TIP841: Palantir – Palantir is Cheaper than I Thought! w/ Daniel Mahncke & Shawn O’Malley

Episode Summary

Executive Summary: The episode analyzes Palantir as a divisive but increasingly compelling business, explaining its ontology-driven software, defense roots, and AI-powered acceleration. The hosts argue that improved onboarding via AIP, strong net dollar retention, and rising margins make the company fundamentally stronger than its reputation suggests, but they still stop short of buying due to valuation, dilution, political risk, and uncertainty about long-term defensibility.

Main Topics: What Palantir Does: Ontology and Digital Twins: The hosts explain Palantir’s core product as a software layer that unifies fragmented enterprise data into an actionable ontology, connecting objects, relationships, permissions, and workflows across systems. Company History and Defense Origins: Palantir was founded after PayPal-era fraud work and 9/11-era intelligence fragmentation concerns, with Peter Thiel and Alex Karp shaping a mission tied to national security and intelligence. AI-Driven Growth Inflection: AIP and customer boot camps reduced onboarding friction, accelerated revenue growth, and increased monetization of existing customers, reversing a prior slowdown. Competitive Moat and Switching Costs: The hosts debate whether Palantir’s ontology, forward-deployed engineers, and first-mover advantage create durable switching costs that rivals like Microsoft, ServiceNow, OpenAI, or Anthropic may struggle to replicate. Margins, Rule of 40, and Valuation: Palantir’s combination of rapid growth and expanding margins makes it look expensive on trailing multiples but potentially reasonable on forward operating profit if growth sustains. Politics, Reputation, and International Risk: Karp and Thiel’s controversial reputations, plus privacy and surveillance concerns, may limit adoption abroad and create legal/regulatory risk, especially in Europe. Capital Allocation and Shareholder Returns: The company holds a large cash pile and no debt, but has no buybacks or dividends and still dilutes shareholders via stock-based compensation.

Key Arguments: Palantir is not a traditional database; its ontology maps entities, relationships, actions, and permissions so organizations can make real-time decisions. The company’s origin in intelligence and fraud detection explains why it is especially good at solving fragmented-data problems across large organizations. AIP changed the business by making demos, onboarding, and customer education much easier, which helped unlock faster growth. Revenue growth slowed to roughly 12% before AI but later re-accelerated dramatically, showing the product-market fit was already there but operationally constrained. Net dollar retention rising toward 160% suggests existing customers are expanding meaningfully, not just adding new customers. The company’s moat comes from switching costs, complex implementation, and first-mover advantage, not from owning a proprietary LLM. Microsoft, Google, ServiceNow, OpenAI, and Anthropic may have some overlapping tools, but the hosts think Palantir’s hands-on deployment model and ontology layer are harder to replicate. Palantir’s valuation looks high on sales, but high margins mean the implied valuation on operating profit is less extreme than it first appears. Despite the business strength, the hosts remain cautious because the outcome depends heavily on sustaining exceptional growth, which is hard to underwrite with confidence. Political reputation and data-sovereignty concerns may be materially limiting Palantir’s international growth versus the U.S.

Data Points: Company founding year: 2003 - Palantir was founded by Peter Thiel, Alex Karp, Joe Lonsdale, Stephen Cohen, and Nathan Gettings. Initial Thiel funding: $30 million - Peter Thiel bankrolled the early company. In-Q-Tel investment: $2 million - CIA venture arm backed Palantir early, mainly for access and credibility. Largest individual ownership stake: ~4% - Peter Thiel’s approximate ownership stake in Palantir. Alex Karp ownership stake: ~2.5% - Karp’s personal stake was cited as significant skin in the game. Customer count: ~1,000 customers - Palantir focuses on large accounts rather than broad small-business distribution. Employee count: ~4,500 employees - Used to contrast with Accenture’s scale and Palantir’s lean structure. Accenture employee count: ~800,000 employees - Mentioned in relation to outsourcing forward-deployed work. Cash on balance sheet: More than $9 billion - Palantir has a major cash position and no debt. Debt: None - No long-term or short-term debt on the balance sheet. Revenue growth trough: As low as 12% YoY - Revenue growth slowed materially from 2019 through mid-2023. Recent revenue growth: ~90% YoY - Cited as current top-line growth in the rule-of-40 discussion. Net profit margin in 2023: ~3% to 4% - Palantir was barely breaking even before the recent inflection. Current net profit margin: ~60% - Shows major operating leverage and margin expansion. Rule of 40 score: 155% - Derived from about 90% growth plus about 65% profit margin. Net dollar retention: From ~100% to almost 160% - Existing customers are spending about $1.60 for every $1 spent the prior year. Quarterly million-dollar deals: 220 deals - Deals closed at $1 million or more in the referenced quarter. Deals of at least $5 million: Close to 100 - Shows emphasis on large enterprise contracts. Deals of at least $10 million: More than 70 - Shows unusually large average deal sizes. Stock-based compensation: 13% of revenue - Ongoing dilution remains a shareholder headwind. Prior stock-based compensation peak: ~20% of revenue - Two years earlier, dilution was substantially worse. U.S. commercial growth outlook: ~150% implied - Karp said growth could match or exceed U.S. commercial growth for 18 months. Potential 2027 revenue under Karp scenario: $17B–$18B - Projected if high growth persists through 2027. Current price-to-sales multiple: ~60x trailing sales - Used to illustrate the market’s lofty expectations. Model fair value under analyst case: $90 - Based on lower growth assumptions, 60% margin, 30x exit multiple, and 10% discount rate. Model fair value under Karp case: $240 - Based on stronger growth assumptions and similar margin profile. Stock move after earnings: +40% - The hosts note the stock rose sharply after earnings, reducing the apparent margin of safety. Stock appreciation since 2023 lows: +2,400% - Used to explain why many investors may tolerate blind spots in the thesis.

Pivotal Quotes: "Palantir is not a value play. But when I started my research, I did it thinking in the back of my mind at least, I might need to abort and look for another company because there's just no sense in pitching it. But that wasn't the case." — Host: Framing the episode’s surprise that Palantir may be more investable than expected. "The magic of Palantir really comes from the context that the ontology creates." — Host: Explaining why Palantir’s value is in its data relationships and execution layer, not just the LLM. "I think the entire story here comes down to: does it make sense for me and do I fully understand the technology behind it? And the honest answer is no." — Host: Why the hosts ultimately decline to invest despite liking the business.

Implications: Palantir looks more like a high-quality compounder than a meme stock, but its valuation, dilution, and political/reputational risks keep it out of the hosts’ buy zone. If growth persists, the market could still reward it heavily, especially in enterprise AI and defense.