Cloud Waste: When Procurement and IT Don’t Talk
Article contents
The Bill Goes to Someone Who Never Made the Decision
Cloud waste in procurement is a growing problem: according to the Flexera 2026 State of the Cloud Report — based on a sample of 753 professionals worldwide — 29% of global cloud spend on IaaS and PaaS is wasted every year. That figure has edged upward compared to 2025, marking the first increase after five consecutive years of decline. The report attributes this reversal directly to the growing complexity introduced by AI adoption and the proliferation of new PaaS services. The problem is not new, but it is getting harder to manage.
This is the pattern I have observed most consistently working as an ICT Buyer: cloud overspend is not a technical failure. It is an organisational one — the result of a structural separation between those who decide and those who pay.
Two Roles That Never Talk to Each Other
Procurement negotiates. It works on rates, volume discounts, committed use agreements and contractual terms. Its frame of reference is unit cost and forecast volume — it optimises what it can see and what it can measure before the contract is signed.
IT and development teams consume. They provision instances, allocate storage, generate data transfer and spin up ancillary services. Their frame of reference is operational need, not budget. They spend what the work requires — or, more often, what is most convenient to spin up in the moment.
Between these two groups, there is rarely anyone whose job it is to connect the two worlds in real time. The end-of-month cloud invoice surprises no one formally — and yet it almost always differs from what anyone actually expected.
Cloud pricing complexity is not an anomaly — it is the baseline. Dozens of variable line items, rates that differ by region and by service, ancillary costs that are structurally difficult to forecast. The person who signed the contract and the person who lives with it operationally are almost never the same. That is where the gap is created.
Where Unmonitored Cloud Spend Hides
The issue is rarely with primary services — those tend to be visible and tracked. The problem lies in peripheral costs: outbound data transfer, test environments running around the clock, forgotten snapshots, workloads deployed in suboptimal regions. These are the items that drift silently through the billing cycle without anyone owning them.
The same Flexera report shows that 17% of organisations exceeded their cloud budget in the past year — not due to exceptional events, but through continuous drift on cost items that no one was actively monitoring. Meanwhile, 85% of IT professionals identify cloud cost management as their primary challenge, and yet 63% of organisations already have a dedicated FinOps team in place. Having the team is not the problem. The problem is that the team is often disconnected from the people who negotiated the contracts in the first place.
These costs do not appear in contracts because they are variable. They do not appear in budgets because no one estimates them with sufficient granularity. They appear in invoices — at which point no one has the mandate, or the data, to contest them.
Cloud Waste in Procurement: The Insight That Changes the Perspective
The real problem is not that organisations spend too much on cloud. The real problem is that they do not know precisely why they spend what they spend. That distinction matters more than it might appear.
Those who optimise without understanding the underlying causes cut in the wrong place. They reduce instance counts when the actual problem is outbound data transfer. They renegotiate committed use terms when the actual problem is that half their environments should not exist at all. Optimisation without diagnosis is just guesswork at scale.
Those who monitor spend by service, by team and by monthly trend have something concrete to bring to the table — before the renewal, and before the vendor already knows what you are going to confirm. That visibility is not a reporting exercise. It is a negotiating asset.
- Separating the people who negotiate from the people who consume — without a shared monitoring process — produces structural waste, not occasional overspend.
- Cloud costs are not optimised at renewal. They are managed continuously, with granular data broken down by service and time period.
- The Cloud FinOps Tracker I published in Storie di Dati was built precisely for this: visibility across AWS, Azure and GCP, segmented by provider, service and project tag, with monthly actuals-vs-budget comparison.
Those who do not monitor do not negotiate. They simply wait for the invoice to arrive.
Further Reading
- Digital Procurement, Cloud and ICT Data
- How an ICT Buyer Uses Data to Make Better Decisions
- FinOps Foundation — What Is FinOps and How It Works in Practice
- Flexera 2026 State of the Cloud Report
What Is FinOps and What Does It Do for a Business?
FinOps is the practice of financial management applied to cloud spend. Its purpose is to connect those who generate costs — IT and development teams — with those who control them — procurement and finance — creating a continuous process of monitoring and optimisation rather than a retrospective audit.
How Do You Monitor Cloud Spend by Provider and Service?
Effective cloud cost monitoring requires tools that segment spend by provider (AWS, Azure, GCP), by service and by project tag, with a monthly comparison between actuals and budget. Without that level of granularity, optimisation is essentially blind — you are adjusting levers without knowing which ones actually move the needle.
Share this article