20VC: Inside Accel's $4BN Growth Investing Machine | Cursor is Dead is Total BS: Here is Why | What Missing Rippling and ElevenLabs Taught Us | Are $2BN-$10BN IPOs Dead | Why Now is a Great Time to be Thoma Bravo with Miles Clements
Summary
Miles Clements (Accel) discusses how growth investing is changing in an era of AI, arguing investors should evaluate AI businesses by both time-to-value and durability of that value. He defends Cursor against narratives that it is "dead," emphasizing multi-model support and agentization as durable product advantages for developer tools. The conversation covers Accel's approach to ownership across stages, lessons from missed deals (e.g., ElevenLabs, Rippling, ServiceTitan), and why many companies avoid $2–$10B IPOs today. There is also practical advice for founders on choosing board members: favor humility, measured input, and quality of feedback over volume of commentary.
Key Takeaways
- 1Assess AI companies on two dimensions: time-to-value and durability of value.
- 2Multi-model architectures and agentization are strategic advantages for developer platforms like Cursor.
- 3Top VCs combine rigorous rules with pragmatic exceptions for outlier opportunities.
- 4Mid-cap public outcomes ($2–$10B) are less attractive today unless there's a clear path to much larger scale.
- 5The quality and demeanor of board feedback matters far more than how vocal a board member is.
Notable Quotes
"Focus on hitting singles and doubles and let the home runs take care of themselves."
"90% of Cursor users are daily active users of the agent product."
"95% of developers switch models on a daily basis."
"Investing is an art and a science. The science is understanding how to properly value a company and the art is understanding when to break the rules."
"You don't need an investor micro managing you through all the little decisions."
"He's got this way of sort of saying, let me politely make an observation and you can sort of choose to accept it or reject it."
"I think there's generally an inverse correlation between how vocal somebody is and how helpful they actually are."
"I think 11 labs is a clear company that we wish we had been a part of. We haven't spent enough time with the founder, which is our loss."
Episode questions
How should investors assess AI company value in a fast-changing market?
Use a two-dimensional framework: measure time-to-value (how quickly users extract benefit) and durability (how persistent that benefit is). Prioritize businesses that are strong on both or have clear paths to durable value even if deployment is slower (e.g., enterprise legal/accounting AI).
Is Cursor's reliance on third-party models (Anthropic/OpenAI) a fatal flaw given cost concerns?
No — Miles argues Cursor's multi-model architecture mitigates vendor risk and provides product advantages: developers switch models frequently, so multi-model support and agentization create a compounding flywheel and resilience to single-vendor cost changes.
Should VCs buy late-stage companies at high prices to achieve the ownership needed for fund returns?
Accel believes being multi-stage and multi-strategy allows them to build ownership over time (early to IPO) rather than only buying late; they also accept that very large outcomes today can compensate for smaller percentage ownership compared to prior eras.
Why do many companies avoid IPOs in the $2–10B range today?
Miles says companies often wait because it's murky for public comparables below ~$5B, and public markets have been unforgiving for mid-cap tech — founders prefer to stay private until they have clearer line-of-sight to larger scale or more favorable market conditions.