The 10x founder and the HUNCH framework: Jeff Bussgang on AI-era product-market fit
Key points
Key takeaways from an Infinite Loops interview with Jeffrey Bussgang, co-founder and general partner of Flybridge and Harvard Business School professor, on his new book The Experimentation Machine (October 2025):
Startups as experimentation machines. Bussgang argues that great startups are not static companies but systems for running disciplined experiments, with AI massively compressing the time and cost of learning.
The 10x founder. AI tools let founders execute roughly 10 times faster than five to ten years ago; as technology moats get harder to build, execution velocity becomes the differentiator that is hardest to replicate.
The HUNCH framework. Product-market fit is judged on a hair-on-fire value proposition (a top one or two priority for the customer), high usage (week-in, week-out, not downloads), a strong net promoter score, low churn with enduring cohorts, and high unit economics (lifetime value vs customer acquisition cost).
Usage beats vanity metrics. Citing research by a Harvard colleague with OpenAI researchers, he notes ChatGPT now reports over 700 million weekly average users — roughly a tenth of the planet — as evidence of real, sustained usage.
The AI-era discovery process. Modern founders can build AI personas of their ideal customer profile for discovery, prototype with tools such as Lovable or Replit in hours using English as the programming language, and use AI agents to generate outbound emails and A/B-test pitches.
Discipline over story. Recalling his time at Open Market after its 1996 IPO, he notes public markets only care about the numbers; at Flybridge he applies that same public-market rigour to the private boards he sits on.
Head in the clouds, feet on the ground. Flybridge looks for founders who are visionary but also execution-minded and results-oriented.
Read more: YouTube — Infinite Loops
