How GenAI Is Poised To Disrupt $30 Billion CTV Advertising Industry

0

Connected TV (CTV) is one of the fastest-growing sectors in digital advertising, expected to grow 25% this year to more than $30 billion. Advertisers are flocking to the channel for its ability to combine the storytelling power of the big screen with the real-time metrics and optimization capabilities of digital.

Internet streaming is now the dominant method through which Americans consume TV, according to a recent Prosper Insights & Analytics survey. Nearly 47% of US consumers said streaming is their most frequent method of TV viewing, while 33% said cable and just 14% satellite.

And, unsurprisingly, it’s most popular among younger generations, with 92% of Gen-Z and 89% of Millennials watching at least one free streaming service such as YouTube or Peacock.

But CTV advertising is still in its early days. And it is about to get a whole lot smarter and more effective thanks to, you guessed it, GenAI, Ron Gutman, CEO of CTV advertising company Wurl, told me in an interview.

“Machines understand emotions better than humans,” Gutman said. “As a result, advertisers can benefit from technology that leverages large language models (LLMs) to analyze the content audiences are consuming and choose the ad campaigns and creative that are most likely to drive conversions based on the emotions of viewers.”

Like the human brain, the latest large language models are now multimodal, which means that they can take in multiple inputs such as text, audio, image, and video. But unlike the human brain, multimodal LLMs (MLLMs) offer computational and mathematical capacities that are far superior in terms of efficiency and scale.

Google released its MLLM, Gemini, near the end of last year and Apple canceled their $10B electric car quest after 10 years of operation to focus on AI, released an image editing model, and last week introduced a model to enhance Siri. In recent weeks, xAI’s chatbot Grok (supported by Elon Musk) was enabled for X premium subscribers and it was reported that OpenAI and Microsoft are working on plans for an AI supercomputer data center at a $100B investment, 100 times the cost of some of the largest data centers in the world. The highly anticipated open-source MLLM Llama 3.0 by Meta is expected to be released in July.

And while a given human individual may have limits in terms of how much information and emotional intelligence they can access, MLLMs have a virtually unlimited store of knowledge and human experiences to pull from thanks to the Internet.

“The best writers, artists, and creative minds have all figured out how to express their feelings with words, songs, and images — and they have shared their work on the Internet for years. MLLMs have been trained on these emotional assets, and, as a result, can understand and describe human emotion with much more clarity than any individual,” Gutman said.

For example, when OpenAI was training its first model in 2017 to analyze Amazon product reviews and predict the next word for a review, they noticed that the model trained itself to understand the emotion or sentiment of a given review. And the model was able to use that understanding to inform text generation. Since then, MLLMs have been able to respond to emotional language that closely resembles responses humans would give, according to recent studies.

For CTV marketers, this means they can use GenAI and MLLMs to dissect the sounds, imagery, and text that appear in CTV content, understand the specific kinds of emotions a given piece of content generates, and translate all the data into measurable, actionable insights — all with unprecedented efficiency and scale.

By leveraging GenAI and MLLMs for these insights, marketers can quickly craft personalized campaigns that better respond to the prevailing moods and emotions of viewers across the globe, foster authentic connections, and drive enhanced brand engagement.

“For marketers, being able to align advertising efforts with the emotional states of their audiences has shown a substantial increase in measured actions and brand awareness,” Gutman said. Consider sadness, which is represented by low levels of serotonin and is the most prolonged emotion. People who are watching a sad scene on TV usually stay sad through the commercial break. And a campaign like one for adopting rescue dogs, say, would likely resonate with them: These viewers would be more attentive to the creative message and more likely to act upon it.

And AI-powered emotion-based analysis won’t just be a boon for the buy side.

“Precise emotional targeting will also benefit publishers, who will reap the reward of greater media investment that comes hand in hand with better performance and reduced churn during commercial breaks,” Gutman said.

The industry is catching on to the role emotion-based targeting can play in creating a better advertising experience. Disney is looking to take advantage through its “Magic Words” tool for advertisers, which brings contextual advertising together with AI and machine learning to analyze the moods its streaming content’s imagery and language evoke. NBCUniversal, too, recently announced AI-driven ads that link its content to emotions, allowing advertisers to serve ads to people deemed likely to buy their products based on the content themes and storylines they’ve watched. Last month, Gutman’s company, Wurl, announced its own GenAI-based CTV advertising solution called BrandDiscovery, enabling advertisers to align their ads with real-time content. BrandDiscovery provides scene-level contextual targeting to help advertisers align the emotional sentiment of their campaign creatives with content closest to the ad break. All of these examples point to the excitement and momentum surrounding emotion-based contextual targeting in the CTV space.

What’s more, tools like Disney’s and Wurl’s can go a long way in averting ad placements that upset viewers or otherwise pose a potential brand safety risk. An MLLM could tag scenes that depict real-word violence, for example, in such a way that prevents upbeat ads (which could be considered insensitive in this context) from being paired with them.

Advertisers and publishers have long understood that emotions play a key role in consumer decision-making — and that the silver screen is the medium best suited for tapping into them. And while we’ve gotten closer to a deeper scientific understanding of emotions and how they influence consumer sentiment, acting on those insights, let alone having the resources to do so, was another matter.

But in light of the latest advancements in GenAI and MLLMs, advertisers and publishers now have the opportunity to capitalize on the biology, neuroscience, and psychology of emotions and unlock unparalleled precision, efficiency, and scale. Performance for advertisers and higher revenue for publishers will follow.

link

Leave a Reply

Your email address will not be published. Required fields are marked *