What Gets You Funded in AI: Insights from Felicis, 500 Global & Mayfield
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
This VC panel discussion features three prominent investors - Patrick Sallier from Mayfield, Tony Wong from 500 Global, and Nastasia Myers from Felicis Ventures - discussing the current AI landscape and what it takes to scale from seed to Series A funding. The conversation covers the dramatic shift in growth expectations, with companies now needing to achieve 2-3x faster revenue milestones than in previous years. (10:36)
- Main themes: AI voice technology emergence, exponential growth demands, business model evolution from SaaS to AI-native approaches, and the critical importance of real customer insights in an environment where anyone can build AI products.
Speakers
Patrick Sallier
Partner at Mayfield, a 55-year-old early stage venture capital firm that invests in seed through Series B rounds ranging from $2-15 million. He focuses on AI middleware and applications, particularly AI agents across various sectors, and helps founders navigate the evolving landscape of AI-enabled business models.
Tony Wong
Managing Partner at 500 Global with $2.1 billion AUM and 15 years of experience as a pre-seed and seed investor. He was an early investor in companies like Canva, Talkdesk, and Intercom, and more recently led investments in AI companies including Sakana AI, DeepInfra, and PlayAI (now part of Meta's Superintelligence Lab).
Nastasia Myers
AI-focused General Partner at Felicis Ventures, an early stage VC fund investing across inception through Series B with check sizes between $1-40 million. She has been part of hyper-growth journeys including Mercor, Canva, Notion, Runway, and Supabase, and focuses on AI across the full stack from infrastructure to applications.
Key Takeaways
AI Voice Technology Represents a Massive Market Opportunity
AI voice is emerging as one of the most compelling investment areas due to its ability to automate previously manual processes with superior user experience. Nastasia highlights that companies like Assort Health are achieving hyper-growth because AI voice can not only automate tasks but often deliver better NPS scores than human operators. (03:06) The ROI for these voice-native applications is incredibly high as customers are eager to adopt solutions that eliminate painful manual call handling processes.
Growth Expectations Have Dramatically Accelerated
The bar for Series A funding has risen significantly, with dev tools needing to reach $1M ARR in 2-3 quarters instead of a full year, while Gen AI applications need $2-3M ARR. (11:05) This reflects a shift where Series A rounds now resemble what Series B rounds used to be in terms of size and expectations. The conversion rate from seed to Series A has dropped below 20%, indicating investors are increasingly selective and demanding exceptional growth velocity.
Customer Insight Velocity Trumps Technical Capability
In an environment where AI tools make building easier than ever, unique market insights become the primary differentiator. Patrick emphasizes that founders need to have conversations with 100+ customers, not just a handful, to develop the texture and understanding that competitors lack. (35:58) This "learning velocity" allows founders to identify latent demand and build solutions that customers will actively pull rather than having to push products into the market.
Business Models Are Evolving Beyond Traditional SaaS
The AI era is disrupting traditional business models in two key ways: companies selling to model labs with massive budgets, and prosumer-focused businesses that democratize specialized skills. (19:19) Tony notes that we're seeing the emergence of "solo interpreters" - individual founders building with AI coding agents without raising traditional funding, representing a shift toward more experimental, RL-style startup building approaches.
Technical Moats Require Domain-Specific Integration Expertise
While cloud providers are ahead in AI infrastructure this time around, defensible businesses emerge from deep domain expertise and complex integrations with legacy systems. (23:23) Companies that can navigate esoteric verticals with specialized knowledge and build difficult integrations create switching costs that are difficult for generic AI solutions to replicate, providing sustainable competitive advantages.
Statistics & Facts
- Conversion rates from seed to Series A have dropped below 20%, representing a significant flight to quality among investors as growth expectations have intensified. (11:26)
- 500 Global manages $2.1 billion in assets under management and has made investments in over 2,100 companies during their 15-year history. (01:11)
- Traditional management spans of 7-10 reports are expanding to 15-20 people as AI tools enable flatter organizational structures and more efficient management workflows. (34:29)
