By Mattias Spetz, MD, Europe
Last week we discussed the nuance of having a blacklist approach that does not overly exclude content (here). On YouTube, keyword blacklists uploaded into GoogleAds or DV 360 are exact match only – on their own, they will act as a blunt instrument and block content without regard for context.
Channel Factory uses keyword blacklists as one part of a multi-pronged approach to delivering brand suitable, efficient YouTube advertising and, in our experience, it must begin at the pre-bid phase of a campaign where inventory selection and pre-optimization is most critical.
Step One – Brand Suitability Parameters
It’s essential for brands to craft an approach that works for them. Brand suitability is all about selecting highly specific content adjacencies and contextual alignments that meet potential customers and brand-loyalists in the right places within YouTube’s massive video ecosystem. Certain contexts work better for certain brands, and that differs by market, by language, and so on. The only way to finetune that resonant environment-building is by sitting down and deciding what contexts work best align with a brand’s values.
Step Two – Ongoing Analysis of Inventory
Channel Factory’s proprietary technology is always-on, always curating content from within YouTube’s constant user-generated content machine. Campaign success occurs as much at the pre-bid phase as it does mid-flight.
Keyword blacklists (approximately 155,000 words in total) are updated regularly in 36 languages to account for real-time, real-world events. They cycle constantly within our technology and data platform, ViewIQ, flagging and categorizing all undesirable content and identifying the highest quality inventory for use in client campaigns.
These blacklists look at all the content signals available within videos and channels, which includes a whole host of metadata (titles, tags, descriptions), as well as the audio tracks of the videos themselves. Unlike exact match keyword blacklists used in live campaigns, these advanced blacklists are capable of detecting context from natural language cues.
Brand unsuitable content is tagged at the keyword, channel, topic and video-level and used to guide whitelist curation, as well as generate video-level and keyword blacklists calibrated to suit a particular brand and run dynamically within campaigns for ultimate suitability and scale.
Step Three – Whitelists Drive Brand Suitable, Contextual Performance
Whitelists help brands curate brand suitable, trusted environments to advertise within. However, smart use of blacklists is a pivotal part of that strategy.
Whitelists are built from the constant cycling of inventory through brand suitability filters. Once these highly refined, pre-bid optimized channel and video lists are built out, however, that’s not the end of the story.
Whitelists need to be updated continually throughout a campaign. Otherwise what we see is a steady increase of CPV/CPM as brands keep bidding against the same static whitelist inventory. It also affects scale as well as relevance – by bidding against static content, brands run the risk of missing out on new, viral highly relevant and engaging content.
Fortunately, blacklists are always running in the background, distilling the best content and extracting the brand unsuitable to ensure there’s always new inventory available for the multiple whitelist iterations it takes to drive dynamic contextual performance on YouTube.
Step Four – Video-Level Blacklists
Once brands have guaranteed pre-bid optimized whitelists which target suitable, aligned content and in-market audiences their campaigns run alongside highly specific video-level and keyword blacklists.
Video-level blacklists enable brands to buy against channels that make overall sense for them, but which have the odd video that sits outside their brand suitability parameters. Without using this technique, brands otherwise have to write off entire channels worth of scale just to avoid one lone wolf video.
Keyword blacklists at this stage in the brand suitability provide “air cover” for a “ground force” of highly curated whitelist inventory. In-flight keyword blacklists are highly brand-customized and, thanks to pre-bid curation, they’re not the only hope of campaign brand suitability once a campaign is live.
As far as COVID-19 content adjacencies are concerned, our clients have taken varying approaches to brand suitability for coronavirus related content. Some of our clients are fine with appearing next to this type of content – in fact some advertisers actively seek it out. However, for some brands it is the right approach to avoid it. A lot of these decisions are driven by an understanding of their consumer, as well as the brand’s creative messaging at this time. For example, brands helping to amplify the WHO’s messaging by providing educational content might find news adjacencies ideally suited to them and their customer’s content consumption choices.
Early on in this crisis we shared a Coronavirus keyword list with all our clients and other advertisers to help them target or de-target content related to COVID-19 as they see fit.
Clumsy blacklisting is a real concern, but that’s only when it’s the only method by which advertisers are driving brand suitability in their campaigns. A lot rests on contextual analysis capabilities and ongoing strategic media planning considerations.
At Channel Factory, we’ve refined a whitelisting methodology which introduces the very best in contextual sophistication. This approach is key to aligning brands, publishers/content creators and consumer stakeholders especially during times where those value systems are in constant flux.