Benchmark Capital invested in Uber in 2011. Three years later, it invested in Lyft. By the time both companies were racing toward their respective IPOs, Benchmark held meaningful stakes in two firms that were actively trying to destroy each other, burning through hundreds of millions of dollars per year in a subsidy war for drivers and riders.
The press called this a hedge. Benchmark called it portfolio management. Both explanations miss what was actually happening.
The Setup
By 2014, ride-hailing was no longer a speculative category. Uber had proven the model worked. The question was whether it would remain one company’s prize or become a contested market. Lyft had launched its peer-to-peer service in 2012 and was growing in cities where Uber had a head start. For Benchmark, the choice was not whether to invest in ride-hailing. The choice was whether to let a competitor fund Lyft while sitting on the sidelines.
The standard investor logic here would be: we have Uber, Lyft is a threat to Uber, therefore we shouldn’t fund Lyft. That logic is wrong, and understanding why it’s wrong tells you most of what you need to know about how top-tier venture capital actually works.
What Happened
Benchmark’s Lyft investment was not primarily about financial diversification. A portfolio hedge at the fund level would require the two investments to behave like uncorrelated assets. Uber and Lyft were almost perfectly correlated: both rose and fell with the same regulatory fights, the same labor classification battles, the same pandemic collapse in ride demand. As a hedge, it was nearly useless.
What the dual investment actually did was buy Benchmark a seat at every important table in the category.
When Uber’s board was navigating the Travis Kalanick crisis in 2017, Benchmark was already a power broker. When Lyft was negotiating terms with potential acquirers, Benchmark had context that no single-company investor could match. When both companies were fighting regulators over driver classification, Benchmark’s partners understood the regulatory landscape from two simultaneous vantage points.
Information, in venture capital, is almost always worth more than the incremental return from any single bet. The firms that consistently generate outsized returns do so not because they pick better companies in isolation but because they understand markets more completely than anyone else, and the best way to understand a market completely is to be inside multiple competing interpretations of it at once.
This is the mechanism that the “hedging” explanation obscures. Sequoia has done this repeatedly: funding both Google and YouTube before Google acquired YouTube, holding positions in companies that later competed directly. Andreessen Horowitz has been explicit about wanting to own the entire category rather than pick a single winner. The language of portfolio theory makes this sound defensive. It is actually offensive.
The Economics of Category Ownership
Here is the math that makes this rational. If a market produces one major winner, and you bet correctly on that winner, your return is determined by your ownership stake and the winner’s terminal value. If a market produces two major winners (as ride-hailing did, eventually), and you own meaningful positions in both, your aggregate return captures more of the total value created by the category regardless of which company wins. But more importantly, if the market consolidates and one acquires the other, you participate in both sides of the transaction.
This is not a subtle point, but it is consistently underappreciated. The relevant competition for a top-tier venture fund is not between Uber and Lyft. It is between Benchmark and the next fund that might have invested in Lyft instead. By owning both, Benchmark denied a competitor the information and influence that would come from owning Lyft alone. The opportunity cost of not investing in a market leader’s primary competitor, when you have the chance, is higher than most founders realize.
Founders, understandably, often object to this arrangement. When Uber discovered Benchmark’s Lyft investment, the relationship became complicated. There are real conflicts: a board member who has fiduciary obligations to two competing companies cannot fully serve either. Benchmark’s partners were not on Lyft’s board precisely to manage this, which tells you something about how the firm valued influence versus governance. Governance can be delegated. Proprietary market intelligence cannot.
The Information Asymmetry Angle
Consider what Benchmark knew by 2016 that no single-company investor could know: Uber’s internal unit economics (from board access), Lyft’s internal unit economics (from their investment), and the competitive dynamics between them from both sides simultaneously. They knew which subsidies were working, which cities were profitable, which regulatory strategies were effective. This is not information that shows up in public filings or press releases.
This advantage compounds. When evaluating the next transportation company, or the next logistics startup, or the next gig-economy labor platform, Benchmark’s partners were reasoning from a depth of proprietary data that no competing fund could replicate. The Lyft investment was, in part, a research budget for understanding a category that would shape several subsequent investment decisions.
The same logic explains why major funds back multiple cloud infrastructure companies, multiple fintech lenders, multiple vertical SaaS plays in the same industry. The “hedge” framing implies risk reduction. The actual goal is information maximization across the category, which produces better decisions downstream and a defensible reputation as the fund that understands the space.
What Founders Should Actually Know
The lesson for founders is not that VCs are adversarial, because the best ones are genuinely trying to help their portfolio companies succeed. The lesson is that the incentive structures are more complex than a simple alignment of interests between a fund and a single company.
When a VC asks probing questions about your competitive strategy, your pricing, your customer acquisition costs, they are building a model of your market. That model does not belong exclusively to you. A VC who has seen fifty B2B SaaS companies explain their sales cycles has a more accurate model of B2B SaaS sales cycles than any single founder, and that model is an asset that accrues to the fund, not to any one portfolio company.
This is not a scandal. It is the mechanism by which information flows through the venture ecosystem and (mostly) toward productive use. But founders who treat their VC relationships as purely bilateral arrangements, without considering the portfolio context, are missing something important about the dynamics of the room they’ve walked into.
Benchmark’s Uber and Lyft positions were not a hedge. They were a claim on a category. When both companies ultimately went public within weeks of each other in 2019, Benchmark had spent nearly a decade understanding ride-hailing from the inside out. The returns were secondary to what the investments made possible next.
