Palantir (Minutes): Projecting a 10-fold increase in U.S. revenue over the next 5 years

The following is a summary of the FY25Q2 earnings call of$Palantir Tech(PLTR.US) compiled by Dolphin Research. For an interpretation of the earnings report, please refer to《Palantir: Rightfully the Leader in AI Applications, Guidance Flaws Are Not an Issue》

I. Review of Core Financial Information

1. 3Q25 Guidance:

a. Revenue: $1.083-1.087 billion (QoQ +8%+, YoY +50%), marking the highest historical QoQ growth guidance.

b. Adjusted Operating Profit: $493-497 million.

2. Full-Year Guidance: Achieving GAAP profitability in each quarter of the year.

a. Total Revenue: $4.142-4.150 billion (YoY +45%), an increase of 9 percentage points from the previous quarter's guidance. U.S. commercial revenue: over $1.302 billion (YoY +85%).

b. Adjusted Operating Profit: $1.912-1.920 billion.

c. Adjusted Free Cash Flow: $1.8-2.0 billion.

d. Rule of 40 Score: Expected to be 94%.

2. The accelerating demand for AIP continues to drive our strong overall business performance in the U.S., with international government revenue primarily stemming from deepening demand from the UK government.

3. Q3 saw a seasonal increase in expenses due to new employee onboarding, but the adjusted operating margin will continue to expand in the second half of the year.

4. $6 billion in cash and cash equivalents on the balance sheet, with a remaining repurchase authorization of $899 million.

II. Detailed Content of the Earnings Call

2.1 Key Information from Executive Statements

1. Overall Revenue: Quarterly revenue surpassed $1 billion for the first time, with a YoY growth rate of 48%, continuing to rise rapidly. The "Rule of 40" score reached 94%, up 11 percentage points QoQ, setting a new record high, reflecting the simultaneous improvement in business scale and quality.

2. Contracts and Customer Expansion:

At the end of the quarter, total contract value (TCV) reached $2.3 billion, and annual contract value (ACV) reached $684 million, both setting new records.

a. Completed 157 transactions over $1 million, 66 transactions over $5 million, and 42 transactions over $10 million.

b. Enhanced customer stickiness: The average annual revenue of the top 20 customers over the past 12 months reached $75 million, a 30% increase YoY.

3. Core Value of Enterprise AI:

a. Emphasized that current large language models (LLMs) can only be considered fragmented intelligences, lacking even basic understanding capabilities and prone to catastrophic errors; whereas Palantir's Ontology achieves "pure cognitive ability" through software, serving as the core support for LLMs in real-world applications.

b. Significant progress in recent quarters: "Ontology Network Services" as a vision for customer architecture concepts is accelerating its implementation. AIP's positioning goes beyond "software used by customers," becoming "a platform for customers to build their own software." Compared to "non-specific tools" from large-scale service providers, AIP's "highly specific modules" enable customers to achieve value at 10 times the speed, attracting many software companies to migrate platforms.

4. Business Progress:

a. U.S. Government Business:

- Secured a $218 million order from the U.S. Space Force to provide seamless multi-domain operational support for space and air combat units.

- The Maven intelligence system contract ceiling increased by $795 million to meet the AI software needs of combat commands over the next four years. Usage has surged, doubling in the first nine months of 2024, doubling again in the following five months, and doubling once more since February 2025, with adoption rates continuing to climb.

- Signed a 10-year enterprise agreement with the U.S. Army, totaling $10 billion, integrating 75 contracts to deepen long-term cooperation.

b. U.S. Commercial Business: New business targets are higher, and the expansion speed of existing customers is accelerating, driving the continuous increase in revenue share.

5. Product Progress:

a. AI FDE: The design goal of AI FDE is to autonomously execute cross-domain tasks, including creating and editing ontologies, building data transformations, writing functions, debugging issues, etc., with closed-loop error handling capabilities (self-correcting and notifying humans when necessary), and integrating with AIP branch functions for seamless human-machine collaboration.

b. Warp Speed: Built on AIP, serving industrial customers in the nuclear energy and defense industrial base. Through the MRP speed module in the "Warp Speed" system, a customer reported that their production line balancing time was reduced from one day to one hour.

As the U.S. seriously undertakes the restoration of its maritime industrial base, we see a series of significant opportunities emerging in the U.S. shipbuilding industry. Last year, China built more ships than the total number constructed by the U.S. since the end of World War II. Now, the powerful engine of American innovation is accelerating to meet this challenge.

6. Future Outlook

Confident in the business for the second half of the year, with a continued focus on creating "practical AI application value" for customers and fully addressing the critical challenges entrusted by customers.

a. The platformization of AIP will be a core investment area in the product roadmap, enabling customers to achieve competitive leaps and capture market share first. It is also an important growth opportunity for both commercial and government businesses.

b. Continuously enhancing customer value through technological innovations (such as Ontology, AI FDE, Warp Speed) to consolidate industry leadership.

c. Collaborating with labor leaders to explore ways for workers to increase income through AI skills, emphasizing that "AI is a revolution in empowerment."

2.2 Q&A Session

Q: Besides using large language models (LLM), how does Palantir make AI more useful for frontline workers and decision-makers, rather than just data scientists?

A: There is a significant difference between ivory tower AI researchers and frontline realities. But we are always rooted in the production frontline—whether it's factory floor workers or fire units. We see on the frontline that AI is empowering American workers with superpowers: ICU nurses have more time with patients, reducing time spent organizing clinical records; submarine base workers are freed from three-day parts quoting work to focus on solving production issues.

What excites me most is the case of Panasonic Energy, where our AI can train workers who previously worked in casinos to operate sophisticated Japanese high-end battery technology. This expands employment, enhances agency and creativity, and is why we promote the "American Technology Fellowship." In practice, some creative AI applications come from blue-collar workers (like self-taught automotive practitioners), who not only use AI but also build AI applications that change business. We need to find and empower these people, releasing their potential in the U.S. economy through training.

Our practical experience shows that some of the most creative AI applications are coming from blue-collar workers—those self-taught automotive molders, rather than formally trained computer scientists. They are not only using AI but, more importantly, building AI applications that change their own business. We need to discover these people, empower them, provide certification, and release their potential in the U.S. economy through training.

Q: Given the enormous success achieved without essentially having a direct sales team, how do you view this? Would you consider forming one, or will you continue to remain cautious?

A: For those new to Palantir, we have made many controversial decisions, one of which is rejecting the common software industry model—relying on networking to sell useless, relationship-dependent products. We firmly believe in benefiting from the value created, and the company's organizational structure (including internal collaboration, flat hierarchy, etc.) revolves around this philosophy.

Our main sales force now and in the future is existing customer referrals, meaning customers tell others "invite them to see how to use their products to collaborate with their own culture." We do have direct sales personnel, but credibility enhancement is key, allowing customers to deeply use our products from the start (rather than old versions), pursuing more possibilities through Ontology, large models, etc. The disadvantage of this model is the lack of a large sales force, but the advantage is that once we engage with customers, we have extremely high credibility.

Customers believe we can help them profit, save money, or make soldiers safer and more efficient, stemming from recommendations by people they trust, and personnel mobility also brings opportunities (people who have interacted with us contact us after changing jobs). Now meetings have shifted from "not knowing how to use" to "asking how to achieve the same impact," driving business data growth.

In the next five years, U.S. business is expected to achieve 10-fold revenue growth, with many companies mentioning us in earnings calls, making cooperation smoother, without starting from a state of unfamiliarity, with a higher starting point and continuous improvement. We maintain a small, flexible sales team, with senior employees handling large transactions, focusing on creating value and then obtaining returns.

Currently, customer leadership is proactively cooperating, with faster and deeper cooperation initiation and expansion, and discussing enterprise-wide promotion from the start. This benefits from the U.S.'s plasticity advantage: compared to countries like Germany with fixed models, the U.S. is more open to creating value, with internal personnel benefiting more from company association, seeking effectiveness rather than empty talk.

Q: Regarding the AI action plan released by the White House a few weeks ago, you participated in the process by providing advice. What excites you most about it? What areas do you think still need improvement?

A: This is essentially about removing all constraints. AI is an empirical exploration journey that requires hands-on involvement and a focus on implementation. It clarifies the critical importance of having an open AI technology stack in the U.S.—it allows everyone to build solutions. Perhaps we were too worried in the past, not acting enough. Therefore, not only Palantir, but the entire industry and customers are eager to take action, realize value, and look forward to related support and subsequent investment.

Q: Considering the intense talent competition among software companies today, and the need to attract, retain, and cultivate the right talent to become mainstream software in the future, how do you ensure Palantir stays ahead in attracting the right talent compared to other companies?

A: In terms of talent, the key is to expose people to meaningful problems. Our retention rate is high because such opportunities are rare elsewhere. While some enjoy pure research, we can attract, retain, and motivate those who want to change history and solve problems that bring tangible results.

For talented individuals who recognize Western superiority (or can tolerate this view), Palantir is a very unique place. Over the past two decades, I've interacted with many Western institutions and companies, and none can offer what Palantir does, allowing employees to have full autonomy from the start, participating in projects that can make headlines, sometimes in as little as three months, and we are enhancing this through improved internal collaboration. Today, the experience at Palantir is already the most significant "passport" in the tech industry, setting the tone for a career.

Additionally, at Palantir, regardless of whether you attended school, the quality of the school, or your background, people only care about your experience, reputation, and projects at Palantir. We are creating a new qualification independent of class and background, which is the most important qualification in the tech industry. It is this full autonomy and new qualification that make us excel in talent recruitment and retention.

<End Here>

Risk Disclosure and Statement of This Article:Dolphin Research Disclaimer and General Disclosure