AWS and Microsoft Azure remain the two leaders in Flexera’s 2026 customer-spending comparison, but the survey does not produce a decisive overall winner. Azure has the stronger concentration in several midrange enterprise spending tiers, while AWS leads the key $200,001–$500,000 enterprise bracket, has more very large VM estates, and is stronger among small and midsize businesses (SMBs). Google Cloud Platform (GCP) is a clear third, while Oracle Cloud Infrastructure (OCI) and IBM Cloud appear in fewer of the largest spending tiers.
The crucial qualification is that these figures describe respondents’ reported monthly spending bands. They are not cloud-provider revenue, dollar-weighted market share, or a ranking of profitability.
As an Amazon Associate I earn from qualifying purchases.
What Flexera’s comparison actually measures
Flexera’s 2026 State of the Cloud report surveyed 753 organizations: 620 enterprises and 133 SMBs. The survey was conducted in winter 2025. CRN reports that respondents were distributed across the Americas (62%), Europe (23%), and Asia-Pacific (13%).
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Respondents reported how much they spend per month with individual cloud providers. Because organizations commonly use multiple clouds, the percentages are not mutually exclusive. A company can appear in AWS, Azure, and GCP spending tiers at the same time.
#1 Best Overall
That creates three different measures readers should keep separate:
- Adoption: whether organizations run some or significant workloads on a provider.
- Spending distribution: the monthly spending band reported by organizations using that provider.
- Market share or revenue: the providers’ total sales across the global market, which this survey does not establish.
For example, AWS’s 16% share of enterprise respondents in the $200,001–$500,000 monthly tier does not mean AWS owns 16% of the cloud market. It means 16% of surveyed enterprise respondents reported spending in that band with AWS.
Overall customer-spending comparison
The following figures are CRN’s presentation of Flexera’s Figure 37. Some provider rows combine adjacent bands or use broader categories, so the columns are not perfectly comparable in every case.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →| Provider | Below $50K | $50K–$100K | $100K–$200K | $200K–$500K | $500K–$1M | $1M–$2M | $2M–$5M | Above $5M |
|---|---|---|---|---|---|---|---|---|
| AWS | 17% | Not separately stated | Not separately stated | Approx. 40% combined | 9% | 9% | 6% | 5% |
| Azure | 18% | Not separately stated | Not separately stated | Approx. 41% combined | 11% | 6% | 6% | 5% |
| GCP | 20% | Not separately stated | Not separately stated | Approx. 28% combined | 6% | 5% | 3% | 3% |
| OCI | 15% | Approx. 15% across $50K–$200K | — | 4% | 3% | 3% | Included above $2M | 3% above $2M |
| IBM Cloud | 12% | 6% | 5% | Approx. 6% across $200K–$1M | Not separately stated | Not separately stated | Included over $1M | 5% over $1M |
| Alibaba Cloud | 8% | Approx. 4% across $50K–$200K | — | Approx. 3% across $200K–$1M | Included in broader bands | Included in broader bands | 3% across $1M–$5M | 0% |
Source: CRN’s presentation of Flexera’s 2026 data. Percentages represent respondents, not provider revenue.
AWS versus Azure: close at the top
Azure leads AWS in two important enterprise middle-spending tiers. It reaches 15% of enterprise respondents in both the $50,001–$100,000 and $100,001–$200,000 monthly bands, compared with 13% for AWS in each.
Rank #2
AWS, however, leads the $200,001–$500,000 tier: 16% of enterprise AWS respondents fall there, versus 14% for Azure. At the highest reported level, both providers have 6% of enterprise respondents spending more than $5 million per month.
The broader infrastructure signal also favors AWS at the extreme end. AWS leads deployments above 1,000 virtual machines (VMs), at 11%, compared with 8% for Azure. Azure leads the medium-sized 51–100 VM category, at 22% versus 18% for AWS.
Free tools Windows power users keep installed
One-click scans. No signup required.
Usage data reinforces the idea that the two providers are closely matched rather than separated by a clear winner. AWS is used for some or significant workloads by 83% of respondents, compared with 79% for Azure. Among enterprises, active workload use is 84% for AWS and 82% for Azure. When experimentation and planned use are included, Azure reaches 94%, compared with 92% for AWS.
Enterprise results
Enterprise spending shows Azure’s strength in the middle and AWS’s strength in larger deployments.
| Provider | Below $50K | $50K–$100K | $100K–$200K | $200K–$500K | $500K–$1M | $1M–$2M | $2M–$5M | Above $5M |
|---|---|---|---|---|---|---|---|---|
| AWS | 11% | 13% | 13% | 16% | 11% | 10% | 8% | 6% |
| Azure | 13% | 15% | 15% | 14% | 13% | 7% | 7% | 6% |
| GCP | 18% | 11% | 9% | 10% | 7% | 5% | 4% | 3% |
| OCI | 16% | 10% | 7% | 4% | 3% | 4% | 2% | 1% |
| IBM Cloud | 13% | 7% | 6% | 4% | 3% | 2% | 2% | 1% |
| Alibaba Cloud | 8% | 2% | 3% | 3% | 1% | 2% | 1% | Not stated |
GCP has the largest low-spend enterprise concentration: 18% of its enterprise respondents report spending below $50,000 per month. AWS is at 11%, while Azure is at 13%. At the top end, GCP has 3% above $5 million, and OCI and IBM each have 1%.
Rank #3
GCP’s profile may reflect a relatively larger population of smaller deployments or exploratory workloads. The survey alone cannot establish whether that results from differences in workload type, customer mix, pricing, technology preferences, or another factor. It does not prove that GCP is cheaper, less capable, or less satisfactory.
Recommended Free Tools
SMB results: AWS has the clearer lead
The SMB data points more clearly toward AWS, although it should be treated as directional because it covers only 133 respondents.
| Provider | Below $50K | $50K–$100K | $100K–$200K | $200K–$500K | $500K–$1M | $1M–$2M | $2M–$5M | Above $5M |
|---|---|---|---|---|---|---|---|---|
| AWS | 45% | 17% | 9% | 5% | Not stated | Not stated | Not stated | Not stated |
| Azure | 38% | 16% | 8% | 4% | 0% | 2% | 2% | Not stated |
| GCP | 31% | 11% | 3% | 3% | Not stated | Not stated | Not stated | Not stated |
| OCI | 11% | 5% | 2% | 2% | 0% | 0% | 0% | 0% |
| IBM Cloud | 7% | 5% | 2% | 1% | 1% | 0% | 0% | 0% |
| Alibaba Cloud | 5% | 3% | 0% | 0% | 0% | 0% | 0% | 0% |
Forty-five percent of SMB respondents report spending below $50,000 per month with AWS, compared with 38% for Azure and 31% for GCP. No SMB respondents reported OCI spending above $500,000 or IBM spending above $1 million. Those zeroes should not be interpreted as proof that no SMB anywhere reaches those levels; they describe this sample.
Where OCI and IBM Cloud fit
OCI and IBM Cloud have much smaller representation in the largest spending tiers in this survey. Each has 1% of enterprise respondents above $5 million per month. Among SMB respondents, OCI has no reported spending above $500,000, while IBM has no reported spending above $1 million.
The survey does not explain why. Plausible structural factors include OCI’s concentration in Oracle databases and enterprise applications, and IBM Cloud’s relationships with regulated industries, hybrid-cloud customers, Red Hat users, and established IBM accounts. Some related spending may also appear through private cloud, hosted infrastructure, software contracts, managed-service providers, or other arrangements that are not captured consistently under each provider’s label.
A smaller share of high-spend respondents is therefore not enough to call either provider irrelevant or to conclude that it is losing. Their strategic importance may be concentrated in particular workloads rather than spread broadly across the survey population.
Does spending correspond to VM scale?
Flexera’s VM data provides useful context, but it still measures deployment distribution rather than provider revenue.
- Roughly one-quarter of respondents have 1–50 VMs with AWS, Azure, GCP, and Oracle; IBM is at 17%.
- Azure leads the 51–100 VM range at 22%, ahead of AWS at 18% and GCP at 16%.
- AWS leads deployments above 1,000 VMs at 11%, compared with 8% for Azure and 5% for GCP.
The practical interpretation is that Azure appears especially strong in medium-sized VM estates, while AWS has the strongest very-large-estate signal. Neither result says which provider delivers the best cost per workload.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Why a high cloud bill is not automatically a bad deal
Flexera reports that 85% of respondents still regard managing cloud spending as a top challenge. Sixty-three percent have established FinOps teams, and 64% report that FinOps has delivered value to business units. Estimated wasted IaaS and PaaS spending rose to 29%, reversing a five-year decline.
These findings change how the spending tiers should be read. A large monthly bill can reflect a large, productive workload, while a small bill may simply reflect limited adoption. Effective economics depend on negotiated discounts, commitment programs, licensing, data transfer, storage, managed services, GPU or other AI accelerator costs, and utilization.
Best Value
AI is an additional source of volatility. Flexera says public-cloud GenAI usage rose from 50% to 58%, while extensive GenAI use increased from 36% in 2025 to 45% in 2026. GPU capacity, model-serving charges, storage, and data movement can quickly change a provider’s cost profile.
For that reason, FinOps teams should compare workload-level unit economics—such as cost per transaction, customer, job, or model inference—not just total invoices. The comparison should also account for the 73% of organizations using hybrid cloud, where duplicated tooling and operations can offset apparent infrastructure savings.
How buyers should use the findings
Enterprise checklist
- Map existing software licenses, enterprise agreements, and discount commitments.
- Compare identity, security, governance, support, and compliance integration.
- Measure data gravity, egress exposure, storage, and cross-cloud transfer costs.
- Confirm required regions, databases, analytics, AI, Kubernetes, and managed services.
- Assess available engineering skills, partner support, and operational maturity.
- Test whether spending can be allocated accurately to teams and business units.
- Include exit costs, portability, migration effort, and the risk of vendor lock-in.
SMB checklist
- Prioritize simple billing, budgets, alerts, and predictable pricing.
- Check for experienced local partners and low-overhead managed services.
- Use account, project, subscription, and tagging structures that remain understandable as the company grows.
- Evaluate managed databases and serverless services where they reduce operational work.
- Control accidental costs from idle resources, high-volume logs, data transfer, and oversized instances.
- Treat credits and introductory incentives as temporary benefits, not proof of lower production cost.
Multicloud checklist
- Compare each workload on the provider where it performs best economically and operationally.
- Standardize cost allocation, identity, security, observability, and policy controls where practical.
- Measure the staff and tooling overhead created by running multiple clouds.
- Review whether commitments and discounts encourage unnecessary lock-in.
- Verify that FinOps data includes negotiated rates, marketplace purchases, managed services, and current usage.
Bottom line
Flexera’s 2026 survey shows AWS and Azure in a close contest, not a definitive overall victory. AWS is stronger among SMB respondents, in the $200,001–$500,000 enterprise tier, and among deployments above 1,000 VMs. Azure leads the $50,001–$200,000 enterprise bands and the 51–100 VM category. GCP is a substantial third provider with a higher concentration of lower-spend respondents, while OCI and IBM Cloud have smaller but potentially important strategic footprints.
The result is best used as a customer-profile and workload-scale signal—not as a cloud revenue ranking. For a buying decision, the relevant question is which provider produces the best risk-adjusted unit economics for each workload after discounts, licensing, data movement, operations, and FinOps controls are included.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
