Nubia Page 2026 Software Ranking Uses Stale Revenue Data

That does not make Microsoft, Alphabet, Oracle, Palantir, Salesforce, IBM, ServiceNow, Intuit, Adobe, and Palo Alto Networks unimportant. They are all consequential technology vendors, and several are central to enterprise Windows, cloud, security, developer, and productivity deployments. It does mean the published ordering should not be read as a defensible 2026 ranking, especially by IT buyers trying to use it as a proxy for product durability, vendor health, or AI execution.

Microsoft’s annual report puts fiscal 2025 revenue at $281.7 billion, not the $261.8 billion stated in the article. Alphabet reported $402.8 billion for calendar 2025, rather than $359.3 billion. Palantir closed 2025 with $4.48 billion in revenue, well above the $2.87 billion figure used in the ranking. Those are not rounding differences; they reveal that the post has combined old revenue snapshots, old market-cap snapshots, and assertions made after later results were already public.

The revenue table does not match its own timeframe​

The article says it weighed revenue for the “most recent fiscal year” and reflects the first quarter of 2026. Yet its cited annual revenue figures do not consistently match the latest completed annual reports available as of August 11, 2026.

The source records are unusually clear here. Microsoft’s fiscal year ended June 30, 2025; its annual report says revenue grew 15% to $281.7 billion. Alphabet’s 2025 Form 10-K reports $402.8 billion. Oracle announced $57.4 billion for fiscal 2025. Salesforce’s fiscal 2026, which ended January 31, 2026, produced $41.5 billion in revenue.

The pattern continues through the rest of the list. Palantir’s 2025 Form 10-K reports 56% growth to $4.475 billion. ServiceNow reported $13.278 billion for 2025, Intuit reported $18.831 billion for fiscal 2025, and Adobe reported $23.769 billion for its year ended November 28, 2025. Palo Alto Networks reported $9.2 billion for the year ended July 31, 2025.

A ranking can use trailing-twelve-month figures, a particular fiscal year, or estimates. But it has to say which one. Nubia Page labels the figures as the most recent fiscal-year results while using a mix that cannot be reconciled to that definition. The result is especially damaging because revenue is presented as a core ranking criterion rather than background color.

IBM’s “70% software” claim fails the basic math​

The IBM entry contains the clearest substantive error. It says software “now represents over 70% of IBM’s revenue mix,” framing this as evidence of a successful shift away from hardware and IT services.

IBM’s 2025 annual report says total revenue was $67.5 billion and software revenue was $29.962 billion. That puts software at roughly 44% of total revenue, not more than 70%. IBM’s consulting and infrastructure segments remain large businesses, even as the company pursues Red Hat-led hybrid cloud, automation, data software, watsonx, and mainframe modernization.

IBM can fairly be described as more software-focused than it was before the Red Hat acquisition. Its software revenue grew 10.6% on a reported basis in 2025, with Automation, Data, and Hybrid Cloud all contributing. But the ranking turns a strategic direction into a revenue fact that IBM’s own financial statements do not support.

This matters beyond a correction to a listicle. IT leaders evaluating IBM’s platform strategy need to understand that IBM is still a mixed vendor: software, consulting, infrastructure, transaction processing, and mainframe platforms are linked in its commercial model. Treating it as a mostly pure-play software company obscures how contracts are sold, deployed, supported, and priced.

“Best” is never defined in a way readers can audit​

The article says it considered revenue, market capitalization, growth trajectory, product breadth, AI and cloud positioning, employee sentiment, and workplace reputation. It does not publish weights, scoring rules, data dates, source documents, or a mechanism for resolving conflicts between those measures.

That omission makes the numbered order arbitrary. A company can lead by market value while trailing in revenue. Another can lead in growth while serving a narrow set of industries. Employee-review scores are volatile, self-selected, and not comparable with audited financial data. “Strategic position in AI” is a judgment, not a measurable financial category, unless the author identifies a test such as AI-specific annual recurring revenue, paid-seat adoption, gross-margin changes, cloud consumption, or customer retention.

Palantir is the test case. Its inclusion near the top reflects an intelligible market view: its commercial business and government contracts expanded sharply in 2025, and its Artificial Intelligence Platform has become a major part of its investor narrative. But placing Palantir fourth, ahead of Salesforce and IBM, requires a stated rule for valuing growth, scale, revenue concentration, government exposure, profitability, and valuation risk. The article provides none.

The same problem applies to Alphabet. Google Cloud, Gemini, Android, Workspace, Chrome, and security products make Alphabet a major software platform owner. But Alphabet is also overwhelmingly dependent on advertising. If the category means the largest American companies with substantial software businesses, its inclusion is logical. If it means software vendors ranked as software vendors, the definition has to explain why an advertising-led conglomerate belongs while other major enterprise technology companies do not.

A ranking is allowed to be opinionated. It is not allowed to disguise unweighted editorial judgment as a financial model.

Market capitalization cannot be presented as a fixed 2026 fact​

The listed market caps are said to be “as of early 2026,” but no date is supplied. That is a major missing detail, not a cosmetic one. Market capitalization changes with every trading session, and the difference between a January 2, 2026 valuation and an August 11, 2026 valuation can materially alter the order of companies whose stock prices trade on AI expectations.

The issue is amplified by the article’s use of valuation as evidence. Its Palantir discussion cites a price-to-sales ratio above 60 times while pairing it with revenue and market-cap figures drawn from different periods. A price-to-sales ratio is only meaningful when the numerator and denominator use a clearly defined valuation date and a clearly defined revenue period. Mixing a later market value with an earlier annual revenue number creates an inflated or distorted multiple.

For enterprise customers, this is not merely an investor-data complaint. A high market capitalization does not establish that a vendor has the most mature admin tooling, clearest licensing, lowest migration risk, strongest security response process, or best support record. Those are the questions that determine whether Windows, Azure, Microsoft 365, Oracle Cloud Infrastructure, ServiceNow, Salesforce, Adobe, or Palo Alto Networks belongs in a production environment.

Microsoft’s presence at the top is easy to defend on scale and enterprise reach. Its $281.7 billion fiscal 2025 revenue, $168.9 billion Microsoft Cloud business, Windows installed base, Microsoft 365 footprint, Azure operations, GitHub ownership, and Copilot product push make it one of the few vendors that touches virtually every layer of a typical corporate IT estate. But that conclusion comes from the underlying business, not from a ranking that misstates the financial baseline.

AI claims need product adoption evidence, not product names​

The article’s broader observation—that AI is now a central strategic concern for the largest software and technology companies—is directionally sound. Microsoft has Copilot, Alphabet has Gemini, Salesforce has Agentforce, Adobe has Firefly, ServiceNow has Now Assist, IBM has watsonx, Oracle is embedding AI in its database and cloud portfolio, and Palo Alto Networks has positioned AI across its security platform.

What the ranking does not show is whether those products are creating durable revenue, reducing churn, improving customer outcomes, or simply increasing vendor marketing activity. The distinction is already visible in public reporting. ServiceNow disclosed that Now Assist exceeded $600 million in annual contract value during 2025, while Palo Alto Networks reports next-generation security ARR as a separate metric. Other vendors disclose less, combine AI with wider cloud or subscription categories, or discuss adoption without a comparable revenue breakout.

That makes “AI leadership” unsuitable as a ranking factor unless the author separates observable commercial performance from announced capabilities. A Windows administrator assessing Microsoft 365 Copilot or Copilot Studio needs data governance, tenant controls, licensing terms, model boundaries, retrieval permissions, auditability, and support commitments. The fact that a company appears in a top-ten list says nothing useful about those implementation questions.

The practical conclusion is straightforward: this is a recognizable vendor roster, not a reliable 2026 ranking. Readers can use it as a reminder of which companies dominate enterprise software conversations, but should discard the numerical order and recheck every financial claim against current filings before relying on it for investment, procurement, or platform-planning decisions.