Venture Capital is Broken. Here is How to Fix It.

Venture capital is not dying. It is stuck.

The industry has produced extraordinary outcomes over the last three decades, funded most of the software, biotech, and consumer internet companies that now shape daily life, and created one of the largest engines of wealth generation in modern history. I am not here to tell you VC is worthless. I work in private markets because I believe in them.

But the specific structure of venture capital in 2026 is broken in several concrete ways. The return distributions are worse than most LPs realize. The access is still gatekept by geography and network in ways that make no economic sense. The information asymmetries are larger than in any other major asset class. The incentives inside firms push partners toward consensus rather than conviction. And the research infrastructure, the thing that should make all of these problems easier to navigate, is two decades behind where it should be.

Each of these is fixable. None of them will be fixed by the people who benefit from the current structure. This post is the direct version of what is broken and what it would actually take to fix it.

The Returns Problem

Start with the uncomfortable truth about venture returns.

The industry-level story (outsized returns, fund-of-funds delivering top quartile performance, 2x to 3x cash-on-cash over 10 years) is true for a small number of top firms. For the median VC fund, it is not. Public benchmarks have consistently shown that median venture returns barely beat public market indexes net of fees, and often underperform.

This is not a secret. It is openly discussed inside the industry and conveniently forgotten in the marketing materials. Most LPs get returns that would be easier to achieve by investing in the Nasdaq 100 with no J-curve, no illiquidity, and no fees.

The top quartile produces real excess returns. The rest of the industry rides the marketing halo of the top quartile while delivering results that do not, in aggregate, justify the structure.

What Would Actually Fix This

Brutal truth-telling in LP reporting, using comparable public market benchmarks, using realistic cash-on-cash methodologies rather than paper markups. Most of the reported performance in VC is theater. LPs who push for honest reporting and invest based on real returns rather than manager stories are the ones who would force this.

Also: smaller funds in most cases. The size of typical VC funds today has grown past the point where fund-level returns make sense. A $1B seed-stage fund cannot return 3x without deploying at a scale that distorts its investment selection. Smaller funds would produce better returns for LPs but worse economics for GPs, which is why they are rare.

The Access Problem

The second structural failure. Access to the best deals in venture is still gatekept by personal network, geography, and brand in ways that are economically indefensible.

The best founders do not get their capital priced by efficient markets. They get their capital priced by which specific partners at which specific firms they happen to know, or be introduced to, at the moment they decide to raise. This produces systematic mispricings, systematic exclusions, and systematic misallocations of capital.

If a founder is in Lagos instead of Palo Alto, their seed round may clear at half the valuation of an equivalent team in California, for no reason related to the actual quality of the company. If a founder does not have a degree from a specific set of universities, their inbound from top-tier funds will be materially lower. If they are not connected to the specific network that a partner respects, warm intros do not materialize.

This is not about diversity for its own sake. It is about economic efficiency. The sorting mechanism currently in place leaves enormous value on the table, because it mispprices large swaths of the opportunity set.

What Would Actually Fix This

Better intelligence infrastructure that makes talented founders visible regardless of geography or network. Modern AI-powered sourcing platforms can surface companies from anywhere, based on real signals (product quality, customer traction, founder execution) rather than social proximity.

This is already starting to happen. Funds that take data-driven sourcing seriously are seeing companies that would have been invisible to them a decade ago. But the majority of the industry still operates on warm-intro sourcing, which means the access problem persists.

The warm intro is dying →, and that is part of the fix. A fuller fix requires explicit investment by firms in global, data-driven sourcing, combined with hiring practices that reward finding under-indexed founders rather than punishing it.

The Information Asymmetry Problem

Public markets have spent 40 years building information infrastructure so that retail investors, institutional investors, and market makers all have access to roughly the same data in roughly the same timeframe. It is imperfect, but it exists.

Private markets have done almost none of this. The information advantages inside private markets are vast, persistent, and largely unaddressed.

A top-tier Sand Hill partner sees a specific funding round closing, knows the lead investor, knows the preferred customer references, knows the competitive dynamics of the sector, and can synthesize this into a thesis in hours. A comparable investor in a second-tier city gets a fraction of this information, weeks later, and has to make decisions with substantially worse inputs.

This asymmetry is not earned through better analysis. It is earned through network position. Over decades, it compounds into returns differences that have nothing to do with skill.

What Would Actually Fix This

Real intelligence infrastructure that operates in private markets the way Bloomberg operates in public markets. Not a database. Not a newsletter. A continuous, source-verified, real-time intelligence layer that is accessible to any serious investor regardless of their position in the social graph.

This is exactly what private markets need →. Public markets have Bloomberg. Private markets do not yet have an equivalent, which is why the information asymmetries persist.

Brevoir is one of several efforts building toward this layer. Whether Brevoir specifically succeeds is less important than the category succeeding. Private market intelligence infrastructure is the missing foundation, and the industry will compound faster once it exists.

The Incentive Problem

Inside most VC firms, the incentive structure rewards being right with consensus more than being right alone.

A partner who backs a company that everyone else backs, and it fails, gets excused: "we all thought this was a good deal." A partner who backs a company that no one else backs, and it fails, takes personal blame: "we told you that was too contrarian."

The asymmetry is real and it shapes behavior. Over time, partners gravitate toward deals that fit consensus shape, because consensus deals are career-safe regardless of outcome. The deals that might produce outsized returns (true contrarian bets) get underinvested in relative to their actual opportunity, because the career risk of backing them is too high.

The math of VC requires outliers. Outliers require conviction. Conviction requires willingness to be wrong alone. The current incentive structure inside most firms punishes the last of these, which undermines the whole chain.

What Would Actually Fix This

Smaller partnerships with clearer accountability per partner. In a three-partner firm, it is hard to hide behind consensus. In a 15-partner multi-stage behemoth, it is easy.

Also: better track-record measurement that rewards high-conviction bets relative to the broader fund context, rather than raw DPI across a diversified pool. This is a compensation and attribution problem, and it is largely unsolved inside most firms.

For LPs, the fix is backing managers who have demonstrated willingness to back non-consensus deals, rather than backing managers who have the cleanest cap tables of consensus deals. Most LPs claim to value this and do not actually underwrite it.

The Research Problem

Every other problem in VC is compounded by the fact that the research infrastructure inside most firms is broken.

Most VC research is useless →. Memos that get read once. Market maps that are stale before they are published. Associate-level work that is performative rather than insight-producing. Internal knowledge that decays faster than it compounds.

This matters because every other problem above is harder to fix when the foundational research layer is broken. You cannot solve information asymmetry with bad information. You cannot make contrarian bets with stale data. You cannot calibrate valuation comparables when your comps are months out of date.

What Would Actually Fix This

Continuous, source-verified, thesis-matched intelligence infrastructure as a first-class part of the firm's stack, not as an afterthought. The firms that have built this (or bought it) are operating on a different information substrate than the firms that have not.

This is also the most tractable of the broken things. The technology exists. The category exists. Firms can adopt modern intelligence infrastructure today. The ones that do are positioning themselves for the next decade. The ones that do not will continue producing decisions based on whatever their associates happened to read this morning.

The Conflict Problem

A smaller but real issue: conflicts of interest inside VC firms are increasingly common and increasingly underaddressed.

Multi-stage funds investing in competitors across portfolio segments. Partners sitting on boards of companies that compete in adjacent markets. Information flows between portfolio companies in the same fund that are uncomfortable even when legally permissible. Incentive-sharing structures that create unspoken preferences for certain outcomes.

Some of this is inevitable at scale. Not all of it. Firms that pretend these conflicts do not exist are storing up reputation damage for the future.

What Would Actually Fix This

Cleaner policies about portfolio overlap, more transparent communication with founders about conflicts, and an honest discussion inside firms about which conflicts are manageable and which should cause them to pass on deals regardless of quality.

Also: smaller, more focused funds. The conflicts inside a focused seed fund are manageable. The conflicts inside a multi-stage fund with $30B under management across every sector are not manageable, even in principle.

What Would a Fixed Industry Look Like

Imagine a version of venture capital where:

• LP reporting uses honest public market equivalents and realistic cash-on-cash, so LPs invest based on real performance.
• Intelligence infrastructure is universal, so information asymmetries based on geography and network are dramatically reduced.
• Smaller partnerships dominate, with clearer accountability and more contrarian conviction.
• Real research replaces performative memos, so decisions are grounded in current data rather than stale assumptions.
• Access to competitive rounds is based on thesis fit and value-add rather than warm-intro proximity.

This is not a utopia. It is a realistic description of what the industry could look like in 10 to 15 years if the structural shifts that are already beginning were accelerated rather than resisted.

It would produce better outcomes for founders (more efficient capital allocation), better outcomes for LPs (higher average returns and honest reporting), and better outcomes for the ecosystem (more capital flowing to underinvested geographies and underindexed problems).

It would produce worse outcomes for the specific set of people who currently benefit from the broken version: large multi-stage firms whose edge is largely network-based, LPs who depend on vague performance reporting to justify their allocations, and partners whose consensus-oriented careers are protected by current incentive structures.

That is why the fixes are hard. The people who could most easily implement them are the ones they would most directly disadvantage.

Where the Pressure Comes From

The industry does not reform from within. It reforms from competitive pressure.

Emerging managers with better sourcing, tighter partnerships, and clearer contrarian conviction are producing outsized returns and forcing larger firms to notice. LPs who take performance reporting seriously and invest based on real numbers are forcing managers to improve or lose capital. Founders with more information and more options are demanding better partners, shifting power away from the old gatekeepers.

The broken parts of VC persist because the incentives of the beneficiaries are strong. They are not permanent. The next decade of private market reform will be driven by competitive pressure from the edges, not by voluntary reform at the center.

That includes intelligence infrastructure as part of the pressure. When every serious investor can access continuous, real-time, source-verified market intelligence, the firms that rely on private information networks as their edge lose some of their advantage. Over time, merit-based sourcing and decision-making becomes more feasible than it has been in decades.

The Honest Conclusion

Venture capital is not broken in a way that makes it worthless. It is broken in a way that makes it much worse than it could be. The gap between the current state and a fixed version of the industry is measured in hundreds of billions of dollars of misallocated capital, thousands of underbacked founders, and dozens of missed category-defining companies per year.

Fixing it is not about throwing out the model. It is about upgrading the specific infrastructure and incentives that have stagnated for 20 years while the rest of the global economy modernized. The asset class deserves better. The founders deserve better. The LPs deserve better.

I am biased, obviously. I am building one piece of the infrastructure I think the industry needs. But I am building it because I genuinely think the current structure is leaving too much value on the table, and someone has to build the layer that makes the next version possible.

This is that layer. Brevoir → exists to solve the information asymmetry and research infrastructure parts of the broken-VC problem. Not because it fixes everything, but because the other fixes are much harder to implement when the information layer is broken. Fix the infrastructure, and the rest of the reforms become more achievable.

The industry is not going to fix itself. The people building the alternative, slowly and stubbornly, are going to fix it one piece at a time.