AI and the Future of Streaming Media Advertising
A few months ago, Tom Hanks warned that his likeness had been used in an AI-generated advertisement without his permission, yet another powerful reminder of what AI can now do. This was clearly a misuse of the emerging technology. However, AI can also serve a more positive role for both streamers and brands, providing powerful new ways to engage audiences and improve advertising performance.
As streaming media companies battle for profitability, AI and machine learning provide them with an opportunity to monetize their content in new and more profitable ways. For agencies and brands that advertise on streaming services, AI allows them to further personalize and customize their ads thanks to more precise targeting data and new forms of creative versioning. All of this should lead to a better experience for the consumer, who will potentially see fewer, more relevant ads.
AI Analyzing Every Scene
How does all this happen? The Tom Hanks example shows how powerful AI has become at interpreting and generating digital video, but it's only one example of what AI can do. Going a step further, AI can parse the content of streaming videos to understand what's happening in each scene, allowing for much more granular ad placement or incremental content at the exact height of engagement.
Currently, brands have limited information about the content they advertise against, as they may know it's a horror movie starring Viola Davis, or a rom com with Jennifer Lopez. Today's AI can go deeper, analyzing the content to show that at minute 15, for example, there's a scene with a family enjoying a beach holiday in the Caribbean — which could be a good moment to display a relevant ad.
This information is stored as metadata, data that describes other data, which streamers and advertisers can then use to optimize which ads are inserted into the content at which time. If the family is enjoying a vacation, it could be a good time to advertise a holiday package. If there's a hurricane blowing, it might be a better time for a home insurance ad. This level of detailed contextual targeting has not been available in the past, and it could greatly boost the ad spend ROI for brands.
AI will also allow streamers and advertisers to adapt content based on viewing patterns they observe. If a streaming company sees that a significant portion of viewers switch away from movies during gory horror scenes, for example, they will know to create content that's less graphic in the future.
If all of this sounds futuristic, look at what's already happening today. In November, YouTube began testing a new feature that will allow creators to auto-generate tunes that mimic the voice of a famous artist, thanks to partnerships it struck with nine artists.
Taking Creative Versioning to New Heights
For advertisers, AI will allow them to do a lot more than just reveal preferences. It can also parse vast libraries of creative content, including images, video, and other IP, and generate derivative content that is completely customized for individuals.
For example, if a brand shoots a car ad on location in the Alps and it performs well, the creative department could use generative AI to swap out the car for a different model, creating new content at minimal cost. Conversely, if viewers are responding well to an ad shot in a different location, the company could swap out the background in the first ad and put the car somewhere else.
AI serves a dual role here, by generating the content itself but also determining, at scale, the optimal content to show each viewer, by analyzing data across potentially millions of viewers. AI can tell brands how to adapt their creative to be most impactful, and then who to show that creative to. Again, we're starting to see the first rumblings of this already: Getty Images recently launched an AI service that allows creatives to generate new visual content based on existing images from its vast library.
For consumers, this use of more precise targeting and content versioning should lead to a more personalized and enjoyable content experience. Advertising is least effective when it feels intrusive, and AI will help brands show more relevant ads and improve the effectiveness of their creative. This in turn may allow them to reduce the volume of ads they need to display, further improving the consumer experience.
The Cloud Plus AI Makes This Cost-Effective
This AI revolution is being made possible in part by cloud data platforms that can collect, store, and analyze large amounts of data cost-effectively. The combination of cloud data platforms and AI makes it much easier to share data up and down the supply chain and to automate processes that are manual today.
The speed with which AI is advancing has astounded many, prompting much hand-wringing and new legislation to protect both consumers and IP. But there is little doubt that AI will transform industries, and so media companies, brands and ad agencies should all be thinking now about how to implement this technology in their businesses.
The views and opinions expressed are solely those of the contributor and do not necessarily reflect the official position of the ANA or imply endorsement from the ANA.
Bill Stratton is the global head of media, entertainment, and advertising at Snowflake.