Cloud Computing Stocks Explained: What Investors Should Know

Cloud computing stocks are often discussed as if they were one clean sector. They are not. A share of Amazon gives exposure to AWS, a share of Microsoft gives exposure to Azure within a broader Intelligent Cloud segment, and a share of Salesforce gives exposure to a mostly subscription-based enterprise software model that behaves very differently from a hyperscale infrastructure business. That distinction matters even more now, while the cloud infrastructure market is still expanding quickly: Synergy Research Group estimated Q1 2026 spending on cloud infrastructure services at $128.6 billion, with AWS, Microsoft, and Google at 28%, 21%, and 14% market share, respectively. (srgresearch.com)

Start by knowing which part of the cloud stack you own

A practical definition helps. NIST’s cloud computing framework breaks the market into SaaS, PaaS, and IaaS, and also separates service models from deployment models such as public and hybrid cloud. In plain English, SaaS companies sell finished software delivered over the internet, PaaS companies provide tools and environments for building and running software, and IaaS providers sell raw computing, storage, and networking capacity. If an investor does not know which layer a company primarily monetizes, almost every comparison that follows will be sloppy. (nvlpubs.nist.gov)

That is why the label cloud stock covers several different kinds of businesses. Amazon, Microsoft, Alphabet, and Oracle have major infrastructure or platform exposure. Salesforce, ServiceNow, Workday, and Adobe are much closer to application-layer cloud businesses, where revenue is recognized over time and the economics are shaped more by seat growth, contract renewals, and upsell than by server utilization. Adjacent businesses such as semiconductors, networking, cybersecurity, and data center REITs may benefit from cloud demand too, but they should not automatically be analyzed with the same yardstick as a company that actually sells cloud services. (s2.q4cdn.com)

Reported cloud revenue is not standardized

This is the first place many investors go wrong. Company presentations can make cloud exposure sound comparable when the underlying disclosures are not. Some companies give investors a nearly pure cloud segment. Others bundle cloud with legacy software, support, consulting, or productivity tools. Before you compare growth rates, margins, or valuations, read the actual segment description in the latest annual report or earnings materials. (s2.q4cdn.com)

The implication is simple: comparing AWS growth to Microsoft Intelligent Cloud growth is not an apples-to-apples exercise, and comparing either one to Salesforce or Adobe is a different exercise again. Fiscal calendars also differ, which can further distort quick comparisons. A careful investor compares like with like, then adjusts for disclosure quality before drawing conclusions. (s2.q4cdn.com)

Use the Cloud Stock Check before you compare valuations

A useful way to avoid narrative investing is to run every name through the same four-part test. Think of it as the Cloud Stock Check: revenue engine, durability, capital burden, and switching pain. It is not a formula for picking winners, but it does help separate businesses that only sound similar from businesses that actually share the same economic model.

1) Read the revenue engine

Start with how the company gets paid. Consumption-heavy cloud businesses can accelerate quickly when customer workloads rise, but they can also decelerate quickly when customers optimize spending. Google says Google Cloud Platform generates consumption-based fees and subscriptions for infrastructure, platform, and other services. Salesforce says its cloud services are generally recognized ratably over the contract term. ServiceNow, Workday, and Adobe also describe predominantly subscription-based models with revenue recognized over time. That means a 20% growth rate at a usage-driven infrastructure provider is not the same thing as a 20% growth rate at a seat-based enterprise SaaS platform. (sec.gov)

2) Check durability, not just growth

For recurring-revenue cloud stocks, the deeper question is whether customers stay, expand, and commit ahead. Salesforce reported total remaining performance obligation of about $72.4 billion as of January 31, 2026. ServiceNow reported $28.2 billion of RPO as of December 31, 2025, with 46% expected within the next 12 months, and a 98% renewal rate. Workday reported $28.1 billion of subscription backlog as of January 31, 2026. Adobe reported remaining performance obligations of about $22.52 billion as of November 28, 2025. Those figures are useful because they add context to top-line growth and can reveal whether demand is merely landing or actually sticking. (sec.gov)

But this is also where investors can fool themselves. Salesforce explicitly says RPO is not necessarily indicative of future revenue growth, because seasonality, renewals, contract duration, acquisitions, and foreign exchange can change the number. ServiceNow says mix of offerings and foreign currency can move RPO, and Workday says backlog can fluctuate with renewals and contract duration. So the right question is not “Is backlog up?” but “Why is backlog up, and how much of that increase is truly durable?” (sec.gov)

3) Match margins to capital intensity

Hyperscale cloud and SaaS may both live in the cloud, but their cost structures can be radically different. AWS generated $45.6 billion of operating income in 2025, yet Amazon’s technology and infrastructure expense also jumped 23% that year. Microsoft said Intelligent Cloud cost of revenue rose 36% in FY 2025 and that Microsoft Cloud gross margin percentage decreased to 69% because of scaling AI infrastructure, partly offset by Azure efficiency gains. Oracle said free cash flow was negative $23.7 billion in FY 2026 as it continued investing to support cloud infrastructure growth. By contrast, ServiceNow reported an 80% subscription gross profit percentage for 2025. The takeaway is that high growth may come with very different reinvestment needs. (s2.q4cdn.com)

4) Ask what makes switching painful

Cloud businesses become more valuable when they sit in a customer’s workflow in a way that is hard to dislodge. Infrastructure providers benefit when core workloads, data, security, and AI tools are integrated into the same environment. Google’s filing, for example, ties Google Cloud to infrastructure, platform, AI, cybersecurity, and data analytics services, while also including Workspace. Workday’s revenue model depends heavily on cloud applications tied to HR and finance processes. Salesforce’s cloud services are built around enterprise systems of record and support. Those are different flavors of switching cost, but all can support durability if the product remains mission-critical. (sec.gov)

Where investors often go wrong

• Treating all cloud revenue as equal. A dollar of AWS consumption revenue, a dollar of Salesforce subscription revenue, and a dollar of Oracle AI infrastructure backlog can have very different visibility and cost profiles. (s2.q4cdn.com)
• Using the headline growth rate without reading the segment note. Microsoft’s Intelligent Cloud is broader than Azure, and Alphabet’s Google Cloud includes both infrastructure and Workspace. (microsoft.com)
• Assuming backlog is a guarantee. Companies themselves warn that RPO and backlog are influenced by timing, term, renewals, acquisitions, and currency. (sec.gov)
• Ignoring the cost of AI capacity. Cloud demand can be strong while margins compress or free cash flow weakens because infrastructure spending rises. (microsoft.com)
• Comparing valuations across unlike models. Hyperscalers, hybrid software vendors, and pure SaaS names deserve different expectations for margin, capital needs, and revenue visibility.

A simple example of how the framework changes the conclusion

Consider a hypothetical choice between two cloud stocks. Company A is growing 30% but relies heavily on consumption-based AI workloads, has falling gross margins, and must keep spending aggressively on data center capacity. Company B is growing 14%, but more than 90% of revenue is recurring, renewal rates are high, backlog is expanding, and margins are steadier. The louder story is Company A. The more predictable business may be Company B.

Neither is automatically the better investment. A more aggressive investor may prefer the upside and optionality of infrastructure-led growth. A more valuation-sensitive or risk-conscious investor may prefer the steadier cash-generation profile of a mature SaaS platform. The point is not that one model wins; it is that the business model should shape the analysis before the ticker does.

A practical checklist before buying a cloud stock

1. Read the latest annual report and the latest earnings release, then write down exactly how the company defines its cloud business.
2. Separate subscription, consumption, support, services, and legacy license revenue wherever possible.
3. Check RPO, backlog, ARR, renewal, or retention metrics, then read management’s caveats about what can distort them.
4. Compare margin trends with the company’s capital needs. If growth is improving while gross margin and free cash flow are deteriorating, ask whether the reinvestment is cyclical, structural, or temporary.
5. Look for evidence that customers are embedding the product into essential workflows, not just testing it during a spending boom.
6. Only after all of that should valuation comparisons begin, and even then the cleanest comps are usually companies with similar revenue models.

What matters most right now

As of Q1 2026, the broad cloud market is still growing fast, and AI is a major reason. Synergy said cloud infrastructure services grew 35% year over year in the quarter, the strongest pace since late 2021, while the top three providers still held dominant share. That is supportive for the sector, but it does not erase the need to tell apart usage-driven infrastructure names, hybrid software vendors, and recurring-revenue SaaS companies. The market opportunity may be shared. The economics are not. (srgresearch.com)

For most investors, the smartest next step is not hunting for “the best cloud stock.” It is building a short list of comparable names, reading how each one reports cloud revenue, and deciding which combination of growth, durability, margin structure, and capital intensity actually fits the kind of risk being taken.

FAQ

References