Premise VC: How to Spot Durable AI Companies and the Power of Product Velocity

Vanessa Larco, former NEA partner and co‑founder of Premise VC, explains why the companies OpenAI won’t kill are the ones that ship fast and keep shipping. She cites OpenAI’s recent deprecation of Sora and the quiet removal of checkout inside ChatGPT as evidence that not every OpenAI feature succeeds, leaving a larger field for startups. Larco stresses that Premise backs only technical founding teams, insisting at least one co‑founder must be highly technical to stay ahead of new AI frameworks. She argues that product velocity is the only early edge; teams that ship in a month will be copied in a month, so continuous delivery is essential. Larco also highlights the shift to a distribution‑first playbook, where founders must own growth and distribution as much as product, especially in dev‑tool markets. She describes Premise as a two‑person fund that operates like a startup, treating fundraising, deployment, and portfolio support as product releases and mapping CAC to LTV. Finally, she warns that conviction becomes harder as the decision‑making group grows, recommending a clear process and “disagree and commit” culture.

The episode covers how to spot durable AI companies, the importance of shipping quickly, the need for technical founders, the new distribution era, and the startup mindset required to run a VC fund. It offers practical takeaways for anyone looking to invest in or build AI‑driven products.