Can Palantir CEO Alex Karp Turn Shorter Contracts Into Long-Term Growth?
Palantir CEO Alex Karp nearly doubled the company’s quarterly revenue, and it still isn’t enough. On Nov. 2, the A.I. giant announced it had generated $2.3 billion in revenue during its third quarter, representing a 96 percent growth rate over the same period the year before. “We have now essentially doubled our entire business since the same quarter last year,” Karp said in the shareholder letter, singling out a “ravenous” demand for its sovereign A.I. solutions.
According to its business update, Palantir closed 212 deals during the third quarter of at least $1 million.
At the same time, U.S. commercial revenue reached $984 million, up from $397 million in the same quarter last year.
U.S. government revenue reached $953 million, up from $486 million a year earlier. The key theme here is that Palantir has traditionally been a company whose revenue has come from government contracts, and now its commercial business has surpassed its government business in the U.S. for the first time.
Palantir reported a 47 percent operating margin under generally accepted accounting principles. Its adjusted operating margin, which excludes stock-based compensation and related employer payroll taxes, was 62 percent.
Palantir’s emphasis on sovereign development appears to be the company’s core approach to turning enterprise interest into paying customers. “We have no interest in building software that leeches the value and alpha from our partners,” Karp said in the Nov. 2 shareholder letter.
However, while its earnings performance was impressive, Palantir’s underlying contract figures are growing more slowly than revenue, and some large initial deals are shorter than usual. Total contract value booked during the quarter rose 6 percent year over year to $2.9 billion, while U.S. commercial contract value rose 11 percent. These bookings include contract options and can differ from the revenue ultimately collected.
Palantir has also acknowledged shorter-term contracts. “We signed multiple eight-figure deals in the quarter that were shorter in duration than usual,” Palantir CFO David Glazer said in the Nov. 2 earnings call. “As we continue to deliver value with Sovereign A.I., over time, we expect both the duration and size of these contracts to increase.” Its commercial strategy is to onboard customers with relatively short engagements before expanding these into longer-term relationships.
“Palantir focused on shortening the sales cycle, and in my view, it pioneered the forward-deployed engineer model, which lets it pilot and implement quickly. Fast conversion is every company’s dream,” Shashi Bellamkonda, principal research director at Infotech, a technology research and advisory firm, tells Observer. “Shorter contracts mechanically shrink total contract value even when annual spend is the same, so part of the bookings slowdown is accounting optics. The other part is real. Customers are hedging. They want proof before committing to multi-year terms.”
The sovereign A.I. pitch
Its real rallying cry for continued growth is sovereign A.I. investment. “The Sovereign A.I. moment is purpose-built for Palantir. Our products, from Foundry and Ontology to AIP with its ontologized, fine-tuned open-weight models, ignite a self-improvement loop that joins and aligns the workflow with its optimal model and us with our partners. This dialectic drives transformational value for both Palantir and our partners and works only in artistic combination. Each becomes stronger as the sum becomes more powerful,” Karp said in the Nov. 2 earnings release.
“Palantir didn’t invent ontologies, but it was early to make them operational, connecting a model of how a business actually runs to real decisions and workflows,” Bellamkonda says.
Part of Palantir’s strategy is that its Artificial Intelligence Platform (AIP) can provide an application layer for deploying A.I. while enabling enterprises to retain control over proprietary information. It’s also established a strategic partnership with Nvidia, announced Sept. 10, to bring Nvidia’s Nemotron open models into Palantir Foundry and AIP, allowing customers to run Nemotron with their own data.
In a frontier A.I. race dominated by OpenAI and Anthropic, Karp has positioned himself as an alternative voice in the sector. In a buzzy July 1 interview with CNBC’s “Squawk Box,” he called out the industry’s token consumption model, arguing “the basic view among enterprises in this country is I’m going to chillax and waste my time with tokens.”
These workflows could keep customers tied to Palantir, according to Alan Webber, an IDC analyst. “This moves the lock-in from the model layer to the data and workflow layer, a whole new revenue model,” he tells Observer. “It is a novel idea, but the proof isn’t in yet.”
Karp has also argued that organizations want control over their compute, models, data and competitive advantage. Palantir is thus emphasizing sovereign development as a means to give enterprises control over the A.I. stack, while simultaneously maintaining a “symbiotic relationship” with Anthropic, Karp said on the Nov. 2 earnings call. The question now is whether this vision can translate into long-term commitment.
