The Procurement Lens.
Category intelligence · Technology · Cloud & hosting

Cloud & hosting

Public cloud infrastructure and platform services (IaaS and PaaS), GPU capacity, hosting and colocation.

Strategic · high spend, three dominant providers, rising AI demand

At a glance

Global cloud market

Cloud services spending, Q3 2026
$168bn
+55% year on year
OfficialSynergy Research, via CIO Dive · Oct 2026
Wasted IaaS/PaaS spend
29%
First rise in five years, driven by AI
AWS GPU reservation prices
+15%
H200 Capacity Blocks, January 2026
AI share of cloud market
≈1/3
AI-specific cloud services
OfficialSynergy Research, via CIO Dive · Oct 2026
Cloud is in a capacity war, not a price war. AI demand has pushed the big three to 39-103% growth and AWS has raised GPU prices for the first time in its history. Meanwhile 29% of spend is wasted. The biggest saving is still governance, not discount.

Exhibit 1The big three are growing faster than ever

Cloud revenue growth, Q3 2026 vs Q3 2025

Google Cloud+103% ($30.7bn)Microsoft Azure+48% ($33.2bn)AWS+39.3% ($46.0bn)

Google Cloud includes Workspace; figures are not strictly comparable.

Showing: Global

The category on a page

What it isCompute, storage, databases and AI platforms rented from cloud providers, plus GPU capacity and hosting.
How bigCloud services spending reached $168bn in Q3 2026 alone (Synergy).
Market structureAWS 26%, Microsoft 19%, Google 16% share; neoclouds (CoreWeave, Nebius) and Oracle gaining in AI.
PositioningStrategic: high spend, deep technical lock-in, three dominant providers.
Key riskAI-driven waste, rising GPU prices, and commitments you can't use.
Primary leverFinOps: right-size, schedule and commit only proven baseline usage.

Exhibit 2Where cloud sits

Profit impact against supply risk, technology subcategories

ROUTINELEVERAGEBOTTLENECKSTRATEGICCloudSoftwareHardwareIT servicesCybersecurityTelecomProfit impact →Supply risk →
JudgementThe Procurement Lens assessment

Exhibit 3Cloud waste is rising again

Estimated share of IaaS/PaaS spend wasted

25%26.5%28%29.5%31%2023202420252026Wasted spend

Exhibit 4Market share

Share of cloud infrastructure services, Q3 2026

AWS26%Microsoft19%Google16%
OfficialSynergy Research, via CIO Dive · Oct 2026

Exhibit 5GPU prices are going up

AWS p5e.48xlarge (8× H200) Capacity Block, $ per hour

Before Jan 2026$34.61After Jan 2026 (+15%)$39.8

A further ~20% rise on Capacity Blocks took effect on 1 July 2026.

Exhibit 6Contracted backlog is huge

Committed future revenue, latest disclosed, $bn

Microsoft (commercial)$677bnAWS$578bnGoogle Cloud$561bn

Providers are capacity-constrained and have little need to discount.

Typical savings from FinOps and commitments

On IaaS/PaaS spend

15%30%
0%20%40%

Value levers

In order of value.

01Cut waste first

Right-size, shut idle resources and schedule non-production. AI workloads are making waste worse.

29% of IaaS/PaaS spend wasted (up from 27%)
02Commit only the proven baseline

Use savings plans and reservations for steady usage; keep spiky and experimental AI on demand or spot.

Underused commitment discounts are a cause of rising waste
OfficialFlexera blog
03Shop GPU capacity across providers

GPU prices differ widely by provider and model; neoclouds and spot can be far cheaper.

Prices for the same GPU class differ widely by provider and purchase model (on demand, reserved, spot)
JudgementThe Procurement Lens
04Negotiate an enterprise discount programme

Commit total annual spend in exchange for a % discount, credits and migration support.

Providers' backlogs of $561-677bn show demand for committed deals
05Use cheaper chips where they fit

Arm-based instances (Graviton) and custom AI chips (TPU, Trainium) often cost less per unit of work.

AWS: backlog for prior Graviton instances doubled after Graviton 5 launch

How to buy it

Sourcing routeOne primary provider plus a second for resilience or AI capacity; marketplace for software.
Pricing modelEnterprise commit for baseline; savings plans; on demand and spot for variable and AI.
Contract termsCommit flexibility (roll-over, cross-service), egress terms, data residency, exit and portability.

KPIs to track

For the monthly review

Waste as % of spendCommitment utilisationUnit cost (per transaction, per user, per model call)GPU utilisationShare of spend tagged to an owner

Where most buyers are, and what good looks like

Most buyers todayBest practice
VisibilityMonthly invoice→Every resource tagged to an owner
CommitmentsBought once a year→Managed continuously, matched to baseline
AIPooled in IaaS line→AI cost forecast and owned separately
ArchitectureLift and shift→Right-sized, scheduled, cheaper chips
ContractStandard terms→Commit flexibility, egress and exit terms

What AI changes

More in the AI disruption tab.

AI demand has turned cloud into a capacity war; GPU prices are rising.
AI workloads are pushing waste up for the first time in five years.
AI FinOps tools automate right-sizing and commitment management.

Full written analysis

Written live for Cloud & hosting and your region, grounded on the figures above. Public data only.

Written live for this category and region when the page opens in a browser.

Sources

Figures checked 16 Nov 2026. Tags: Official = published by the source; Derived = calculated from published figures; Judgement = The Procurement Lens view.