A military software contract can sound impressive without telling investors much about the business behind it. The announced value might cover several years, include spending that has not been funded, or require extensive implementation work before revenue appears. For investors researching defense tech stocks beyond aircraft and missile manufacturers, Palantir (PLTR) provides a useful company to examine. Its role is not to build the weapon platform, but to help military users combine information and coordinate decisions. Understanding that distinction is the starting point for judging both its opportunity and its share price.
What Palantir Does—and Where It Fits
Palantir builds software that connects data from separate systems and makes it usable within an organization’s operations. In defense, those systems can include intelligence databases, logistics records, maps and information from sensors. The practical problem is that collecting data does not automatically make it accessible, trustworthy or useful to the person making a decision.
Its Gotham platform serves defense and intelligence workflows. Foundry addresses data integration and operations across commercial and government organizations. Apollo supports software deployment and updates across different computing environments, including restricted networks. Its Artificial Intelligence Platform, or AIP, adds tools for using AI models within organizational workflows, subject to access controls and other constraints.
These product names matter less than the underlying task. A customer needs software that connects its information, preserves permissions and lets users take practical action. A logistics team, for example, might need to connect equipment condition, available spare parts and maintenance schedules rather than inspect each database separately.
Palantir is not a newly established business. It belongs in the software-defined military discussion because its business model differs from the traditional defense prime model, not because it appeared recently.
Lockheed Martin (LMT), for example, earns revenue from major defense platforms and associated systems and services. Palantir primarily supplies software and related implementation capabilities. AeroVironment (AVAV) offers another form of technology exposure through unmanned systems and related products, but manufacturing, component availability and production capacity matter more directly to that business.
| Business model | What the customer receives | Main economic questions |
|---|---|---|
| Defense software | Data integration, applications and operational workflows | Does usage expand, and how much support does each deployment require? |
| Unmanned systems | Aircraft, payloads and related services | Can production scale reliably at acceptable margins? |
| Traditional prime contractor | Large platforms, systems integration and support | How do contract terms, program execution and cost overruns affect profit? |
These categories overlap. A software supplier may work with a prime contractor rather than replace it. Large contractors also develop software internally. The useful question is not which category will win all military spending, but which company controls an important capability and can earn attractive profits from delivering it.
For the broader procurement framework, our guide to how defense contracts become revenue explains why awards, funding and reported sales need separate treatment.
How the Company Makes Money
Palantir earns revenue through contracts for access to its software platforms and related services. Deployments can involve cloud subscriptions, software running in customer-controlled environments, implementation and ongoing support. Revenue recognition depends on the contractual obligations and how the services are delivered; an announced award does not necessarily become revenue immediately.
The economic ambition is straightforward: establish a useful deployment, integrate it deeply enough to become important, then expand across more users, functions or organizations. Once a customer has connected sensitive information and built operational processes around the platform, replacement can require substantial work.
That creates potential staying power, but it is not the same as guaranteed renewal. Customers can reduce scope, delay expansion or decide that maintaining another supplier is worth the disruption.
Separate Defense Exposure From Total Company Growth
Palantir reports government and commercial revenue. Government revenue includes more than US military work, while commercial revenue includes customers outside defense entirely. Investors therefore should not describe every dollar of company growth as evidence of rising military software demand.
As a clearly dated reference point, Palantir reported approximately $2.87 billion of total revenue for fiscal 2024. Government revenue was approximately $1.57 billion, with commercial revenue accounting for the remainder. Those historical figures establish the business mix at that time; they are not a substitute for the latest quarterly filing.
This distinction affects valuation. An investor paying for rapid commercial AI adoption is making a different assumption from one expecting steady expansion of government programs. Both can support the same company, but they carry different sales cycles, competitive conditions and risks.
Contract structure also matters. Consider the difference between these measures:
- Contract ceiling: The maximum potential spending permitted under an award, subject to its terms.
- Funded commitment: Spending backed by an actual funding obligation, which may cover only part of the potential award.
- Recognized revenue: The amount recorded as the company satisfies its accounting obligations.
- Cash collected: Money received, which can arrive before or after revenue recognition.
A contract can be strategically important without providing its full headline value in near-term revenue. Government awards may contain options, termination rights or future funding requirements. Even a funded order can take time to turn into software delivery and cash receipts.
Investors should also distinguish a repeatable product from a labor-heavy project. Software can support strong margins when the same core platform serves many customers. If every deployment needs extensive custom engineering, revenue growth may require much more hiring than the software label suggests.
The Bull Case: Important Software Can Expand With the Customer
The strongest argument for Palantir is that military organizations need working connections between systems they already own. Buying another sensor or unmanned aircraft does not resolve the problem of sharing information across different networks, services and command structures.
Software that makes existing equipment and data more useful can address that problem without requiring the supplier to manufacture every physical component. If an initial deployment proves valuable, the customer may extend it into additional operations. That expansion can spread product development costs across more revenue.
Palantir also has experience operating in demanding government environments. Security requirements, permissions, procurement procedures and deployment restrictions create obstacles for suppliers unfamiliar with the customer. A commercial software demonstration is not enough to establish that a product can work reliably on a restricted network with operational consequences.
The potential competitive advantage is therefore broader than access to an AI model. It includes integration work, deployment infrastructure, institutional knowledge and applications that users already depend on. Models can change while the surrounding operational system remains in place.
For investors comparing that business promise with the share price, the useful exercise is to track growth, profitability and valuation together rather than treat a rising chart as confirmation of military adoption.
A successful investment case would require more than a series of announcements. Evidence would include customers extending deployments, government revenue growing across multiple programs and operating expenses increasing more slowly than revenue over time.
Commercial adoption could strengthen the economics by supporting product development that also benefits government customers. However, that benefit should be demonstrated rather than assumed. Commercial and military customers have different security requirements, purchasing processes and implementation needs.
There is also a distinction between helping users make decisions and automating decisions. Defense customers need accountability, appropriate human oversight and reliable performance. A system that fits those requirements may be more valuable than one that offers a striking demonstration but cannot be deployed responsibly.
The bull case ultimately rests on a measurable outcome: increasingly important software that customers continue funding, delivered at a cost that leaves a growing share of revenue as profit and cash.
The Bear Case: Procurement, Competition and Valuation
Government demand is durable in some respects, but procurement is neither smooth nor automatic. Budget negotiations, changing priorities, contract protests and administrative delays can interrupt the path from customer interest to revenue. A software supplier may have a useful product and still face a slow purchasing process.
Dependence on a limited set of important customers or programs can magnify those delays. A broad customer count does not necessarily remove concentration if a small group accounts for a large share of revenue. The latest filing’s customer concentration disclosures are more useful here than the total number of announced relationships.
Competition is another constraint. Defense agencies can choose established contractors, specialist software vendors, internally developed systems or combinations of suppliers. They may also seek to reduce dependence on a single platform through interoperability requirements and control over their own data.
Switching costs help a supplier only while the customer believes the product delivers enough value. High integration costs can also make prospective customers cautious about adopting the platform in the first place.
A Strong Business Can Still Be an Expensive Stock
Valuation can overwhelm good operating results. As a hypothetical illustration, suppose a company’s revenue doubles while its enterprise-value-to-sales multiple falls from 20 to 10. Its enterprise value would be unchanged. Shareholder outcomes could still differ because cash, debt and the diluted share count may also change.
That example is not a forecast for Palantir. It shows why revenue growth alone cannot establish whether a stock is attractively priced. The starting valuation embeds expectations about future growth, margins and the durability of both.
Enterprise value measures the market value of the operating business after adjusting equity value for items such as cash and debt. Comparing it with revenue can be useful, but it leaves profitability out of the denominator. Two businesses with identical sales can deserve different valuations if their sustainable margins differ substantially.
Stock-based compensation deserves separate attention. It is a real cost to shareholders even though it does not require an immediate cash payment. Free cash flow can look stronger than accounting profit partly because stock compensation is added back in the cash-flow statement.
The practical checks are whether stock compensation is becoming less significant relative to revenue, whether diluted shares keep rising, and whether growth in profit and cash flow remains attractive on a per-share basis. Repurchases should be assessed against issuance rather than treated as an automatic reduction in shares.
Finally, operational failure carries unusual consequences in defense. Security incidents, unreliable outputs or poor integration can damage customer trust and create legal or contractual exposure. Investors should not assume that a general-purpose AI improvement automatically translates into a dependable military application.
What to Watch Next: Evidence Before Headlines
The next useful milestone is not necessarily a larger contract announcement. It is evidence that deployments are expanding, funding is becoming firmer and the company is keeping more economic value from each additional dollar of revenue.
A repeatable review after each earnings release can focus on six questions:
- Where did growth come from? Separate US government, other government and commercial trends where disclosures permit. Avoid treating commercial AI demand as military procurement evidence.
- What has actually been funded? Read the terms behind major awards. Identify options, delivery periods and dependencies on future appropriations.
- Are customers expanding real usage? Look for disclosed deployment extensions and broader operational adoption, not just trials or announced partnerships.
- Are margins improving for sound reasons? Compare revenue growth with implementation costs, research spending and sales expenses. Separate recurring improvement from one-time effects.
- What is happening per share? Review diluted shares, stock compensation and cash generation together. Company-wide growth can overstate the benefit reaching each existing share.
- What does the valuation require? Test whether a reasonable range of growth and margin outcomes supports the expectations embedded in the current price.
Use the latest annual report, quarterly filing, earnings release and relevant contract disclosures as the primary record. Management presentations explain the strategy, but they do not replace the financial statements or the terms of an award. A simple research log can record the claim, the supporting disclosure and what evidence would contradict it.
Portfolio context matters as well. A holding can combine government procurement exposure with the valuation sensitivity of an AI software company. Owning it alongside other highly valued AI businesses may create more shared risk than a defense label suggests. Our guide to position sizing for high-risk technology stocks explains how to translate a downside assumption into a defined portfolio loss budget.
For comparisons with funds, inspect the actual holdings. A defense ETF dominated by established manufacturers may offer little exposure to software-led suppliers. A broad technology fund may already hold the same company, adding overlap rather than a genuinely different source of returns.
Palantir illustrates why defense tech stocks require two separate judgments. First, does the company provide a capability that customers will keep funding, with economics that improve as adoption expands? Second, does the share price leave room for procurement delays, competition and results that fall short of expectations? The business case can strengthen while the valuation becomes less forgiving. Keeping those judgments separate is the most useful discipline when assessing military software beyond the primes.