Interview, Fireside Chat
Gokul Rajaram on the 8 Moats Companies Need & Why Dropouts are "AI Maxing" the World
- Companies possessing four or more of eight defined modes (data, workflow, regulatory, distribution, ecosystem, network, physical infrastructure, scale) are expected to be "pretty damn secure," with specific entities like Amazon and TSMC leveraging low-cost replication barriers via scale.
- The outlook for physical infrastructure moats faces potential erosion from humanoid robots within "about a few years," while data portability trends may drive business-to-business switching costs toward zero over the next one or two years, diminishing brand relevance.
- Labor markets are predicted to shift from outsourcing to third-party BPOs, to non-replacement of leavers, and finally to layoffs as the primary response to AI efficiency gains.
- Vertical SaaS firms targeting BPO and human capital spend are expected to reach $10 billion valuations, though reaching $100 billion is deemed less likely for pure vertical players compared to horizontal platforms, while full-stack verticals (e.g., ServiceTitan with 32 products) face displacement risks despite higher defensibility.
- Pricing models are anticipated to evolve from seat-based access for products like ChatGPT Enterprise to outcome-based pricing for work-products where the user is no longer the constraint.
- AI-native businesses with strong leadership could potentially triple or triple-triple revenue to reach multi-billion dollar valuations by capturing service budgets, whereas legacy platforms like Salesforce and NetSuite must commoditize complements by offering free workflows to remain defensible against agentic AI.
- Future "durable" software companies must demonstrate data assets that improve with every interaction and deep workflow embedding, requiring a complete product re-frame and UX rebuild rather than simple feature additions.
- Product roadmaps must shorten as model capabilities improve every six months to prevent being "blown out" by subsequent iterations, while early-stage software firms focusing solely on MOIC without regard for IRR may fail to meet obligations.
- A wave of "zombie companies" with high valuations and moderate growth is expected to be sold to Private Equity or merged for scale, as PE struggles to digest assets and mid-level partners initiate spin-outs to rebuild deep entrepreneur relationships.
- Venture capital returns are projected to rely on doubling down on specific winners rather than maintaining diversified 30-company portfolios, with "propertied founder access" described as a myth requiring verification through direct founder interaction.
- Fund strategies face pressure to mix incubation, seed, and Series A bets because Series A-only deals at $300–$500 million valuations fail to provide sufficient ownership percentages for returns.
- Investment focus is shifting toward early-stage "AI maxed" young founders and dropouts, with "king making" in VC valid only if execution matches vision, while non-consumption markets creating new behaviors represent the largest sizing opportunities.
- Market dynamics in the first few years of a company's life will prioritize durability and defensibility over margins, with the best returns often found in early high-conviction deals rather than expensive later-round (Series B+) investments.
- The next decade is characterized by ambitious entrepreneurs tackling significant societal challenges, moving beyond small apps, while the speaker anticipates investing in more dropouts than in the previous 15 years.