The State of Cloud 2026.
Read it from inside the product.
An Emerging Markets view of global cloud, grounded in primary telemetry from 25 production cloud accounts. This edition renders the report as a live ZopDev console. The numbers are measured, not modeled.
Cloud has stopped being new. So why is the bill still a mystery? After fifteen years of migration the question is no longer whether to run on cloud. It is why the bill is what it is, and what it could look like if we treated it like a product. Ten falsifiable hypotheses were written before a single dashboard was opened. All ten were confirmed.
Every pre-registered hypothesis confirmed by the evidence. The conventional wisdom is right. The will to act is what is missing.
System overview.
Three trajectories.
Every enterprise we examined runs on one of three cost trajectories. Most are drifting. A few are cutting. The rest are quietly quadrupling.
Active FinOps embedded in engineering. Recommendations get triaged, owners get assigned, savings get tracked. One Optimizer halved daily spend in 9 weeks. No migration. No rebuild. Just hygiene.
Detection in place, action absent. Cost grows quietly, 5 to 10% a quarter, from accumulated unfixed waste. No new product launch. Nobody owns the output. The default state of enterprise cloud.
Growth-stage spending. Each new team gets its own workspace, region, NAT gateway, snapshot policy. The growth is legitimate. The waste embedded inside it is not, and it is baked in for later.
Account count and spend are not the same shape. Azure commands disproportionate enterprise spend on far fewer accounts. AWS still takes two-thirds of every cloud dollar.
The average enterprise cloud bill is a meaningless number. One customer halves spend while another quadruples it in the same quarter.
The defensible number.
If you take only one statistic from this report, take this one. It is the floor, not the ceiling. We only counted what we could prove.
| Annualized spend · cohort | $2,222,930 |
| Recoverable savings · proven | $375,217 |
| Waste-to-spend ratio | 16.9% |
For an enterprise spending $10M a year on cloud, that is $1.69M on the table. Across the near-$680B global market, the implied addressable waste is north of $100B annually.
Waste telemetry.
87% is just forgotten.
Vendor narratives focus on Reserved Instances and Savings Plans. Those are 6% of flagged items. The real waste is mundane: things nobody remembers creating.
The biggest cloud waste problem is not that you bought the wrong thing. It is that you forgot to delete the right thing. Each orphan takes about 5 minutes to fix.
Detection is solved.
Action is not.
Cloud teams identify waste roughly 30× faster than they fix it. The bottleneck is execution, not awareness. The gap is entirely human.
The platform stops at the dashboard. The work happens, or does not, in the cloud console. Fewer than 1% of detected signals are formally applied through a remediation workflow.
Architecture signals.
Sprawl is the new debt.
The visible compute layer of any production account is dwarfed by its metadata. And the metadata, unlike compute, has no lifecycle.
| Pattern | Example | Regions | $/region/mo |
|---|---|---|---|
| Global sprawl, one tenant | Sporting goods, Azure | 32 | ~$60 |
| Concentrated production | Global CPG, Azure | 3 | ~$5,900 |
| Single-region monolith | Consumer internet, AWS | 2 | ~$12,500 |
| Spread without intent | E-commerce, AWS | 19 | ~$560 |
Two of these reflect deliberate architecture. Two reflect accident. Footprint is a choice, but most enterprises have not made it consciously.
One customer ran 33 distinct Azure subscriptions in a single tenant. Another ran 3 for comparable spend. Databricks workspace proliferation is the most visible symptom.
Every AWS enterprise in the cohort had at least one Single-AZ production database. The fix is trivial and the cost is a small storage uptick. The benefit is the difference between a 5-minute outage and a multi-hour one when an AZ fails. The reason is never a deliberate choice. It is always: we ran the migration script in 2021 and never went back.
The lens, and the next shift.
The 16.9% finding holds from Mumbai to Mountain View. How the waste accumulates is what differs by geography. And the next inflection has three more zeros.
Today AI/GPU is under 2% of enterprise bills. By Q2 2027 it crosses 15%. The same anti-patterns return with three more zeros on every waste number.
GPU utilization becomes the new leading indicator. Most enterprises will discover their GPUs run under 30% utilization, the same way they discovered their EC2 fleets did in 2015.
System recap.
Ten hypotheses, all confirmed. One defensible number. Three trajectories. The whole report on one board.
It is not can we cut cloud spend.
It is why haven't we already.
16.9% is the headroom you did not know you had. Detection is a solved problem. The bottleneck is ownership. ZopDev closes the detection-to-action gap: someone owns the list, every day, and the fixes actually get applied.