Everywhere you look, our industry is talking about AI: AI-powered planning, AI-driven optimization, agentic media buying, automated curation.
The promise of greater efficiency, better performance and smarter decision-making is compelling. But amid all the excitement, one important question isn’t getting enough attention: How good are the data sources that are feeding this automation?
No matter how sophisticated the technology becomes, its effectiveness will always depend on the quality of data that underpins it. If those signals are incomplete, inconsistent or overly broad, automation can only take us so far.
Streaming TV has outgrown broad contextual signals
CTV has evolved dramatically over the past decade. Today’s viewers expect personalized experiences, advertisers want measurable outcomes and publishers are producing more premium content than ever before. Yet much of the advertising infrastructure supporting CTV still relies on only the most basic contextual signals that were designed for a much simpler ecosystem. They don’t accurately describe what viewers are actually experiencing on-screen before an ad appears.
Consider a drama series. One moment may feature an intense argument, the next is comedic relief. Minutes later, the story shifts into an emotional family conversation. From an advertising perspective, those moments create very different environments. Still, they’re often treated the same because the program carries a single genre label.
Context doesn’t work that way. The moments surrounding an ad break shape how viewers receive advertising. They influence attention, mindset and campaign performance. That’s why advertisers have long sought more contextual transparency to understand what’s happening when their message appears.
Understanding the moment
Rather than relying on static metadata, AI can now analyze content the way viewers experience it through imagery, dialogue, audio, emotion and deeper contextual cues. The result is a far richer understanding of the moments leading up to an ad opportunity, giving advertisers the ability to align with the viewer’s mindset versus relying on broad assumptions.
We’ve already seen how contextual signals can help advertisers move beyond blanket approaches to brand safety, allowing them to evaluate individual moments instead of excluding entire genres or publishers.
News is a perfect example. While news is often treated as a single high-risk category, recent research found that more than 35% of scenes in news content are fully brand safe, highlighting the significant opportunity advertisers may be leaving on the table when they rely on broad category exclusions rather than evaluating content moment by moment.
The same principle applies across campaign planning. Better contextual understanding helps advertisers make more intentional decisions while allowing publishers to unlock greater value from premium inventory.
Better signals only matter if buyers can use them
The industry’s next challenge is making signals consistently available throughout the programmatic ecosystem.
Today, content data varies from one publisher to the next. Metadata standardization is lacking. Signal availability is inconsistent. Buyers frequently need separate integrations to activate different contextual and brand safety solutions. As a result, advertisers aren’t always able to access the richest signals across the full breadth of streaming inventory.
But as AI becomes more central to media buying, that fragmentation becomes increasingly difficult to ignore.
The advertising industry has an opportunity to make identity, data, intelligence and decisioning capabilities work better together on CTV.
Allowing specialized technologies to bring together the right signals, intelligence and inventory is a must. The next phase of CTV will be defined by how easily those signals can be activated.
The value is in creating the infrastructure that allows the best capabilities across the ecosystem to work together. That means reducing friction, improving interoperability between buyers and sellers and making high-quality contextual intelligence available wherever advertisers buy streaming inventory. Contextual targeting shouldn’t depend on which publisher, platform or technology provider happens to sit in the middle of a transaction.
Success in the next era of programmatic advertising will depend on the quality of the signals available to buyers and how easily those signals can be activated at scale. The individual pieces are largely there. The next phase is about filling the gaps so that all of this intelligence can translate into actionable signals, more prescriptive targeting and better outcomes.
