EO StudioThe 8-Year Journey to Building a $4B AI Company | Synthesia, Victor Riparbelli
At a glance
WHAT IT’S REALLY ABOUT
Synthesia CEO on finding real AI video product-market fit
- Synthesia endured several unglamorous early years where the core tech was immature, yet the team maintained conviction that AI-generated video would become broadly valuable.
- The company learned that “cool” viral usage doesn’t equal retention, and sustainable growth comes from users with recurring, painful jobs-to-be-done rather than novelty seekers.
- After an initial attempt at AI dubbing proved to be a “vitamin,” Synthesia found a “painkiller” wedge in corporate communication videos where quality expectations and workflows differ from Hollywood-grade production.
- They combined bottoms-up product-led growth with top-down enterprise selling, describing an effective PLG “vacuum” that converts traffic into both self-serve revenue and high-quality enterprise leads.
- Riparbelli argues that hiring hungry, unconventional “underdogs” can outperform big-tech pedigree hires in early-stage environments, and predicts a long-term shift from text toward video/audio as default information media.
IDEAS WORTH REMEMBERING
5 ideasAnchor the product on a customer problem, not the model.
Riparbelli emphasizes users aren’t trying to “use AI”; they’re trying to make a video, train employees, or communicate a message. Framing around the job-to-be-done helps prioritize features that drive repeat usage and willingness to pay.
Viral interest is a signal—retention is the truth.
Synthesia saw many people generate demo videos because it was fun, but they didn’t return. The sustainable segment was the smaller cohort with recurring needs; digging into that cohort revealed the real value proposition.
A “vitamin” can be impressive and still fail commercially.
AI dubbing was valuable in theory, but it didn’t scale and demanded near-perfect quality from video professionals. Because customers wouldn’t “scream and yell” if it disappeared, it lacked the painkiller urgency needed for early traction.
Choose a wedge where quality thresholds match the alternative.
Synthesia’s breakthrough came when customers compared AI avatar videos to text docs—not to studio video. That reframed acceptable quality, changed feature requests, and unlocked a practical initial market (corporate communications) before broader ambitions.
Combine PLG and enterprise by designing for both—then hire for culture clash.
Bottoms-up access lets users self-qualify and creates enterprise-ready leads, but roadmap priorities can conflict (e.g., admin tooling vs SMB usability). Riparbelli notes the strategy also requires merging distinct “self-serve” and “enterprise sales” talent/cultures.
WORDS WORTH SAVING
5 quotesWe got turned down by like 100 investors, something like that, until we eventually sent a cold email to Mark Cuban.
— Victor Riparbelli
You need one person to say yes. That's all it takes.
— Victor Riparbelli
It was just clear to us that it was a vitamin, it wasn't a painkiller, right? Like, people liked it, they thought it was cool. Um, but if we disappeared, no one is gonna yell and scream.
— Victor Riparbelli
Early days of 2021, we have customers who will scream and yell if we shut down the product tomorrow. That's always, I think, a good acid test of, like, do you have something real or not, or do people kind of mildly care, but not really?
— Victor Riparbelli
I think that there is nothing that prepares you to really start a business like starting a business.
— Victor Riparbelli
High quality AI-generated summary created from speaker-labeled transcript.
