But because it’s paired with these two other safety mechanisms, I’m less concerned right now.”
For more tech news and advice, visit CROWD. or send an email to firstname.lastname@example.org.