Much has been made of AI-powered videos that can either make a real person speak different words or other languages, or generate a synthetic person that can speak on-demand. A few years ago former football star David Beckham helped push the idea into the mainstream via a video from Synthesia, where he spoke different languages as part of a Malaria charity campaign.
Aside from being a source of gimmicks, as well as more serious disinformation, these tools have several real-world applications, and the market is booming. The aforementioned Synthesia raised $90 million this year for its platform, which is used for several things, including generating product videos, as it can generate personalized videos from plain text. To date, it’s now raised a total of $156.6 million.
However, while it’s possible to generate these videos, doing so millions of times over for millions of products is not easy, and can require a lot of manual intervention.
Now Oxolo, a German startup out of Hamburg, has hit up on a novel way of doing so, using generative AI.
Here’s how it works: Say a company wants to do a corporate training. Using Oxolo, it can build a video for an employee type, based on management level, gender, geography, etc. It can then send out videos that will automatically A/B their own performance. If an employee clearly loses interest in one video, then the video is adjusted and replaced with a better performing one on the fly. The same application can be made for product videos, for instance, creating better and better performing product videos.
“Other platforms have avatars and maybe templates and you have to manually put in the text,” explained Elisabeth L’Orange, co-founder and CCO of Oxolo, over a call. “With Oxolo you just take a product link — say for socks — paste it into Oxolo, hit generate and you get the whole product video. You don’t need to add anything, and you can edit it. You don’t need to manually think about what you need to do.”
Oxolo was co-founded by L’Orange, a former venture partner with Neuhaus Partners, and Heiko Hubertz, former founder & CEO of Bigpoint games. Oxolo claims its AI can integrate images, context, text-to-speech and human-based avatars.
As a result of this rather magic approach, Oxolo has now raised a €13 million Series A funding round led by DN Capital. Also participating were German angel investors, including Christian Vollmann, Johannes Plehn, Thomas Hagemann, Stefan Wiskemann, Alex Täubert and Mischa Rürup.
According to Market Growth Reports, the current market size for tools that enable the creation of video advertising content is estimated at $1.0-$1.5 billion in 2023, and is expected to grow at a 15-20% CAGR over the next five to seven years to $3.4 billion in 2028 and $4.3 billion by 2030.