Mitchell Green: Why 50% of VCs Should Not Exist

Mitchell Green of Lead Edge Capital argues 50% of VCs destroy value, China wins AI, and gross dollar retention is the only metric that matters.

• Green believes ByteDance will do $70–100B in earnings within 5 years and is the most advanced AI company globally, underestimated by the West.
• China wins the AI race due to power infrastructure, PhD density, and ability to build nuclear plants in years vs. decades in the US.
• 50–60% of VCs add negative value; spinning out of Anthropic or OpenAI and raising $2B on a napkin idea is ‘complete lunacy.’
• Gross dollar retention thresholds: 90% = good, 95% = great, 98% = incredible; below 80% Lead Edge won’t invest.
• Lead Edge targets 2–5x returns in 3–7 years at ~25% IRR; roughly one-third of deals are secondary sales — they are explicit traders.
• DPI is math, marks are opinions: liquidity windows open and close, sell 20–30% of winners when windows open rather than waiting for a unicorn exit.
• Levered companies cannot innovate during tech disruptions — the pattern from e-commerce (Sears, Kmart dead; Walmart survived) repeats in AI.
• Lead Edge found Grafana Labs via cold call when it was a bootstrapped $12M revenue company; Lightseed was the only co-investor at entry.

2026-03-07 · Watch on YouTube