The Case for Scaling Venture

Software now 22% of S&P 500; top firms are Google, Amazon, Nvidia — not GE.
~150 companies/year hit $100M revenue now, up from ~15 previously.
- Winner valuations: $1B → $100B+ territory.
- AI labs now build data centers and chip supply chains — capital intensity exploded.
Directly rebuts Roelof Botha: fixed-winners-per-year thesis is wrong.
- Outcomes larger, companies stay private longer, capital requirements vastly higher.
Picking alone is dead — winning the deal is now equally or more important.
- Scaled firms offer recruiting, GTM, legal, gov-relations, Fortune 500 CEO access.
Mega-funds can 10x a $1B fund; top-decile VC shows meaningful persistence.
Venture bifurcates: 4-5 mega-platforms + hyper-specialized boutiques; middle dies.
VCs back scale-disruption in every sector but resist it in their own.

Erik Torenberg (a16z General Partner) · 2026-02-11 · Read on a16z.com