Interview, Fireside Chat
Mamoon Hamid: AI - Where Value Accrues, Startups vs Incumbents & Scaling Laws | E1217
- Market Creation Strategy: Mamoun emphasizes a preference for companies that create new markets rather than competing in existing ones, citing Slack (created a market), Figma (collaborative design), and Glean (enterprise search) as ideal examples where the investor can "create the playing field, play on it, and win."
- Targeting High-Value Scarcity: Investment focus is directed at software and AI solutions that supercharge the top 20 highest-paying US jobs—specifically doctors, lawyers, and developers—addressing the scarcity of highly skilled talent in these sectors.
- Specific Portfolio Examples: The firm has backed co-pilot tools including Harvey (legal), Ambience (medical), and Codium (developer tools) to assist these specific high-value professions.
- Differentiation via Technical Depth: In crowded AI application spaces (e.g., 10+ medical transcribers), Mamoun asserts success relies on "technical depth" achieving near-perfect output accuracy (e.g., 99% vs. 87%), requiring founders with deep ML expertise paired with domain experts.
- Incumbent vs. Startup Dynamic: While incumbents (Google, Microsoft, Amazon, Meta, Oracle) are spending hundreds of billions on frontier models, Mamoun believes the opportunity for venture capital lies in the application layer, where trillions of value will be created over the next decade.
- Pricing Strategy Exception: While the firm generally targets owning 15–20% of early-stage companies at $5–10M checks, they maintain a "YOLO bucket" for exceptional founders, willing to invest at much higher valuations (e.g., $750M pre-product) if conviction is extreme, though this is treated as a one-in-20 exception.
- Labor vs. Software Pricing: AI companies are shifting from seat-based pricing ($30–$40/month) to labor-based pricing ($300–$500/month) because they provide capability to do 10x the work, enabling faster revenue scaling from zero to $4–5M ARR.
- Build vs. Buy Decision: Mamoun doubts the trend of companies like Klarna building custom internal AI tools to replace SaaS (e.g., Salesforce, Workday), arguing that specialized AI vendors will likely offer better value for outcomes (e.g., per ticket resolved) than in-house development.
- Investment Thesis on Middleware: He views the "middle layer" between foundation models and applications (e.g., vector databases, fine-tuning tools) as potentially over-invested and fleeting, given the rapid change and risk of being subsumed by foundational model providers.
- Foundation Model Business Viability: Mamoun notes that while GPU and infrastructure providers (e.g., NVIDIA) are highly profitable, the "selling tokens" business for LLMs is currently not a great business due to price dumping, though margins may improve as models get more efficient.
- Token Price Trajectory: Token prices have dropped 200x in the last 18 months; Mamoun predicts a further 10x–20x decline, while anticipating 10x–20x better models, creating a massive opportunity for technology adoption.
- Revenue vs. CapEx Gap: Addressing the "$600 billion AI question," Mamoun argues that while CapEx is high ($200–600B), the potential revenue is justified by technology's shift from 15% to 20% of global GDP, driven by addressing labor shortages rather than just software sales.
- Kleiner Perkins Positioning: The firm characterizes itself as a "boutique" with a small team (7 investors) managing both an $800M early-stage fund and a $1.2B growth fund, preferring a craft-focused approach over scaling through headcount.
- Reserve Management Tactics: The firm allocates roughly 60% of capital in the initial check and reserves 40% for follow-ons, though they emphasize that "pay-to-play" dynamics often force participation in subsequent rounds to avoid dilution.
- Box Investment History: Kleiner Perkins provided multiple bridge rounds for Box during the 2008–2009 financial crisis at a $25M valuation, accepting significant founder dilution to navigate the market dislocation, eventually yielding an IPO.
- M&A Market Status: The M&A market is described as "slow" and "gun-shy" due to regulatory fears (not solely the FTC) and high valuations, leading strategic acquirers to prefer buying smaller private companies rather than large public ones.
- IPO Outlook: Mamoun anticipates a reopening of the IPO market in the coming year, contingent on a "bellwether" large tech company (e.g., Databricks, Stripe, Starlink) successfully going public.
- Figma Investment Rationale: The decision to invest in Figma (at ~$100M post) was driven by the realization that the product finally achieved low-latency multiplayer functionality in the browser (around 2017), evidenced by near-daily usage metrics despite low revenue.
- Founder Archetypes: The firm prefers two founder types: (1) First-time founders who are hyper-obsessed with building a product in a non-obvious market (e.g., Figma's Dylan Field, Intercom's Owen Mahoney), and (2) Repeat founders looking to surpass previous successes (e.g., Slack's Stewart Butterfield, Rippling's Parker Conrad).
- Valuation Philosophy: Mamoun disagrees with the strategy of founders raising maximum capital at the highest possible valuation at early stages, arguing that "fair" pricing preserves flexibility and prevents future down-round dilution; he cites raising at $35M post (vs. $200M) as a strategic choice for long-term partnership.
- Investment Decision Structure: The firm operates without a voting structure for early-stage deals; decisions are made based on individual partner conviction and discussion, allowing non-consensus bets to proceed.
- Learning from Losses: A $30M loss in a consumer lending startup called Tally occurred when interest rates rose, teaching the firm that consumer lending is inherently difficult and highlighting the risk of over-leveraging in volatile interest rate environments.
- Board Governance: The best board meetings focus on deep-diving into only one or two critical issues rather than high-level status updates across many topics, allowing the board to meaningfully influence the company's trajectory.
- Personal Beliefs: Mamoun notes that faith heavily influences his work ethic, emphasizing humility, empathy, and care, while admitting he is a "dreamer" who sometimes holds belief in underperforming founders for too long.
- Quick Fire Insights:
- Disbelief: Most people underestimate the difficulty of venture capital, assuming it is glamorous rather than a grind.
- Respect: He cites Matt Kohler (Benchmark) as a key investor he respects outside his own firm.
- Ideal CEO Role: If he could be CEO of any company for a day, he would choose OpenAI to understand the progress toward AGI.
- Global Concern: He identifies geopolitics and polarization as his primary concern for the world.
- Career Reflection: The hardest lesson learned over 19 years was simply that "venture is a grind."