Zephyrnet Logo

Synthesia’s Series C – And a Few Lessons Learned Building a Generative AI Startup

Date:

Synthesia, the market leader in AI video generation, is announcing this morning an exciting $90M Series C financing.

This is a great milestone for the company, making it a newly-minted unicorn and adding strong new partners (Accel, Nvidia, and some great individuals) to the team. Beyond the fundraising momentum and accolades, however, Synthesia has first and foremost been building a really impressive business. Over 35% of the Fortune 100 now use Synthesia’s enterprise offering, and over 50,000 businesses use its self-serve product.

Much to their credit, the company’s founders (Victor, Steffen, Lourdes and Matthias) were very early to the Generative AI wave, starting the company at a time (2017) when there was significant technology risk and limitations, to the point that what they were doing was probably a bit weird, if not outright crazy. Several years later, by the time FirstMark led the Series A early 2021, the term “generative AI” was still not a thing – my blog post announcing the round used the term “video as code” and in my 2021 MAD landscape, Synthesia appeared in a box we called “synthetic media”, for lack of a better term.

Anecdotally, this Series C is another example of the growth market showing some sign of life lately, at least for A+ companies, in the same vein as Pigment’s recent announcement, albeit perhaps less surprisingly given the hype around Generative AI.

Given the explosion of Generative AI and the flood of brand new startups that were created in the space over the last 6 months, there are interesting early lessons to learn from Synthesia’s journey so far:

Proprietary AI does matter: In the current feverish AI hype cycle, many people have been quick to jump to conclusions – building an AI company means leveraging OpenAI or other foundation model providers; there’s no competitive advantage or sustainability to build around technology; AI is getting commoditized, etc. In sharp contrast to that line of thinking, Synthesia has built a very strong AI applied research team over the years (co-founder Lourdes Agapito and Matthias Nissner are both prominent AI academics). While it can leverage third party foundational model technology opportunistically, the company has developed industry-leading, proprietary AI in the field of AI-generated, life-like avatars, making it clearly stand apart from any other solution on the market.

A great product is broader than just AI: While it is known for the quality of its AI avatars, the Synthesia platform is much broader than that. It is an enterprise-grade AI video generation platform that enables teams to collaborate and build great videos for all sorts of business purposes. Avatars are not even required. The Synthesia platform replaces cumbersome processes where actors, directors and cameras are required. It also replaces boring PDF content and even Powerpoint. There’s a lot of infrastructure, workflow, permissioning, collaboration, security etc involved in building such a product.

Solving business problems: In every AI hype cycle I have experienced, there is a first phase where many get enamored with the power of the technology and dream up a million use cases – always a fun phase. However, quickly enough, the dust settles and a simple reality emerges: It is not about the AI, it’s about what business problems one solves with AI. As the cycle progresses, it’s always telling to see vendor website messages evolve from being very AI-centric at first, to being very business solution-centric a few months or years later, barely mentioning AI on their home page. The Synthesia founders understood this reality very early on, and a big part of the success of the company has been their instinct to eschew flashy use cases, and instead point the technology at core enterprise problems and big, Fortune 100 type customers.

Congrats to the team on this great milestone, and excited for the journey ahead!

Oh – and here’s my use case:

spot_img

Latest Intelligence

spot_img