Product placement is one of the oldest marketing techniques introduced in this book.
Brands negotiate for years to get their products mentioned in movie scripts, and even a sponsored social media post can take over a month via phone call or email to arrange.
But startups, including Renbrand, a new venture launched in February by ad tech veteran Omar Tawakol, are starting to use generative AI to flip the script for product placement in video.
On Monday, Rembrandt closed an $8 million seed round led by Graycroft and UTA Ventures, the venture capital arm of United Talent Agency. The round included participation from L’Oréal’s VC fund (BOLD) and Good Friends, a venture fund run by the founders of Harry’s, Warby Parker and Allbirds.
“It makes sense to have some brands around the table because then we can understand exactly what their priorities are,” said Tawakol. “All money is green, but strategic investors are interesting because they give us more insight into what matters.”
However, it’s worth pointing out that Tawakol has been very good at reading his own tea leaves over the years as co-founder of data management platform BlueKai and later of enterprise voice assistant Voicea. there is. (He sold his BlueKai to Oracle in 2014, and in 2019 he sold Voicea to Cisco.)
draw a picture
Tawakol launched Renbrand to capitalize on multiple intersecting trends.
People are watching content from many creators online.
Small and medium-sized creators have a passionate audience that brands miss. Because the product placement negotiation process is so lengthy, advertisers often focus on larger influencers with large subscriber bases.
People hate being forced to watch intrusive ads. (Preroll’s anger is real.)
Also, while pre-recorded host-read ads can be dynamically inserted within audio content, it doesn’t work well with video and makes the format less scalable.
“Interrupting a show to read an ad for a minute or two is considered real in the industry,” said Tawakol. “Our approach is that creators just create organic content and don’t care about brands until they have a chance to choose what content goes into their videos in post-production.”
Fusion of physics and AI
Renbrand’s technology works through what the company calls “physics-informed” generative AI (i.e., virtual elements follow the laws of physics).
Rather than prompting an image or text generator, Rembrandt analyzes every frame in an existing video to understand light, shadows, reflectance, motion, occlusion and camera angles. increase.
From there, Rembrand turned photorealistic product images and animations into creator videos, like an OLIPOP soda can on a side table, or a Garnier shampoo bottle dancing in the frame and resting on a shelf in the background. can be inserted digitally into
“The product is there and people can watch it throughout the video, as opposed to ads that interrupt the show or can be skipped after five seconds that are not part of the content,” said Tawakol. rice field.
AdSense for AI
But how do creators get paid?
Renbrand has two business models.
The first is a two-sided network with influencers and content creators on one side and brands on the other. Creators can choose which brands they want Rembrand to feature in their videos, and brands can reach out to influencers who have a captivating audience for their products.
It’s a bit like AdSense, but for the placement of virtual goods. Renbrand is in growth mode, so the revenue share is generous. Sold on a CPM basis, the creator earns his 75% of net revenue on each placement.
Tawakol said the sales share will eventually even out, but the majority will always go to influencers.
In addition to its marketplace-based business model, Rembrand also has an up-and-coming API business licensing its insertion technology to streaming platforms and the entertainment industry to help automate the process of inserting virtual goods into their content. I’m here.
Create a plan
Renbrand plans to spend most of the new funding on engineering and AI enhancements.
For example, today, Rembrandt is primarily focused on creator content featuring people sitting and talking, shot indoors with walls and shelves in the background and a coffee table in the foreground. These are all great places to park a virtual soda can, a bag of crisps, and more. promotional poster.
The next step will be to manage film and TV content and outdoor scenes with more variables and moving parts, so more development work will be required.
Renbrand is also in hiring mode. He currently employs just over 20 people, all of whom are engineers, except for Mr. Tawakol himself and another one of him in a foreign sales position. Renbrand plans to assemble a full sales team, a go-to-market team and someone to manage influencer relationships later this year.
“But the primary goal is to generalize the applicability of AI,” Tawakol said. “All of these videos have to look correct and completely authentic, and that’s what we’re training our neural networks to do.”
