Focus on investing in software companies that create "hostages, not customers" by becoming deeply embedded in a business's operations. Vertical SaaS companies like Toast (TOST) are a prime example, serving as the indispensable operating system for the restaurant industry. Similarly, established "systems of record" like Workday (WDAY) are attractive due to their high switching costs, making them more resilient to AI disruption. The key is to find businesses whose products are so integrated into customer workflows that leaving is prohibitively expensive and risky. Be cautious of AI application companies that are merely "thin wrappers" over foundational models, as they lack a durable competitive advantage.

Stripe (Private)

• The podcast revisits the early days of Stripe, when it was called Dev Payments. Alex Rampell admits to passing on the seed round because he couldn't understand where the customers would come from, as most businesses used established players like Chase Payment.
• Stripe's founders, Patrick and John Collison, had a key insight: their customers did not exist yet. They were betting on a wave of new internet companies that would be created and would choose the best, most developer-friendly product, rather than trying to steal customers from incumbents.
• This strategy is described as "Greenfield Bingo": building a better version of a software category and selling it to the wave of new companies being created, who are not "hostage" to an existing provider.
• The founders were also noted for having deeply studied the history of the payments space, a trait Alex Rampell looks for in great entrepreneurs.

Takeaways

• Look for companies that aren't just competing for existing market share but are enabling and capturing a brand-new market of customers. These are often startups selling to other startups.
• The "Greenfield" strategy works best in markets with a high rate of new company creation (e.g., e-commerce, tech startups) and is less effective in markets where new entrants are rare (e.g., hospitals).
• Investing in a "Greenfield" opportunity is a bet on the future and the creation of new businesses. If the rate of new company creation slows, this strategy becomes less effective.

Plaid (Private)

• Alex Rampell led deals in Plaid and discusses it as an example of correcting a past mistake. He debated investing in the Series B at a $130M vs $135M valuation and ultimately passed.
• He later admitted he was "stupid" and invested in the Series C at a $2.4 billion valuation, demonstrating the importance of being willing to pay up for a clear winner, even if you missed the earlier, cheaper round.
• The key takeaway from this anecdote is that in venture, and by extension growth investing, it's better to be "rich than right." Stubbornly sticking to a past decision (like passing on a company) can be more costly than admitting you were wrong and buying in at a higher price.
• The failed acquisition of Plaid by Visa is also mentioned, highlighting the regulatory risks that can come with being a highly successful and strategic company in the fintech space.

Takeaways

• When a company establishes itself as a clear market leader, don't let a past decision to "pass" prevent you from investing. The opportunity cost of missing out on a dominant winner can be far greater than the pain of paying a higher valuation.
• For private companies that become market leaders, follow-on rounds will be expensive. However, the potential upside can still justify the high entry price if the company continues to dominate its category.
• Be aware of regulatory risk for dominant companies in critical sectors like finance. As they scale, they attract scrutiny that can block potentially lucrative exits (like an acquisition by a major incumbent).

Toast (TOST)

• Toast, the restaurant point-of-sale (POS) company, is highlighted as a prime example of a successful Vertical SaaS company.
• The challenge for a company like Toast is immense at the beginning. They have to convince their first customer—a restaurant—to bet their business on a brand-new startup with no other customers and limited cash.
• The ability to overcome this and "materialize customers" is a sign of an exceptional founder and business model.
• The discussion implies that while hard to start, Vertical SaaS businesses that become the "operating system" for an industry can be incredibly sticky and valuable.

Takeaways

• Vertical SaaS—software built for a specific industry—can be a very powerful investment theme. These companies solve unique, industry-specific problems that general software cannot.
• Look for companies that are becoming the system of record or operating system for their niche. Once a business runs its entire operation on a platform like Toast, switching costs become extremely high, creating "hostages, not customers."
• Early-stage traction in these markets is difficult but is a strong positive signal. A company that can sign up its first few customers in a skeptical industry is demonstrating significant product-market fit.

AI: Infrastructure vs. Application Layer

• The podcast breaks down the AI landscape into two main layers:
  • Infrastructure Layer: The foundational model providers like OpenAI and Anthropic. They are powerful but face the risk of customer "promiscuity," as businesses can switch between models relatively easily to get the best performance or price.
  • Application Layer: Companies that build products on top of the infrastructure layer (e.g., an AI-powered customer service tool). They face immense competition, as the barrier to creating a "thin wrapper" around an API is low.
• The speaker notes that the speed of innovation is accelerating dramatically. A market-leading product could be unseated in weeks, not years, because AI tools make it so easy to replicate and improve upon existing software.
• A company like Zendesk (customer support) is at risk because AI can automate its core function, potentially reducing the need for licenses. In contrast, a company like Workday (HR system) is more insulated because it's a "system of record" with high switching costs.

Takeaways

• When investing in AI, it's crucial to determine where the company sits in the stack and how it plans to build a durable advantage or "moat."
• Application-layer companies are a risky bet unless they can create stickiness. The key question to ask is: "How do you get hostages?" They must move beyond being a simple wrapper and become a system of record by integrating deeply into customer workflows or by owning unique, proprietary data.
• Infrastructure players are in a powerful position but may compete heavily on price and performance, potentially squeezing margins. The ultimate winners may be those who can specialize in a specific, high-value domain (e.g., coding, legal analysis).

Investment Theme: "Hostages, not Customers"

• This is a core investment framework mentioned throughout the podcast. The best businesses are those where it is incredibly difficult or painful for a customer to leave.
• This "stickiness" is not about having a bad product, but about being deeply embedded in a customer's operations.
• Examples include:
  • Workday (WDAY) or NetSuite: Once a large company runs its HR or finances on these platforms, ripping them out is a massive, expensive, and risky undertaking.
  • Vertical SaaS companies like Toast (TOST) that become the central operating system for a business.
• The opposite is a business where customers are "promiscuous" and can easily switch, such as with the foundational AI models.

Takeaways

• When analyzing a company, especially in software, ask: How high are the switching costs?
• Look for businesses that are a system of record. These are platforms that hold a customer's essential, mission-critical data (e.g., financial records, employee data, customer lists).
• A strong indicator of a "hostage" situation is when a company's product is integrated with many other systems and workflows within the customer's organization. This creates a web of dependencies that is hard to untangle.

Investment Theme: Software Displacing Labor

• One of the three main investment theses discussed is investing in software that does the job of labor. These are often the fastest-growing companies because the return on investment for the customer is immediate and obvious.
• The example given is Eve, a company selling to plaintiff attorneys. The software automates the work for small-ticket cases that attorneys would otherwise ignore because the manual labor isn't worth the small payout. The software turns unprofitable cases into profitable ones.
• These companies can "hyperscale" because they are not just making a process 10% more efficient; they are fundamentally changing the economics of a job by replacing an $80,000/year salary with a $20,000/year software license.
• The primary risk for these companies is that they may not be "sticky." If they are just a feature (e.g., an AI agent making outbound calls), they can be easily replaced by a competitor.

Takeaways

• Look for companies whose value proposition is replacing or augmenting expensive human labor with software. This creates a clear and compelling ROI for customers.
• The fastest growth often comes from selling into markets that previously used zero software for a specific task (e.g., relying on Microsoft Office and manual processes).
• Critically assess the company's long-term strategy. To be a durable investment, the company must have a plan to evolve from a labor-displacing "feature" into a sticky system of record that creates "hostages." Ask: "What happens when a competitor shows up and offers to do it 50% cheaper?"