Most VCs try to pick winners. Kingmakers try to create them — backing a company early with a check large enough to crown it the category leader before the market decides.

Not all kingmaking is the same. Sometimes capital is a genuine prerequisite — frontier AI, defense, infrastructure that can't be built incrementally. But more often it's deployed as a weapon: to buy distribution, subsidize growth, and manufacture the appearance of inevitability before anyone has earned it.

Capital creates signal. Signal shapes belief. Belief drives behavior. That's why it can work in specific conditions. 

In capital-intensive categories, money is a raw material, not a strategy. In true network markets, capital can subsidize one side long enough to lock in the other. Enterprise buyers genuinely prefer well-capitalized vendors — stability is a buying criterion. And when the underlying thesis is right, capital accelerates genuine momentum. 

Databricks is the landmark case: a16z led every round from the $14M Series A onward after they asked a16z for $200k. They pushed seven PhDs to think beyond a $50M academic exit. Thiel bet on SpaceX because he believed rockets were a monopoly waiting to happen. The $671M became $18.2B because the thesis was right. The capital was conviction made tangible.

But those conditions are rarer than the strategy is deployed. Capital cannot manufacture product-market fit — it only buys time and talent to find it. Large balance sheets breed parallel bets, premature hiring, and confusion between growth and product validation. Premium valuations compress returns even on winners. And the best founders have leverage to take less capital; the ones who take kingmaker checks often need them to paper over weak unit economics.

The base rate is damning. Most $100M+ rounds since 2018 have not produced commensurate outcomes. Inflection, Adept, and Character are early evidence that kingmaking in AI is already failing. Convoy, Bird, and Fast raised enormous sums and failed anyway. 

Figma beat InVision. Cursor is beating the incumbents. Customers, not capital, get the final vote.

This isn't new. In 1903, Samuel Langley had the full backing of the Smithsonian, $50,000 in War Department funding, and a team of engineers. The Wrights had roughly $1,000 and a wind tunnel they built themselves. When the Aerodrome collapsed into the Potomac twice, the Wrights flew at Kitty Hawk nine days later.

There's also a cost the kingmakers don't absorb. Kingmaking concentrates capital in a handful of firms, inflates valuations across the stage, and forces every other investor to pay up or sit out. 

It creates zombie unicorns — companies too big to fail gracefully but unable to grow into their valuations, locking up talent and capital for years. It pulls LP money toward mega-funds even though fund size is inversely correlated with returns. And it trains the next generation of founders to optimize for the next round rather than the customer.

The kingmakers themselves can survive on brand and secondaries. The companies, the LPs, and the ecosystem absorb the cost.

The current AI cycle is the largest live kingmaking experiment ever run. The next 24 months will produce real evidence. But the history is clear: the best venture returns have come from disciplined check sizes and founder-driven momentum, not manufactured inevitability.

The market still decides. Capital just speeds up the answer — and often the answer is no.