| Episode | Status |
|---|---|
| Episode | Status |
|---|---|
David George is a General Partner at Andreessen Horowitz, where he leads the firm's Growth investing team. His team has backed many of the defining companies of this era, including Databricks, Figma, ...
David George, General Partner at a16z's Growth fund, defends mega-fund economics while revealing the firm's investment philosophy. He argues that large funds can generate 5x returns, citing their best-performing $1B fund with Databricks returning 7x and Coinbase 5x. George discusses the private market expansion creating $5T in market cap, the shift from public to private value creation, and a16z's strategy of investing in founder strengths over lack of weaknesses. He shares insights on AI company margins, competitive dynamics, and controversial bets like Flow and Waymo.
George counters the notion that large funds can't generate high multiples, revealing a16z's best-performing fund is actually $1B with Databricks returning 7x and Coinbase 5x already. He explains how private markets have grown 10x to $5T, with 53% of value creation happening post-Series C, and how tech waves create massive opportunities that now stay private longer.
George argues private markets still offer cheaper capital than public markets despite high valuations, contrasting with conventional wisdom. He identifies avoiding stock price volatility and employee management issues as the biggest advantage of staying private, using Stripe, SpaceX, and Databricks as examples of companies benefiting from controlled private valuations.
George advises institutional investors to recognize that private markets now represent what used to be public small-cap, with 8 of top 10 companies being US West Coast tech that were venture-backed. He argues top venture funds outperform top PE funds and that AI implementation will make this gap more extreme, recommending increased allocation to venture.
George discusses backing European founders like Matty at 11 Labs and Alex at Deal, highlighting Alex's relentless sales culture. He reveals a16z's 'fix the mistake fund' strategy where growth team catches early-stage misses, with 50% of investments being venture follow-ons, 15% growth follow-ons, and 33% net new companies with pre-existing relationships.
George reveals a16z's core investment rule from Ben Horowitz: always invest in spiking founder strengths rather than lack of weaknesses. The biggest mistake pattern is overweighting fear of theoretical future competition (the 'isn't Google going to do this?' trap), which can talk you out of any investment. He also notes consistently underestimating market sizes.
George ranks disruption factors for AI attacking SaaS incumbents: business model shift (most disruptive), UI/workflow changes (second), and data access (third). He uses Decagon in customer service as example of business model shift from seats to task completion. Argues human labor to technology budget transition must be product-driven and pulled by market, citing C.H. Robinson's 40% productivity increase.
George explains revenue means the same if retention and engagement are high, but the bar for assessing AI companies is way higher due to fast scaling. Companies like Gamma, 11 Labs, ChatGPT, and Harvey show 'magic' organic customer acquisition. He emphasizes ROIC (return on invested capital) as the #1 company measure, with momentum being relative to peer set growth rates.
George distinguishes between helpful capital deployment and failed 'capital as weapon' strategy. SoftBank Vision Fund's mistake was believing capital alone could create winners through adverse selection - companies opting for capital-as-weapon strategy lacked inherent competitive advantages. He argues a16z's brand and capital help already-winning companies via preferential attachment, not king-making.
George defends investing in crowded customer support AI space, explaining it's already better/faster/cheaper with current models. Notes 50% of SaaS markets are winner-take-most, 50% are distributed (like payroll). Whether Decagon wins outright or market distributes, the growth and market pull are 'staggering' with extremely high EBC conversion rates.
George addresses AI margin concerns, predicting rationalization similar to cloud infrastructure with model companies becoming oligopolistic like AWS/Azure/GCP. Currently muddy due to token cost decreases offset by reasoning usage increases. Companies with SaaS-level margins in AI raise red flags - likely means customers aren't using AI features. Defends high valuations for 3x faster growth than predecessor SaaS.
George reveals constant reassessment of OpenAI entry price, noting their 2019 Databricks investment at $6B never predicted current scale. Uses Google/Facebook example: 10 years ago monetizing users at 1/7th of today. Emphasizes investing in companies where core market can be bigger than expected (Stripe, SpaceX Starlink, Waymo) and founders have advantage in finding next products (Anduril).
George shares his biggest disagreement with Marc and Ben over Waymo's 2020 investment as only VC investor. Despite his analysis showing high valuation, Marc and Ben argued it's 'endless market size' and they're market leader. Started smaller, maintained relationship, wrote much larger check in recent round. Product now 7-10x safer than human drivers per medical professional analysis.
George defends controversial Flow investment, arguing Adam Neumann has extraordinary strengths in brand building, company building, product, and hiring - the most important ingredients for early-stage companies. Insight: renters spend 30% of disposable income on only unbranded experience in their lives. Thesis is bringing brand premium to rental housing with Adam's unique real estate + brand intersection skills.
George's biggest mind change: initially thought models would eat everything 18-24 months ago, now believes application companies will thrive in every direction. Uses radiology example: AI better than humans at scans pre-current wave, yet radiologist numbers increased because they spend 60-70% of time on other tasks. Biggest error of omission: not investing in Anthropic, which he sees winning B2B while OpenAI dominates consumer.
George identifies Chris Dixon as clearest articulator of early-stage strategy at a16z. Distinguishes Marc's superpower as seeing the future with accurate 10-year predictions vs Ben's as best management coach for executive dynamics. Most excited about personal health management (proactive AI coaching on health decisions) and robotics as largest AI category, though form factors still being determined.
20VC: a16z's David George on How $BN Funds Can 5×, Do Margins & Revenue Matter in AI & the Most Controversial Bet at a16z
Ask me anything about this podcast episode...
Try asking: