Digital advertising is overdue for a change. Video viewership has increased and the audience has gone to streaming. As the audience has shifted, the industry simply adopted old ways of advertising in a new medium, but the old model for commercial interruption is flawed. The audience has proven they are no longer willing to endure commercial interruption.
Depending on the source you choose to believe, consumers skip anywhere from 56-90% of the ads they are exposed to. According to some of those same sources, if they don’t skip them, they ignore them. At the same time, paid streaming services continue to grow with as many as 89% of households now engaged in streaming video. Coincidentally, those paid streaming services are predominantly ad-free, with only recent announcements from companies like Netflix and Disney rolling out ad-supported versions (as of writing this, only 5M Netflix subscribers had signed up for the lower-cost ad-supported version vs. 232.5M paid ad-free subscribers). Across the board the message is clear; consumers do not want to see commercial interruption.
The key takeaway here is that while brands want to capture the attention of the consumer, they can no longer rely on the old ways of doing so because there are too many options to avoid them. In today’s modern digital landscape, brands stand out by either creating the content, or becoming part of the content. That is the foundation for Product Placement and that is the goal of bringing into the category..
Product Placement generates approximately $23 billion in annual revenue, primarily focused on large studios as well as on the fast growing influencer economy. Digital Video advertising generates around $74billion in annual revenue for content creators of all shapes and sizes. The overlap of these two categories is where AI product placement operates. AI can be a tool that democratizes access to product placement by merging together the best of native advertising, traditional product placement and programmatic platforms with the advent of new generative AI technologies. The result is a new ad format where advertisers can become a part of the story, rather than an interruption. We see the opportunity for a marketplace business that matches creators with brands, identifies placement opportunities, and inserts photo-realistic product placement as well as a platform. The platform enables creators and studios to monetize their content seamlessly. Product placement has existed for years and while there are other companies historically in the category, they have relied on manual visual effects artists to create and insert the placements. This doesn’t scale.Scale comes from AI and automation being used to generate and insert the placements. It is all done with AI. These placements can be simple or they can be animated or with brand call-outs to ensure attention and activate or engage the audience. To make this happen, there is a form of Generative AI we call physics informed fusion. This Generative Fusion process combines two assets, one a long-form video and the other a 3-dimensional product rendering, into a seamless, single, high-quality composite output where the inserted product is undifferentiatable from a real item placed on set.
This form of generative AI is new and takes advantage of the recent advancements in technology to overcome some of the challenges currently seen in generative AI. Many of the image-oriented AI tools in the market today are great in producing artistic novelty but are challenged when it comes to producing photo-realistic moving video of high resolution. They even suffer from hallucinations, or the inability to be rooted in what is real. If you give an AI the prompts to create product placement insertions, they will come up with images that are fake and can be spotted quickly, standing out in the scene. They don’t respond to the laws of lighting, or spatial depth and motion. Newer models are being trained to be “physics-informed” so that insertions appear real, and are not a novelty in the video. Product Placement needs to be realistic to be considered authentic, and when the AI enables animations as a way to augment the insertion, it does so with a nod or a wink to the consumer to call attention without being disingenuous. This is a new use case for AI offered in a B2B manner, unlike most of the AI category that is going B2C. We see the vision of offering these tools to creative and creators alike, allowing them to leverage the process of Generative Fusion to come up with new and interesting ways to blend objects in video for the purpose of fueling creativity.
It should also be noted that more than just the realistic approach to avoid hallucinations and applications that are unreal, these models are created to fuse long-form, multiple hour video, with objects and products. They offer a solution so the creator knows exactly what the output will be, rather than some “black box” approach that requires trial and error to get it right. This advanced knowledge model for Generative Fusion means creators can predict the outcome, and rely on the technology for accuracy in multi-shot, multi-scene video with actions and occlusions and know the product insertion will be natural through elements like fine-grain control and more.
Generative AI is a fast-expanding area of creative development, but most of the new platforms to date are seen solely as productivity enhancers rather than category creators. Generative Fusion, using AI, can create an entirely new form of advertising that enables brands to reach their audience without interrupting the content, and content creators of any size can monetize their content with no loss for their artistic integrity or fear of diving too far into the realm of “influencers”.