Xumo, the joint venture between Comcast and Charter, today announced expanded integrations with Gracenote and IRIS.TV, adding two distinct layers of contextual intelligence across its free ad-supported streaming inventory - a move that extends capabilities historically reserved for on-demand libraries into the more complex territory of live linear FAST channels.

The announcement, made on June 22, 2026, in New York, positions Xumo among the first FAST operators to bring both program-level metadata and video-level frame analysis to linear streaming at scale. The integrations cover Xumo's global footprint of 2,000 channels distributed across more than 30 platforms, spanning both linear and on-demand programming.

What the integrations actually do

The two partnerships address different technical layers of the contextual targeting stack and are designed to work in combination rather than as alternatives.

Gracenote CTV Content Intelligence supplies standardized, source-validated program-level metadata, content IDs, and taxonomy to Xumo's streaming inventory. According to Xumo, the system creates a common language between buyers and sellers, allowing marketers to evaluate FAST opportunities and activate campaigns with greater transparency and precision. Gracenote, the content intelligence business unit of Nielsen, covers more than 50 million titles across 80,000 channels and catalogs in over 70 languages and 80 countries. Its taxonomy infrastructure has been the subject of significant development over the past two years as contextual advertising has moved toward the center of programmatic CTV strategy.

IRIS.TV, now a Viant Technology company following its acquisition in November 2024, brings a different capability: video-level contextual intelligence delivered at the individual content level. According to Xumo, leading AI models analyze content frame-by-frame, surfacing signals across both linear and on-demand programming through IRIS.TV's technology. The core identifier behind this system is the IRIS_ID, a universal content identifier that IRIS.TV uses to tag video at the asset level so that data partners can analyze it and return contextual classifications into the bidstream.

The distinction between the two layers is worth spelling out. Gracenote operates primarily at the program level - genre, rating, content type, advisory categories, and standardized IDs that allow buyers to recognize and transact on the same show across multiple distribution points. IRIS.TV operates below the program level, at the video or scene level, using AI to classify what is actually on screen at a given moment. Together, according to Xumo, these capabilities enable advertisers to improve relevance, reduce waste, and support stronger performance outcomes.

Why FAST has been a harder problem

FAST channels present a structural challenge that on-demand streaming does not. With on-demand content, every individual title is known in advance and can be classified, tagged, and made available to buyers before any impression is ever served. Linear FAST channels air content continuously - sometimes live sports, sometimes older library content, sometimes programming whose metadata has never been properly structured.

The gap in FAST channel metadata has been documented in detail by Gracenote's own research. According to a Gracenote survey of 600 U.S. brand and agency executives published on October 1, 2025, nearly 70% of respondents identified lack of standardization and unification in content metadata as at least a modest challenge when developing CTV campaigns. Among sports programs airing between mid-July and mid-August 2025, 55.4% lacked original air date information before metadata enrichment - a figure that directly affects whether a buyer can distinguish a live game from a replay, which has direct implications for pricing and placement decisions.

Earlier Gracenote data published by PPC Land in August 2025 showed that in January 2025, 66 active FAST channels operated without any genre information at all. Across all FAST channels, only 8% contained secondary genre classifications, compared to more than 50% of CTV content having two or more genres. That absence, as the same article noted, directly limits the biddable surface area for contextual campaigns: when programmatic bid requests contain Gracenote content IDs, advertising opportunities increase by 712%, according to a Q4 2024 analysis of 6.8 billion publisher-supplied requests.

That 712% figure is the commercial case for what Xumo is deploying today. Without structured metadata, large volumes of FAST inventory sit outside the reach of contextual buying systems regardless of how sophisticated the DSP or how refined the audience model.

The purchase intent claim

According to the announcement, relevant ads in contextually appropriate environments can drive 5.2 times higher brand purchase intent. That figure appears in the press release without an attributed source or a published study reference, so it should be treated as a promotional data point rather than independently verified research. The broader directional argument - that contextual alignment improves campaign performance - is, however, consistent with findings from multiple industry studies.

IRIS.TV and Upwave published research in April 2025 showing campaigns using IRIS-enabled data demonstrated a 2x lift in awareness, a 3x lift in ad recall, and a 5x lift in favorability compared to standard CTV benchmarks. A separate study covered by PPC Land in June 2026 documented that CTV ad spending is projected by eMarketer to reach $38 billion in 2026, representing 43% of all TV ad budgets - a scale that gives contextual infrastructure investments real commercial weight.

Viant's role and the IRIS.TV ownership structure

One detail worth noting in this integration is the corporate structure surrounding IRIS.TV. Viant Technology acquired IRIS.TV in November 2024, and the company now operates as a Viant subsidiary. Viant closed its $40 million TVision acquisition in May 2026, adding second-by-second attention measurement to a platform already built around the IRIS_ID contextual classification system and Household ID identity infrastructure.

Richie Hyden, SVP of Publisher Solutions at Viant Technology, is quoted in the Xumo announcement. "Together, with Xumo, we at Viant and IRIS.TV are giving brands the ability to activate against the precise content moments that drive attention, from live sports to premium streaming, and tie that directly to outcomes," Hyden said.

The Viant involvement here matters structurally. When an advertiser buys IRIS-enabled inventory on Xumo through Viant's DSP, the system can in principle connect content-level targeting signals with Viant's Household ID coverage and, following the TVision integration, with attention measurement. That creates a chain from contextual adjacency through to outcome measurement - which is what most CTV buyers say they want but have historically struggled to assemble across multiple vendor relationships.

PPC Land's coverage of Viant's SupplyIQ launch on June 11, 2026 noted that 85% of CTV spend on the Viant platform currently flows through Direct Access, the company's direct publisher integration program. Xumo is among the pilot CTV partners listed in Viant's publisher ecosystem, making today's IRIS.TV integration consistent with that broader supply relationship.

Gracenote's expanding CTV infrastructure

The Gracenote side of today's announcement reflects a multi-year infrastructure build that has accelerated throughout 2025 and into 2026. Gracenote first introduced contextual CTV ad targeting in July 2024, partnering with Peer39 as the DSP-side distribution partner and Magnite as the first SSP to support the new categories. Xumo was among the initial FAST publishers in that 2024 announcement.

The trajectory since July 2024 has been systematic. Index Exchange became the first SSP to embed Gracenote contextual intelligence directly into its platform in September 2025, adding brand safety segments and Do-Not-Air controls to programmatic workflows. Gracenote then launched Content Connect in December 2025, providing agencies, brands, SSPs, and DSPs direct access to standardized program-level metadata via its proprietary content ID graph - a system built on unique identifiers known as TMS IDs that enable consistent show tracking across multiple distribution platforms.

The TMS ID infrastructure is the connective tissue running through Gracenote's commercial proposition. When a buyer specifies a content exclusion or targets a genre category, the TMS ID ensures that specification applies consistently regardless of which streaming platform or distribution service carries the content. That interoperability problem - the same show appearing in multiple places under different metadata schemes - is what makes standardized IDs commercially valuable.

Gracenote research published in May 2026 found that 86% of U.S. media planners said they would shift linear TV budgets to CTV if show-level data were available to them. That finding quantifies the commercial opportunity that both today's Xumo integration and the broader Gracenote infrastructure build are oriented around capturing.

Jake Richardson, Vice President of Partnerships at Gracenote, addressed the FAST-specific application directly. "As FAST platforms continue to grow into go-to destinations for viewers, precise ad placement within premium content has never been more critical," Richardson said. "Through our partnership with Xumo, buyers gain clearer insight into content environments to maximize the impact of their FAST investments with greater confidence, contextual alignment and brand suitability."

Xumo's operational scale and recent performance

Xumo entered 2026 with reported momentum across its key operating metrics. According to Xumo, Xumo Play recorded approximately 40% year-over-year monthly active user growth, a 64% increase in total hours watched, and a 22% rise in time spent per user during what the company describes as a record year.

Those figures provide context for why today's integrations are commercially significant beyond their technical merits. A FAST platform with flat audience growth and thin engagement would have limited appeal as a contextual targeting destination regardless of how sophisticated its metadata infrastructure becomes. The combination of audience scale, engagement depth, and now contextual signal availability changes the buyer's equation.

Jerrold Son, Vice President of Ad Revenue Operations at Xumo Advertising, framed the integration in terms of the live streaming environment specifically. "By bringing deeper contextual signals across all our FAST programming, including linear, we're helping advertisers improve brand safety and suitability, better navigate content genres, and bring more actionable context into programmatic buying," Son said.

The emphasis on "including linear" is deliberate. Linear FAST - channels that simulate traditional broadcast schedules rather than serving content on demand - has historically been the least transparent environment for programmatic buyers. A news channel airing live coverage, a sports channel running a game replay, a movie channel with a late-night slot: without program-level metadata tied to a live schedule, buyers cannot evaluate those placements at the granularity that contextual advertising requires.

Context in the broader CTV advertising environment

The Xumo integrations arrive during a period of significant consolidation in CTV contextual infrastructure. The broader market context is one where the tools for contextual buying are becoming more available, but where the adoption of those tools has lagged behind their availability.

A study Gracenote released in June 2026 tested ungrounded large language models against 2,600 streaming titles across 13 countries and found the models invented all metadata entirely for nearly one in five titles. In the United States, cast accuracy from ungrounded models ran at 53%. That study, which carries obvious competitive motivation given Gracenote's commercial interest in selling verified data as the alternative, nonetheless highlights the practical risk of contextual systems built on AI-generated rather than human-verified metadata: if the classification is wrong, the adjacency decision built on it is wrong.

PPC Land's coverage of ad tech trends in June 2026 noted the broader pattern: CTV contextual advertising depends on content metadata to place ads, and any platform generating those classifications through ungrounded AI is operating with significant accuracy risk at the content level. Verified, source-validated metadata - which is Gracenote's explicit product positioning - exists partly in response to that risk.

For programmatic buyers evaluating FAST inventory, the practical implication is straightforward. As PPC Land documented in its coverage of Viant's Wurl integration in August 2025, scene-level contextual intelligence at that level spans 95 billion monthly available CTV impressions. The combination of IRIS.TV's frame-level analysis and Gracenote's program-level taxonomy, deployed across Xumo's 2,000-channel FAST footprint, gives buyers a surface where multiple layers of contextual assurance can be stacked - not just what the show is, but what is actually on screen at the moment an ad is placed.

The FAST channel universe has grown substantially, from 424 active channels in 2020 to 1,700 in 2026 according to Luminate Intelligence, with active U.S. FAST households expanding from 22 million to an estimated 54 million over the same period. Grand View Research estimates the global FAST market reached $9.4 billion in 2024, with projected compound annual growth of 23% through 2030. As that inventory expands, the quality differentiation between contextually classified and unclassified supply is expected to increase rather than decrease - which is the commercial logic underpinning Xumo's decision to add two contextual partners simultaneously rather than one.

What this means for media buyers

For buyers running programmatic campaigns against FAST inventory, the practical effect of today's announcement depends on how the signals are exposed in the bidstream and which demand-side platforms support them. Xumo has not detailed the specific DSP integrations through which Gracenote taxonomy and IRIS.TV classifications will be accessible, which is a meaningful gap for planning purposes. The signals exist; their accessibility for any given campaign depends on the buying path.

The brand safety dimension is the most immediately applicable. FAST channels have historically posed brand safety challenges because content scheduling and metadata coverage have been inconsistent. A buyer blocking certain content categories across CTV inventory could not always rely on those exclusions applying accurately to linear FAST channels in the absence of structured program data. Gracenote's taxonomy, applied at the program level with TMS IDs, addresses that problem directly. IRIS.TV's frame-level analysis provides a secondary check at the content level.

For contextual targeting beyond brand safety - reaching viewers of live sports, navigating genre preferences across linear channels, or building daypart strategies around known content types - the value depends on the precision and coverage of the metadata actually passing through bid requests. As documented in prior Gracenote research, only 32% of bid requests included genre content attributes before metadata enrichment programs of this kind, meaning the majority of FAST inventory was invisible to contextual buyers regardless of targeting intent.

Today's integration is a structural supply-side move. It does not automatically change what every buyer can do immediately, but it creates the metadata layer on which more sophisticated contextual campaigns can be built.

Timeline

Summary

Who: Xumo, the streaming platform joint venture between Comcast and Charter, together with Gracenote (the content intelligence business unit of Nielsen) and IRIS.TV (a Viant Technology company). Key executives quoted include Jerrold Son, VP of Ad Revenue Operations at Xumo Advertising; Jake Richardson, VP of Partnerships at Gracenote; and Richie Hyden, SVP of Publisher Solutions at Viant Technology.

What: Xumo today announced expanded integrations with two contextual intelligence providers. Gracenote CTV Content Intelligence delivers standardized, source-validated program-level metadata, content IDs, and taxonomy to Xumo's FAST inventory. IRIS.TV brings video-level, frame-by-frame AI content analysis across both linear and on-demand programming. Combined, the integrations cover Xumo's global footprint of 2,000 FAST channels on more than 30 platforms. Xumo reported approximately 40% year-over-year monthly active user growth and a 64% increase in total hours watched in the prior year.

When: Announced on June 22, 2026, in New York.

Where: Xumo's FAST inventory across more than 30 streaming platforms globally, spanning both linear FAST channels and on-demand content libraries.

Why: FAST inventory has historically lacked the content metadata infrastructure that contextual programmatic buyers require. Without structured program-level data, large volumes of FAST inventory are invisible to contextual targeting systems, limiting advertiser spending. Xumo's integration of Gracenote's standardized taxonomy and IRIS.TV's frame-level AI classification is designed to bridge that gap, making FAST linear channels accessible to the same contextual buying tools that have long served on-demand streaming libraries.