A former Kleiner Perkins partner who helped close Google's Series A just said the VC playbook he helped write is broken.
The Summary
- Former Kleiner Perkins VC who worked on Google's funding reveals that Apple, Amazon, Microsoft, and Google combined raised less than $35M in VC before becoming trillion-dollar companies
- The "get big fast" model emerged in the late 1990s at firms like Kleiner Perkins and became the default playbook, despite evidence it wasn't necessary for building valuable companies
- After running his own VC-backed company (Good Technology), he now argues the pressure to raise massive rounds and hit $20B+ valuations creates more problems than it solves
The Signal
The math is stark. Apple raised under $1M. Amazon about $8M. Microsoft $1M. Google $25M. Total: $35M in VC funding (about $74M in today's dollars). Combined current value: $14 trillion. That's a 189,000x return on capital for the entire foundation of modern tech.
Now compare that to 2024-2025, where Series B rounds average $50-100M and AI startups raise hundreds of millions before they have real revenue. The article's author, who sat across from Larry and Sergey reviewing their term sheet, watched this transformation from inside the machine. He helped build the "get big fast" doctrine at Kleiner Perkins starting in 1997, right as Amazon went public.
"The expectation for me was that the company would be worth $20 billion. That was a ma—" [The article cuts off, but the point lands.]
The timing matters. This isn't some bootstrapping evangelist or lifestyle business advocate. This is someone who lived at the center of power when the current VC model was being invented. He's not arguing against venture capital. He's arguing the dial got turned too far, too fast, and now founders are trapped in a game where anything less than decacorn status feels like failure.
Here's what changed between Google's $25M raise and today:
- Cloud infrastructure dropped server costs by 90%+, but round sizes went up 10x
- Distribution got cheaper through social platforms and app stores, but marketing budgets exploded
- Remote work made global talent accessible, but comp packages ballooned to match SF rates everywhere
- AI tools made small teams more productive, yet headcount expectations stayed stuck in the pre-LLM era
The capital efficiency argument hits different in 2025 than it would have five years ago. LLMs can now do the work of entire departments. A two-person team can build and ship products that would have required twenty people in 2019. But the VC model still assumes you need to hire those twenty people, raise $50M to pay them, and burn through runway racing to the next round.
The Implication
Watch for a quiet counter-movement. Not loud anti-VC posturing, but founders who've seen the cap table math and realized they can build bigger companies by raising less. The tools exist now to do what Google did with $25M, but with even less capital. AI agents handle customer service. Code generation tools shrink engineering teams. No-code platforms eliminate entire layers of infrastructure spend.
The question isn't whether you should raise VC. It's whether you need to raise as much as everyone assumes you do. If a former Kleiner partner who helped fund Google is saying the playbook broke, founders should at least run the numbers on what capital-efficient growth looks like for their specific business.