The Trillion-Dollar Industries AI Is Disrupting: Voice, Law & the End of the Billable Hour
with Mati Staniszewski
14 Jul 20264 min read31m
TL;DR
ElevenLabs CEO Mati Staniszewski reveals the company hit $600M ARR — going from $100M to $200M in 10 months, then $200M to $300M in just 5 months — by building an AI-native voice platform for enterprises replacing broken IVR systems. Staniszewski argues voice is identity and IP, describing how ElevenLabs has paid over $22 million back to voice creators through its marketplace, and worked with the James Earl Jones estate to preserve iconic voices. The biggest competitive moat, he says, isn't scale but proprietary labeled audio data and research architecture that frontier models like OpenAI and Anthropic have struggled to match.
Key Moments
Mati Staniszewski
“Roughly 20 months to get to the first 100 million in ARR. Roughly 10 months to get to 200, 5 months to get to 300. Um and um and that's how we closed end of the last year and now we are at 600.”
Staniszewski is walking through ElevenLabs' revenue ramp timeline since launching their first human-sounding TTS model in early 2023.
“The a specific emotional example, we work with a lot of um financial services companies uh Revolute, Clara, Pogbank. And some of the frequent case not in all of them is of course how you remind people about payment or that you collect the de that from the people that aren't answering and frequently people would naturally feel ashamed of telling the real situation with AI people are much more open to share what actually happened”
Staniszewski is explaining a behavioral shift he's observed: people are more honest with AI voice agents than with human agents when discussing sensitive financial situations.
“Today, we paid back over $22 million back to the community of talent.”
Staniszewski is describing ElevenLabs' voice marketplace where creators upload and license their voices, positioning it as a new revenue stream for voice-over actors.
“It's on the research side. It's the architecture that matters, not the scale. You really need to change how the model operates. Two, you need very specific data that there's of course a wide set of data out there, but it's unlabelled data and where we spend a lot of time. So we build a internal team of over thousand contractors that label all those audio assets to make them to make them good.”
Staniszewski is explaining why ElevenLabs has been able to out-compete OpenAI and Anthropic on voice models despite their much larger scale.
Mati Staniszewski is the co-founder and CEO of ElevenLabs, an AI audio and voice platform he started in 2022. Under his leadership, ElevenLabs grew from zero to $600 million in ARR in roughly four years, building what he describes as the leading text-to-speech and voice AI platform. The company powers voice agents across financial services, healthcare, and telecom, and has partnered with celebrities and estates including Matthew McConaughey and the James Earl Jones estate.
Takeaways
1
People are more honest with AI agents than humans In financial services debt collection use cases with clients like Revolut and Klarna, ElevenLabs found that customers disclose their real financial situation more freely to AI voice agents than to human agents, removing an emotional barrier that previously blocked resolution. This shifts the value prop of voice AI from cost-cutting to actually better outcomes.
2
Embed engineers in every team, including non-technical ones ElevenLabs places engineers inside legal, talent, and go-to-market teams — not just product and engineering. These engineers serve dual roles: building internal automations and acting as security reviewers for AI-generated code that non-engineers deploy, preventing silent production leaks.
3
Voice is licensable IP, not just a biological trait ElevenLabs has paid $22M back to voice creators through its marketplace and struck deals with celebrities including Matthew McConaughey and the James Earl Jones estate to extend iconic voices across languages and interactive use cases. For creators, this opens a passive income stream; for companies, it's a new class of IP asset to acquire or protect.
4
Voice AI revenue can accelerate faster than SaaS ElevenLabs took 20 months to reach $100M ARR, then just 5 months to go from $200M to $300M — and is now at $600M. This compressing growth curve suggests voice AI adoption has crossed a tipping point where enterprise deals compound quickly once a platform proves reliability.
5
Labeled audio data beats raw scale in voice AI Staniszewski argues ElevenLabs out-competes frontier labs not by spending more on compute but by maintaining a team of over 1,000 contractors who label audio assets. In voice AI, the architecture and data quality matter more than parameter count — a meaningful counter-narrative to 'scale is all you need.'
6
Voice cloning safeguards require three simultaneous layers Staniszewski describes ElevenLabs' approach as: (1) trace everything generated so action can be taken retroactively, (2) moderate at both voice and text input levels to block commercial misuse in real time, and (3) build public detection tools that identify AI-generated audio regardless of which model produced it. Any one layer alone is insufficient.
7
No PMs, ever — engineers own the full product loop ElevenLabs has never hired a product manager. Instead, they hire people who are expert in one discipline but conversant in at least one other, and rely on AI to bridge the remaining gaps. As AI lets individuals move from amateur to advanced in adjacent skills, the classic PM coordination role becomes redundant.