Ben Horowitz on Raising a New Fund and How Venture Firms Scale

Ben describes his unique 30-year working relationship with Marc Andreessen using a Quincy Jones/Michael Jackson analogy. He explains how their complementary skills work together, with Marc as the visionary talent generator and Ben as the operator who maximizes that talent through structure and execution.

• •Marc generates more ideas while Ben focuses on editing and execution, creating a productive tension
• •Ben is more decisive as a personality type, helping commit resources and set direction
• •The relationship works because they're similar enough to collaborate but different enough to be complementary
• •They argue frequently on firm direction and timing, but work through ideas together before implementation

Ben explains why managing a venture firm is fundamentally different from running an operating company. The key challenge is managing brilliant, disagreeable investors who can destroy each other's work through conflicts. He emphasizes the critical importance of organizational design to minimize conflict rather than relying on process or authority.

• •Great VCs are typically high-IQ and disagreeable by nature - wanting to be liked is a liability
• •Heat-seeking investors (who chase hot deals) tend to be more agreeable but less successful long-term
• •Organizational design must solve conflicts structurally - you can't tolerate cross-functional dependencies like in operating companies
• •Conflicts in VC firms are 'kimchi problems' - they get worse over time if buried, requiring immediate resolution
• •Each fund is designed like a small VC with no more than 5 GPs to maintain cohesion

Ben contrasts a16z's mission-oriented approach with 'heat seeking' investment strategies. Rather than chasing whatever's hot or optimizing purely for returns, a16z focuses on helping the best entrepreneurs build companies that strengthen America technologically, requiring belief in both the technology and the entrepreneur.

• •The firm's mission is to help the best entrepreneurs build the best companies and make the West technologically strong
• •They evaluate entrepreneurs on magnitude of strengths (world-class abilities) rather than ruling out deals based on weaknesses
• •Heat seekers show up during booms and disappear after - they care about the chase, not breakthrough technology
• •This mission-driven approach means they can't ignore important sectors like crypto or AI, even if concentrated strategies might be more financially optimal
• •Every good investor can find what's wrong with anyone - the key is focusing on what they can do exceptionally well

Ben explains the core thesis behind building a large venture firm, rooted in Marc's 2011 'Software is Eating the World' piece. The conventional wisdom was only 15 companies per year would reach $100M in revenue, but they bet it would be 150-200, requiring a bigger firm to address the market.

• •In 2011, conventional wisdom said only 15 companies per year would reach $100M revenue
• •Marc's bet that 'software eats the world' meant it would be 150-200 companies, which proved correct
• •You need a bigger firm to get into that many great companies - five people can't address the market
• •The more important rationale: entrepreneurs need capabilities beyond smart advice - networks, customer access, government relations, recruiting
• •A platform with real capabilities is a better product than 'I'm smart and will take you to coffee for two months'

Ben discusses which platform services provide real value versus those that don't. The key insight is that general services across domains don't work well, but specialized services within specific sectors (like AI model evaluation or crypto research) provide significant acceleration for portfolio companies.

• •General research ideas for the entire startup community didn't work well
• •Sector-specific services work extremely well - like evaluating AI models and posting results for portfolio companies
• •Talent recruiting is most valuable for 'seed corn' - the first 3-5 stellar hires with strong networks
• •Companies must ultimately build their own recruiting muscle or they'll never be good companies
• •Cross-domain services ended up being less interesting; specialization within sectors is key

Ben makes a strong case for why boards matter, both for legal protection and company success. He debunks the idea that founders should avoid boards, explaining that boards provide critical fiduciary protection and create valuable rhythm for companies. Y Combinator's data showed companies with boards significantly outperformed those without.

• •Running without a board after selling equity is 'the most dangerous fucking idea' - you lose legal protection from personal liability
• •Board approval for material decisions (like equity grants) provides complete legal protection; without it, you're vulnerable to lawsuits
• •Y Combinator found companies with boards significantly outperformed those without (though with confounding factors)
• •The rhythm of reporting to outsiders every 2-3 months creates valuable internal pressure and organization
• •Board members can be existentially valuable at key moments (like preventing a premature $4B sale of Databricks, now worth $100B+)
• •Platform support allows board members to scale beyond the traditional 8-board limit

Ben reveals a counterintuitive insight: the ability to win deals is a much bigger driver of returns than picking ability. Being able to win automatically gets you to top-tier returns, while picking moves you up within that tier. This creates a self-reinforcing cycle where the best pickers want to join firms that can win.

• •Winning deals is a much bigger percentage of getting good returns than VCs like to admit
• •You could be a roughly average picker but a great winner and outperform the inverse
• •Being able to win automatically gets you to top-tier returns; picking moves you up that list
• •This is self-fulfilling: the best investors want to join firms where they can actually invest in their best ideas
• •If you can't win the deals you want, being a great picker is extremely frustrating and ultimately pointless

Ben explains how a16z's media strategy has evolved from the traditional press model to direct channels. The fundamental physics changed from limited channels and formats to unlimited options, and from company brands to personal brands. This required building an entirely new marketing system under Eric Torenberg.

• •When a16z started, the primary information channel was traditional press with very different 'laws of physics'
• •Old world: fixed channels, fixed formats, company brands, focus on what NOT to say
• •New world: unlimited channels, unlimited formats, personal brands (Jensen vs NVIDIA, Karp vs Palantir)
• •Podcasts work best now because people can consume while exercising and content targets specific understanding levels
• •The answer to a gaffe is 'flood the zone' not 'don't make the gaffe' - never apologize, just create more content
• •Other VCs are still mostly in the old world because they copied a16z late and don't understand the underlying physics

Ben outlines why most venture firms can't scale like a16z has. The two fundamental barriers are shared control structures (which prevent effective reorganization) and lack of operational leadership. Most firms have shared economics and shared control, making it impossible to redistribute power through necessary reorganizations as the firm grows.

• •Most VC firms have shared control structures that prevent effective reorganization as they scale
• •When you reorganize, you redistribute power - people who had power lose it, others gain it
• •If reorganization is democratic/consensus-based, it will fail because everyone does local optimization except the CEO
• •You need a leader with operational caliber to make tough decisions, which is rare in VC
• •The main limit for a16z is market size - how many great entrepreneurs exist to fund, not internal constraints
• •Investing teams must stay small (like little VCs) to maintain effective conversation about finding truth
• •a16z's advantage: centralized control for reorganization + operational leadership + platform that makes small teams effective