Digital-POV: YouTube Brand Safety Updates, June 2018
Following last year’s well publicised brand safety issues, Google has announced several changes to its platforms to address the concerns raised by advertisers and agencies. This week, it announced the following changes:
Details and Implications:
Google announced updates to efforts to improve brand safety on YouTube, including details of a new three-tiered inventory structure for YouTube True View. In addition to enhancements to human verification and higher channel monetisation thresholds, the new structure includes Standard, Expanded and Limited inventory tiers designed to better align with individual brands’ risk tolerance profiles.
- Expanded Inventory: This setting includes “Videos that focus on sex…without visible contact or nudity” and “expletives/profanity used multiple times in the context of comedy, documentary, news and education”.
- Standard Inventory: This is the new default setting that allows for “moderate profanity used in a non-hateful comedic or artistic manner” and “moderately suggestive behaviour or dancing as well as educational discussion of sex, artistic nudity and discussions of sexuality and relationships”. This new setting eliminates approximately 10% of the inventory available in the Expanded Inventory setting.
- Limited Inventory: This setting will further reduce these two categories to what Google calls ‘light or very fleeting strong” use of profanity and “light romance, kissing, or general discussions of relationships and sexuality”, reducing overall inventory by 30% vs the Expanded Inventory setting.
This update represents another step forward in Google’s efforts to clean-up the YouTube ad experience. Coupled with expanded relationships with expert NGOs, increased human content review, decreased channel monetization and expanded third-party monitoring, this structure seeks to offer clients flexibility in determining their own risk tolerance.
However, Google is yet to address some other issues in the True View model:
- Inventory default settings continue to rely on Google’s content evaluation as opposed to independent third-party evaluation and classification.
- Google content avoidance categories are less comprehensive Vs 3rd party controls; Google has limited or no controls in Standard IAB Avoidance Categories such as Fake News, Spam or Harmful Content, Online Piracy and Death or Injury.
- New brand safety integrations with IAS, Double Verify and soon Moat are limited to post campaign reporting of any brand safety infractions, after any reputational damage will have been done.
- YouTube remains the only inventory source that can’t be pre-bid filtered in DBM.
- Google AI is still having difficulty determining context in music lyrics promoting violence.
We continue to recommend activation carefully apply maximum internal YouTube brand safety settings for Digital Content Labels, Content Types, Sensitive Category Exclusions and Placement Exclusions as recommended in the Mindshare Global DVA guidelines.