The news cycle is tumultuous, fake news is still proliferating online, and it probably won’t surprise you to learn that adult content is still pretty popular. Brand safety challenges abound and for marketers who need scale in their campaigns, the risk of winding up next to something unsavory has never been greater. We organized a webinar to explore the new boundaries of the brand safety problem and to share some of our accumulated wisdom for how to keep your brand safe from risky content. These are some of our favorite questions from the Q&A session that followed:
1.) Can you share more about the brand safety beta between YouTube and IAS? What have beta clients said so far about their experience?
As of today, more than 90 advertisers across a variety of verticals including Telecom (Verizon) and Consumer Health (Bayer) have been invited to participate in this closed beta for brand safety protection across both auction and reserve. Participants are able to monitor their campaigns across five brand safety categories including Offensive Language and Violence. Using the insights gained from IAS reporting, these brands are able to customize their Brand Suitability settings on YouTube to meet their specific needs.
Feedback has been overwhelmingly positive and clients in the beta plan to roll it out across additional markets as well as recommend the integration be used across their brand worldwide. Many have reported that the video level insights, in particular, have helped them to meet their YouTube brand safety and suitability standards.
2.) How does IAS monitor brand safety?
We dynamically analyze content based on multiple variables in the environment such as copy, inbound and outbound links, files, and metadata inputs. We continuously score and rescore web pages based on risk categories and thresholds. This approach ensures that if the environment changes, our models will still protect brands.
Traditional models, such as blacklists, tend to be limited and inefficient. Plus, this method eliminates entire domains, rather than the specific pages within a given domain that pose harm to brands, which could negatively impact the ability to scale.
3.) How long do you recommend brands block keywords based on violent events? For instance, is the term “Virginia” still risky following the violence surrounding the Unite the Right rally in Charlottesville last year?
We believe that each individual brand should be empowered to decide what they consider acceptable. We’ve heard agencies recommend that brands review keyword lists on a monthly basis. As we mentioned during the webinar, brand safety is not a set-it-and-forget-it problem. It requires vigilance and rigorous monitoring, and that includes the keyword lists you create to protect your brand.
4.) How do you control for false positives on keywords? For example, the sentence “PG Steph Curry shot poorly last night,” contains the word shot, but isn’t about violence. Similarly, how do you avoid blocking the verb “trump” while blocking against Trump the person?
We recommend that you use IAS’s models as your first line of defense, and then to strategically consider which keywords you chose to additionally block on. Blocking on keywords is risky and requires thoughtful consideration, especially if words can have multiple meanings such as “trump” but it is still extremely valuable in ensuring your brand doesn’t show up near content that is against your brand guidelines. If you’d like additional guidance, please check out our blog, Avoiding false positives: Strategic implementation of keyword blocking.
5.) News publishers have an obligation to keep their audience informed of world events, even if those events are violent, concern objectionable people, or touch on adult themes. This can be a challenge when brand safety is concerned. However, news publishers have actually seen positive results when brands surround breaking news content because audiences are highly engaged. Do you have an opinion on this perspective?
Whether content is “brand safe” depends very much on the brand in question. For example, political TV shows like “House of Cards” or “Veep” should want their ads to appear next to breaking news about American politics. But other brands won’t. It’s up to each brand to decide for themselves whether news is an appropriate context for their advertising. Perhaps they will tolerate brand risk in exchange for the desired engagement; perhaps not. There’s no one-size-fits-all answer here.
6.) Have you seen more publishers offer new or customized products that directly counter brand safety-related concerns, and if so, what type of functionality is there?
In most cases, publishers aren’t building products because this is cost prohibitive and is not scalable. Instead, they are leveraging tools like IAS’s Publisher Optimization solution to help them to automate the process of preventing brand safety violations proactively.