Does IRIS.TV’s Tagging Partnership With Tubi Signal a Moneyball Moment for Longtail CTV Content?
IRIS.TV, which was acquired by Viant a year ago, is a leader in tagging shows by mood and theme—the company provides the examples of “joyful” or “travel”—in order to maximize opportunities for CTV advertising. IRIS.TV asserts that its approach facilitates more precise targeting that can be used with massive streaming content libraries.
IRIS.TV recently partnered with Tubi so IRIS.TV could provide advertisers with the appropriate level of AVOD addressability, contextual targeting, and measurement. Viant’s press release states, “With a new, powerful ID sync between Viant and Tubi, who offers the world’s largest Hollywood collection, thousands of creator-led stories and hundreds of Tubi Originals, advertisers can reach addressable audiences—spanning more than 100 million monthly active users and over 300,000 movies and TV episodes—with greater accuracy and scale.”
To learn more about IRIS.TV’s tagging strategy and collaboration with Tubi, I spoke with Rohan Castelino, CMO of IRIS.TV.
Rohan Castelino, CMO, IRIS.TV
Brandi Scardilli: What are the biggest misconceptions advertisers still have about long-tail content on FAST and ad-supported platforms? How is IRIS.TV leveraging AI and data-driven tagging to address those misconceptions?
Rohan Castelino: The biggest misconception is that long-tail content isn’t premium. In streaming, viewers have unlimited choice, and content is fragmented across hundreds of apps and services. Platforms like Tubi already have some of the largest premium content libraries in streaming, drawing millions of daily viewers. Unlike traditional broadcast TV, FAST offers personalized viewing at massive scale, not mass appointment viewing.
FAST and AVOD services often share content libraries with SVOD platforms, but also go deeper with a broader range of premium titles. Viewers who stream The Pitt on HBO Max might also watch reruns of ER on The Roku Channel. As Scott Sutton, director of media for CKE Restaurants, put it at the ANA Masters of Marketing Conference, “Premium content is the content that converts, and IRIS.TV helps us find that content.”
We achieve this by integrating directly with publishers and connecting their video libraries to leading AI models that analyze each show and movie frame by frame. These models generate standardized content categories that can be activated across the ecosystem. Scott knows his customers are into “anime,” “sports,” and “video games.” With IRIS.TV, he can place ads in shows aligned with those interests and see which categories drive conversions in real time.
When content is accurately and consistently categorized across all publishers, it unlocks the scale to capture more ad spend. For FAST platforms, that means expanding the total addressable market and ensuring every piece of high-quality content can deliver value.
Tell me about how IRIS.TV’s contextual data pipeline works. Is it AI-driven? How do you ensure content gets all the tags it needs to increase discoverability?
IRIS.TV connects three primary customer types: publishers, data providers, and ad platforms. For publishers, we ingest and normalize their metadata via their CMS, then assign a unique IRIS_ID to each asset. To date, over 70 million videos have IRIS_IDs, representing over 30% of all CTV ad requests.
We share each asset’s metadata with our content data marketplace partners, who apply proprietary AI to classify it by context, emotion, and brand suitability. Their categories are then linked to the IRIS_ID.
Publishers include this IRIS_ID in their ad requests, enabling any SSP or DSP integrated with IRIS.TV to unlock targeting and measurement use cases.
What makes this approach powerful is its open and scalable design. It does not rely on one taxonomy or one model. Multiple AI providers can analyze the same content, encouraging innovation and allowing advertisers to choose the models best suited to their campaign goals. Some models can identify nuanced categories such as “interior decorating” or “parenting babies and toddlers,” which map directly to consumer behavior and spending.
For brand safety, these models also classify content according to the GARM framework, helping advertisers confidently target categories like “news” while avoiding sensitive topics such as terrorism or violence. This protects brands and supports quality journalism.
How has Viant’s acquisition of IRIS.TV helped you be more effective?
While IRIS.TV remains an independent subsidiary, and the IRIS_ID can be activated across other platforms, Viant’s acquisition has accelerated our roadmap and expanded our product capabilities. Together, we have launched several first-to-market innovations within Viant DSP.
In June, we released the IRIS-enabled Content Report and Pre-Bid Targeting. These tools go beyond contextual targeting, offering category-level performance insights that help advertisers understand what content drives engagement, even on campaigns not originally using contextual data. This creates an onramp to contextual activation and deeper insights into viewer preferences.
We have also launched exclusive inventory and data partnerships for Viant DSP clients, including scene-level contextual and emotional targeting for live sports and online video. This helps brands connect with viewers at their most engaged moments.
How does the shift to content-based targeting change how ad buyers think about audience versus context in CTV?
Content-based targeting is about efficiency and effectiveness. It complements audience-based strategies rather than competing with them. Even with the best identity graph, CTV remains probabilistic because a smart TV is a shared household device. To reach the right viewer, advertisers need to understand what is on screen in real time.
Content adjacency increases the likelihood that the intended viewer is the one watching. For example, an outdoor apparel brand we worked with targeted both men and women using a mix of audience and contextual categories. For women, “self-care” content performed best; for men, content associated with the emotion “reliability” drove stronger engagement. Combined with geo-targeting near retail locations, this approach expanded reach and improved relevance.
The Viant-Tubi press release states, “These capabilities enable advertisers to engage audiences at the most relevant moments in premium programming, driving stronger outcomes and increased demand for Tubi’s supply.” What does this look like in practice?
Ads remain in-stream and appear during standard ad breaks, but relevance begins with mindset. When viewers sit down to stream, they are intentionally engaging with content. Whether watching NBA highlights or Hoosiers, the emotional and thematic cues around “basketball” are active.
While IRIS.TV enables scene-adjacent ad delivery, the true moment of relevance extends across the viewing session. CTV viewers are among the most engaged audiences, with research showing they watch four times more ads when those ads align contextually with the content on screen.
IRIS.TV calls its data-driven approach a “Moneyball strategy.” Why is that an apt metaphor for today’s CTV advertising?
At its core, Moneyball is about using data to challenge assumptions and find value others overlook. That is exactly what we are doing with content data in CTV. Instead of relying on traditional notions of what is “premium,” advertisers can now use objective, real-time insights to identify the types of content that actually drive performance. Often, the categories that deliver the strongest results are not the ones marketers would have picked at the start of a campaign.
This approach turns CTV fragmentation into a strength. Every show, movie, and clip becomes an addressable opportunity once it is properly understood and categorized. It allows brands to invest more intelligently, expand reach, and improve ROI.
Most importantly, this is not about replacing creativity with algorithms. It is about using AI and data to make creative content more discoverable and monetizable. The Moneyball moment in CTV is here, and the marketers who embrace data-driven content strategies are the ones setting themselves up to win the next decade of streaming.
Related Articles
Tubi, the free streaming service owned by Fox, flew under the radar for years, but now the industry has taken note of its surging popularity. At NAB New York, Peter Kafka, Business Insider's chief correspondent about the TV/streaming landscape, talked with Tubi CEO Anjali Sud about the TV/streaming landscape and how her company is changing the ways that people watch.
10 Oct 2024
In the era of AVOD, Jeffrey Johnson, Senior Director, Supply and Demand at Verve Group, breaks down which types of metrics matter most, categorizing them by engagement, visibility, and performance.
17 Jun 2024
Tubi, Fox Corporation's ad-supported streaming service, released findings from The Stream 2024: Streaming Insights for Marketers. In this year's report, Tubi, who partnered with The Harris Poll to conduct research, provided a deep dive into the behaviors and preferences of today's streamers to help inform marketing strategies. Streaming Media's Tyler Nesler spoke with Tubi's Cynthia Clevenger, Senior Vice President of B2B Marketing, about many of the report's key findings and what they reveal
04 Apr 2024