Mar 16, 2026, 11:19 AM ETInterview, Fireside Chat

Gokul Rajaram on the 8 Moats Companies Need & Why Dropouts are "AI Maxing" the World

Rajaram proposes an "Eight Modes" framework to score the durability of software companies, where a score of four or more indicates a secure, defensible business.
Mode 1: Data Mode: Requires proprietary data that is difficult for competitors to replicate (e.g., Spotify's decade-long listening behavior).
Mode 2: Workflow Mode: Depth of embedding in company operations determines strength; NetSuite (ERP) is a "1.0" workflow, while Zendesk is a "0.5".
Mode 3: Regulatory Mode: Moats built on licenses, capital requirements, or multi-year procurement contracts (e.g., Coinbase's money transmission licenses).
Mode 4: Distribution Mode: Proprietary, exclusive distribution channels that are hard to displace (e.g., Intuit's network of CPAs trained exclusively on QuickBooks).
Mode 5: Ecosystem Mode: Reliance on third-party developers and partners (e.g., Shopify's hundreds of thousands of integrated apps).
Mode 6: Network Mode: Structural liquidity and density effects that AI cannot "wipe code" (e.g., DoorDash's restaurant/runner/reputation density).
Mode 7: Physical Infrastructure: Moats based on "atoms" and physical assets, which remain hard to displace despite robotics advancements.
Mode 8: Scale Mode: Cost advantages derived from massive scale that prevent replication (e.g., Amazon, TSMC).

Market Overreaction: Rajaram characterizes the current software market correction as a "100% overreaction" where all software is being unfairly painted with the same brush as AI commoditizes basic coding.
Brand Irrelevance: He argues brand is no longer a primary moat in B2B as switching costs approach zero due to improved data portability and AI cloning capabilities.
Data Portability Risk: AI agents will make migrating data between ecosystems easy, eroding traditional lock-in for systems of record like Salesforce.
Salesforce Strategy: He suggests Salesforce must commoditize its complements (agentic workflows) and decide whether to monetize data or workflows, potentially offering free workflows to sell data.
Multi-Product Requirement: Single-product companies struggle to reach $10B+ valuations; companies must own the full stack with adjacent products (e.g., Square Capital) to drive retention.
Retention vs. Profit: Not every product in a portfolio needs to be profitable; some serve as retention drivers while others generate the profit pool.

Margins: Early-stage gross margins are less critical than defensibility; margins are expected to improve as inference costs drop and pricing power increases.
Pricing Models: "Seat pricing" will persist for access products (e.g., ChatGPT Enterprise) but will shift to "outcome-based" or consumption pricing for "work products" where user count is not the constraint.
BPO Spend Shift: AI investment is targeting Business Process Outsourcing (BPO) budgets first, followed by non-replacement of departing staff, and finally direct layoffs.
Vertical SaaS: Vertical software can build $10B companies by taking over human labor costs (BPO and capital), not just software budgets.
Growth Expectations: Rapid "1-to-10" growth is becoming common; investors should prioritize Net Revenue Retention (NRR) and Gross Retention over top-line velocity.

Market Sizing: Focus on "non-consumption" markets where products create entirely new behaviors or demand (e.g., Shopify, Granola, Gamma).
Founder Archetypes: Distinguishes between repeat founders (e.g., security sector), first-time extraordinary founders (e.g., consumer internet), and AI lab researchers (e.g., Anthropic).
Mega Fund Dynamics: Mega funds are using Series A checks as "optionality" to index all potential winners and double down later, potentially cannibalizing traditional Series A returns.
Capital Allocation: Funds should balance seed bets with "incubation" bets on known super-founders to maximize ownership and control.
Liquidity: Investors should prioritize IRR over MOIC; selling parts of winners before IPO or via secondary markets is necessary to maintain fund IRR if the asset's future outlook slows.
Remote Work Regret: Rajaram reversed his view on remote work, now believing in-person presence (3-4 days/week) is critical for early-stage founder alignment and iteration speed.

Top Misses: Rajamar identifies Quince (dismissed due to D2C trend bias despite 40% retention) and Facebook (underestimated its trajectory to a $1T company) as his biggest misses.
Highest Multiple: His Figma angel investment returned 500x to 1,000x at IPO, though he notes the multiple has compressed since.
Biggest "Miss" (Visionary Scale): Admitting he failed to predict that Google and Facebook would become multi-trillion dollar companies, a failure of scale prediction rather than investing logic.
Vanta Miss: Regrets passing on Vanta because he had already committed capital in the space, missing a "N of one" category winner.
Advice to Young Founders: Recommends 2–3 years of work experience at a good company before launching, as the network and operational knowledge are invaluable.
Future Optimism: Expresses high excitement for ambitious entrepreneurs tackling "hard problems" (e.g., energy, longevity, space) rather than small incremental apps.

"You cannot be a single product company. Vertical products, you've got to really own full stack. It's harder otherwise to be a 10 plus billion dollar company."
"If you have a product that is compelling enough... you don't look at margins. You look to see whether or not they have a product that is compelling enough."
"The best companies have a remarkable product at their core... I look for what the remarkability is in the core product or value proposition of the company."
"If they don't build agentic workflows and commoditize a complement by figuring out where the profit pool is, I think they have to figure out, is the profit pool in the data or the workflows?"
"You've got to have a sunk cost. You've got to abandon sunk cost fallacy... I'm going to be ruthless about migrating the customers from the current business to the new product."