How CTV Marketers Are Using Gen AI to Improve Contextual Advertising

Generative artificial intelligence has already unlocked a host of new capabilities for marketers, particularly regarding ad creative. Still, the technology could soon transform the world of contextual advertising as well.

Using gen AI, marketers can now scan the content of a series and gather more granular metadata, such as program details and closed captioning, to gain a clearer picture of the emotional tenor of a scene. 

At the Possible conference in Miami, executives from CTV, marketing and measurement firms shared the results of their early efforts with the technology.

According to Peter Crofut, the vice president of business development at Wurl, marketers can use this metadata to serve an ad whose tone reflects the mood of the content the viewer has just watched, heightening its contextual relevance and improving the performance of the creative.

In one case study with a financial advertiser, the technology drove 200% more brand engagement than the rest of the campaign, according to Laura Grover, the senior vice president and head of client solutions at EDO.

“This is such a key factor in scaling CTV from a programmatic lens,” said Kathy Argyriou, the head of publisher sales at FreeWheel. “We are now able to identify more context within commercial breaks, then figure out where you want to send targeted, contextual ads within a pod.”

Context is everything

Though contextual advertising has always sought to serve ads appropriate to the context, gen AI enables marketers to track the emotional sentiment of a show by scene rather than series. 

For example, an episode of The Office could be a comedy or a romance depending on the scene, Crofut said. Similarly, while Pillsbury might happily serve an ad during Shrek, it would want to avoid the scene where Lord Farquad throttles the Gingerbread Man.

By inferring the mood of a scene and serving a similarly toned ad, the technology ensures a higher level of relevance, a cornerstone of effective advertising. 

Companies are already looking to include the innovation. NBCUniversal’s ad sales chief Mark Marshall told ADWEEK in January that its One Platform Total Audience utilizes AI to account for genre, tone and emotion, adding incremental reach to make media plans more effective.

A generation of AI

As the technology improves, these pairings will only grow more precise, said Matthew Kramer, the head of brand investment at Media.Monks.

The more metadata gen AI technology can gather, the more information it can pass in the bidstream, allowing marketers to even more granularly target ads that match the emotion and subject matter of the content. And 2024 will bring more data.

IAB recently released its digital advertising report ahead of NewFronts presentations, indicating generative AI was one of the trends to watch for in the upcoming year.

But while the technology is sophisticated, its efficacy is rooted in the commonsense nature of contextual advertising. 

“If the last few years have been all about data, we are now seeing a swing back toward the intuitive logic of contextual relevance,” Kramer said. “Being in the middle of those two poles is exactly where ads are most effective.”