While streaming accounted for nearly 40 percent of total TV usage in July, advertisers are seeking more content transparency in order to shift additional video ad dollars into streaming. Over the past several years, publishers have invested in building authenticated first-party data that assures marketers they are reaching real, individual users within their target audiences. However, as connected TV (CTV) grows, one major challenge persists: sharing content information, including standardized contextual and brand-suitability data that can be activated programmatically.
Linear TV is delivered to millions of viewers who typically see the same ads simultaneously. Targeting is based on the network and viewership demographics, while brand suitability is determined by program restrictions at the show level. In CTV, millions of viewers stream their favorite shows, but viewers see different ads because of advanced targeting. When it comes to brand suitability, legal and technological reasons prevent CTV buyers from enabling show-level restrictions programmatically.
In addition, unlike a mobile device which is personalized to the viewer, a smart TV is shared in a household of several viewers. Combined with questionable accuracy of ad targeting and measurement data, at best buyers have an unclear idea of who saw the ad and within what type of content the ad appeared.
The emergence of AI has been a focal point of industry disruption, most notably in the entertainment industry. As the studios and gilds settle on the role of organic intelligence AKA people, AI, ironically, may offer the best path for buyers to scale investment in ad-supported streaming.
Publisher-declared metadata can’t walk the last mile alone
CTV buyers, whether placing their orders directly or programmatically, have the same needs as traditional TV buyers — scale, relevance, and brand suitability. And while they may ask for program-level information like genre, title, and episode, publishers often can’t make that information available on streaming.
To address this transparency issue, publishers are passing more information in the bidstream and building their own taxonomies that allow them to tag content in a way that provides more nuance and granularity. While laudable, this solution only partially addresses the problem of transparency and only within the confines of a publisher’s own walls. One publisher may tag golf content under “sports,” or “lifestyle.” Extrapolating, a comedy could be tagged “funny,” “situation comedy,” “sitcom,” “family fun,” and so on.
In a recent analysis, IRIS.TV found there were over 740 different ways that “comedy” was labeled in only 100,000 ad impressions. Jounce Media has published similar research in “news” content, a genre that is over-blocked by advertisers, explaining that there is not a single demand-side platform that offers genre targeting or genre reporting in CTV because “the data is simply too messy.” This fragmented data makes it nearly impossible to scale programmatic campaigns across publishers. Still, even when following a standard like the IAB taxonomy, consistency across supply sources is elusive, as the information behind sight, sound, and motion is greater than the sum of its parts.
AI knows what the story’s about
Imagine now that publishers can move beyond sharing only broad categories like genre but also video-level contextual segments enriched by AI. These are insights that marketers have never before had access to, and that allow them to move past legacy segments like “show” and “episode” and target in an entirely different way. AI can scan video frame by frame and analyze it for content, emotion, objects, and people, then assign it with contextual and Global Alliance for Responsible Media (GARM) brand-safety and brand-suitability segments. If these segments were made available to advertisers, it would ensure that every impression is sold at its premium for two reasons:
- Brand suitability at scale. Highly watched content that would otherwise be blocklisted (like news) or mature content (like crime dramas) have precise GARM brand-suitability segments that differentiate between low- and high-risk content.
- Relevance resonates. With the exception of live events, streaming viewers watch when they feel like it. Ads served programmatically through CTV could be delivered in a million different permutations — at night, in the morning, or watching a comedy, drama, or home cooking show. Studies have shown that when streaming ads are more aligned with the subject matter or mood of the show, ads are more effective, as viewers pay more attention, have double the unaided ad recall, and have higher search and purchase intent than when demo data or generic contextual data was used for ad targeting.
AI unlocks budgets with data that delivers relevance, resonance, and scale.
For over a decade, viewers have become accustomed to binging every season of their favorite show ad-free. While the leading subscription services are introducing ad-supported tiers, viewers are sensitive to disruptive, irrelevant, and unsuitable ad experiences. And with free ad-supported television (FAST) owning an increasing share of the available inventory, the supply chain becomes more complex as smart TVs, connected devices, apps, and other ad tech enter the mix.
Relevance matters and the leading contextual intelligence and brand-suitability solutions have traditionally applied their AI to display and social video advertising by analyzing content via its page URL. In streaming, there is no analogous publicly accessible signal. To address this challenge, the IAB Tech Lab released a community extension for extended content identifiers (ECID) to act as a “clean room” to hash a publisher’s first- and third-party enriched content metadata. In this use case, the content ID serves as a common signal across supply sources and contains video-level contextual and brand-suitability metadata provided by the currencies that marketers rely on for planning, targeting, and measurement.
Ad spend in CTV is expected to exceed $41 billion within five years and agencies like GroupM and Dentsu believe that AI is the path forward to enhancing the way they invest in streaming. By integrating these capabilities into their media strategies, they are elevating the art and science of advertising and leading the way to better viewing experiences that drive better advertising outcomes.
The views and opinions contained in this op-ed represent the views and opinions of the author(s) and do not represent the views or opinions of The Current, a division of The Trade Desk, nor of The Trade Desk. The appearance of the op-ed on The Current does not constitute an endorsement by The Current or The Trade Desk of the content.