This episode unpacks the operational playbook of ElevenLabs, revealing how CEO Mati Staniszewski navigates the tension between foundational AI research and rapid product deployment to dominate the voice AI market.
Introduction: The ElevenLabs Growth Engine
- Mati "Maddie" Staniszewski, co-founder and CEO of ElevenLabs, opens by discussing the company's rapid expansion from voice synthesis to licensed music generation and AI agent platforms. He attributes their high shipping speed and quality to a combination of top-tier research talent and a unique organizational structure.
- The company is built on a foundation of strong AI research, led by co-founder Piotr Dabkowski, which enabled the creation of their initial text-to-speech model capable of understanding context and conveying emotion.
- To maintain velocity, ElevenLabs operates with approximately 20 small, independent product teams of 5-10 people each. This structure fosters high ownership and allows the company to move quickly, despite occasional risks of duplicative work.
- Strategic Insight: For investors, ElevenLabs' model of decentralized, highly autonomous product teams offers a compelling blueprint for scaling AI companies. This structure allows for parallel innovation across multiple product lines without creating bureaucratic bottlenecks.
Balancing Foundational Research with Product Development
- Maddie details the inherent conflict between long-term research goals and immediate product demands. He shares a key operational principle his team developed to manage this tension, using the example of a user request for a "speed slider" on voice generation—a feature the research team initially resisted in favor of a more elegant, AI-native solution.
- The team initially resisted building a simple speed toggle for nine months, hoping to solve the problem at the core research level. When the research solution proved elusive, they implemented the simpler product feature, which was highly successful.
- This experience led to a new internal guideline: "If we think the research work will take more than three months, then the product team can do anything they want to start adding other models or extensions."
- Actionable Takeaway: This "three-month rule" is a critical heuristic for AI companies. Researchers and investors should note that balancing pure research with pragmatic, short-term product wins is essential for maintaining market momentum and user satisfaction.
Building a Global Talent Engine
- Maddie explains that ElevenLabs' European origins were fundamental to its creation, inspired by the poor quality of monotone voice dubbing in foreign films in Poland. This perspective drove their strategy to build a global, distributed team from day one to access the best talent wherever it exists.
- The company started fully remote to hire the best researchers and engineers from across Europe and Asia, deliberately avoiding a Silicon Valley-centric approach.
- They adopted unconventional hiring methods, discovering a top researcher who had developed an open-source text-to-speech model while working in a call center.
- As the company scaled beyond 30 people, they established physical "hubs" in London, Warsaw, and San Francisco to help new hires immerse themselves in the company culture while still supporting a remote-first ethos.
Organizational Design: Flat Hierarchy and No Titles
- To foster agility and meritocracy, ElevenLabs adopted a flat organizational structure and eliminated formal titles a year ago. Maddie explains this approach is designed to empower individuals based on impact rather than tenure, allowing passionate and talented people to rise quickly.
- The structure minimizes hierarchy, with a small layer of leads overseeing subdivisions (Research, Creative, Agents, Go-to-Market) and flat, small teams underneath.
- This model encourages ownership and allows the company to deploy the best people to engage with partners and customers, regardless of their formal position.
- A key challenge is ensuring leads can manage complexity and facilitate cross-team collaboration. To combat distraction, the company intentionally limits access to certain communication channels like Slack to enforce focus.
Engaging the Creative Industry: From Disruption to Collaboration
- Maddie outlines ElevenLabs' strategy for working with the creative industries, particularly voice actors and music labels, which were initially resistant to AI. Their approach focuses on collaboration and creating shared economic incentives rather than outright disruption.
- They launched a Voice Marketplace where creators can clone their voice, share it on the platform, and earn money when it's used. The marketplace now has nearly 10,000 voices and has paid out $10 million to creators.
- Negotiating with major music labels to license their catalogs for ElevenLabs' music model was an 18-month process. Maddie notes that setting "forcing functions" or deadlines was crucial to reaching an agreement.
- Maddie's Perspective: "Just speaking through exactly the technology, showing the examples and kind of avoiding this initial knee-jerk reaction that AI is bad has been tremendous."
Transitioning to Enterprise: From PLG to a Hybrid Sales Model
- ElevenLabs began with a product-led growth (PLG) model, where the product itself drives user acquisition, attracting significant inbound interest from enterprises. Maddie describes the company's deliberate and challenging transition to serving large enterprise customers.
- An early experiment to have engineers handle sales failed, leading them to build a hybrid go-to-market team that is "80% sales, 20% engineering."
- Working with early enterprise partners like Hippocratic AI revealed the need for an entire orchestration layer, combining speech-to-text, LLMs, and text-to-speech with integrations for telephony and deployment.
- This shift required building robust infrastructure focused on security, compliance, and reliability—foundational elements for enterprise trust. The cultural shift from rapid PLG cycles to long enterprise sales cycles (6-12 months) was a significant internal challenge.
Managing the Product Lifecycle for Enterprise Scale
- To serve both agile creators and demanding enterprise clients, ElevenLabs developed a dual-track product development system. This allows them to ship innovative features quickly while ensuring stability for production-level deployments.
- Products are clearly delineated as "alpha" or stable. Enterprise customers can opt-in to access alpha features, understanding they may be less reliable, giving them a choice between cutting-edge innovation and stability.
- Internally, teams are designated as working on "pre-product-market fit" or "post-product-market fit" products. Pre-PMF teams are tasked with shipping rapidly to find market validation within a six-month window, while post-PMF teams focus on long-term stability and evaluation.
The CEO's Scaling Challenge: Aligning Incentives with Strategy
- Reflecting on his journey as CEO, Maddie identifies the transition from a passion-driven startup to a 350-person company with a formal sales machine as his hardest challenge. He realized that as the company grew, incentive structures, particularly sales commissions, began to drive behaviors in ways that could diverge from the overall company strategy.
- He learned that sales commissions are a "lagging indicator of strategy." If not carefully designed, they can incentivize short-term gains that conflict with long-term goals.
- To resolve this, ElevenLabs made its strategy explicit to the sales team, empowering them to flag deals that, while profitable, might be strategically unwise—such as selling to a foundational model competitor. The company commits to paying the commission even if they decide to kill the deal.
- Strategic Insight: For investors evaluating scaling AI companies, the alignment between sales incentives and long-term strategy is a critical, often overlooked, indicator of operational maturity and sustainable growth.
Conclusion
This discussion provides a masterclass in scaling an AI company by balancing research with pragmatic execution and strategic team design. Investors and researchers should track how AI companies structure their research-to-product pipeline and align incentives, as these operational details are crucial indicators of long-term market leadership and adaptability.
