Right viewer, right time: the benefits of scene-level targeting on CTV

Contextual targeting isn’t a new idea. Horace Schwerin‘s first study into the effectiveness of contextual advertising on television was conducted in 1958. His research, titled ‘Do Today’s Programs Provide the Wrong Commercial Climate?,’ examined whether TV programming had any impact on the ads around it.

It turns out it does!

Appearing in the right context can make or break the effectiveness of your ad. Or as Schwerin put it in the 1950s: “TV is not an advantageous medium for every type of product … [Many TV ads fail because media buyers are] college men … not in rapport with the people they are communicating to.”

And now — 65 years later — we have a wealth of data and research proving him right.

Studies show that contextual advertising is more effective for brand recall, conversion, activation, and performance.

So why isn’t contextual advertising more common on CTV? Well, until now, it was hard to do at scale.

The programmatic advertising landscape is riddled with transparency challenges, stemming from multiple sources: publishers focusing on direct sales, buyers zeroing in on specific audiences, and the ongoing issue of inconsistent metadata. Additionally, targeting parameters are highly variable, often only letting buyers target to the level of apps, channels, or at most, programs. This inconsistency complicates the process of pinpointing effective advertising contexts, making it difficult for advertisers to strategically time their ads for maximum impact.

Why is program-level not enough?

When advertisers focus on broad categories like programs or channels to target their ads, they gain some contextual relevance — but miss the granular details of individual scenes. This can lead to ads appearing out of place, disrupting the viewer experience and diminishing the campaign’s effectiveness. Imagine you are buying an ad during The Lion King. This is an ideal spot for a family-friendly product — but not if it appears directly after Mufasa’s death.

Similarly, imagine targeting advertising during news programs. While news content can be a strategic choice due to its consistent viewer base, the lack of scene-level targeting means advertisers risk associating their brands with potentially sensitive or controversial news items. Many brands may steer away from running ads in news content for fear of alienating viewers instead of engaging them.

Scene level targeting would enable advertisers to avoid these scenarios. Ads could be aligned with specific content that matched the sentiment of their brand. This is important, as research by De Pelsmacker, Geuens, and Anckaert (2002) highlights that ads embedded in congruent contexts are perceived as clearer and more likable, especially among ‘low-involvement’ audiences — viewers who are watching less intently.

The power of scene-level targeting

One of the standout features of scene-level targeting is its ability to match ads with the emotional tone of the surrounding content. Imagine watching a deeply emotional scene and suddenly being interrupted by a jarring, upbeat commercial. This mismatch can create distrust and a negative viewer experience. By aligning ads with the emotional context of the scene, advertisers can enhance viewer receptivity and brand affinity.

The psychology of why aligning your ad to the surrounding content relies on two established psychological principles.

The first is Priming. This happens when exposure to one stimulus influences the response to another, often without conscious awareness. For example, viewers watching an exciting sports event are more likely to react positively to ads for sports gear or energy drinks. Placing ads in these contexts naturally primes the audience, making your message more relatable and impactful and is probably part of the reason why advertisers value sports content.

The second principle is the Halo Effect. This cognitive bias means that our overall impression of something influences how we perceive its individual attributes. In advertising, this means ads that appear in situations where they gain positive attention are viewed as more trustworthy and appealing. This is why celebrity endorsements or sponsorships work so well.

Interestingly, Wurl’s research has also found that emotions are key. Ads that match the emotional tone of the preceding content can significantly enhance engagement. This concept builds on the idea of Emotional Resonance, allowing advertisers to leverage both the principles of Priming and the Halo Effect. The content primes the viewer’s emotions, setting the stage for the Halo Effect to enhance the perceived trustworthiness and appeal of the ad. This helps brands to ‘meet the consumers where they are’.

How? Wurl’s BrandDiscovery allows advertisers to analyze their creatives. GenAI then scores the creative based on 8 primary emotions and matches the creative to the most effective ad placements.

How does GenAI help with scene-level targeting?

Scene-level targeting involves analyzing and targeting the specific scene just before an ad break rather than broader categories like entire programs or channels. With hundreds of thousands of hours of content available on CTV, it was not feasible for humans to analyze content at the scene level. Gen-AI, however, is a game changer – unlocking our understanding of content in real-time.

As an example, Wurl’s proprietary large language model (LLM) plays a pivotal role here. It evaluates and classifies the dialogue, sounds, and imagery of content as it passes for distribution. Each piece of content is automatically assigned a score across emotion, genre, and brand safety. As Wurl identifies the content being streamed, these scores are appended to the content metadata before it is sent to the Demand-Side Platforms (DSPs), where the segments can then be targeted.

By analyzing millions of content assets, models can consistently and accurately identify the contextual and emotional characteristics of each scene, ensuring ads are placed in the most relevant contexts.

Independent research, as well as studies carried out by Wurl through the development of BrandDiscovery, have demonstrated that ads that resonate emotionally with the content they follow can enhance performance by 2-3 times.

Of course, scene-level targeting also extends to targeting for genre and brand safety – so a light-hearted ad is not misplaced in the midst of a grim battle scene, and ads for family-friendly products are only placed in brand-safe environments. By analyzing scenes individually, BrandDiscovery ensures ads are always contextually appropriate, protecting brand integrity and maximizing effectiveness.

Conclusion

Combining the power of Gen-AI with smart creative placements, advertisers can achieve stellar results by targeting the scene level. Since each scene carries a unique mood and message, scene-level targeting that leverages emotions ensures that ads are not just seen but felt, aligning perfectly with the viewer’s emotional state and the content’s context.

The shift to scene-level targeting comes at a critical time. As the advertising industry moves away from reliance on cookies, contextual targeting on CTV is gaining momentum because it allows for precise ad placement — but in a cookieless environment. The fact that the CTV ecosystem never relied on cookies is a big reason why advertisers are now increasingly moving their spend to CTV. However, to continue this growth CTV marketers need greater control over where their ads appear – control that is key to achieving the performance marketing scale, provability, and repeatability they’re accustomed to in digital and mobile environments.

There’s a wealth of performance benefits to shifting your mindset to scene-level — why not check out how Media.Monks improved cost per engagement by 200% by embracing this granular targeting approach.