ITAM Meets FinOps

There is a conversation happening in a lot of enterprises right now that goes roughly like this. The FinOps team is pulling together a cost review and notices a line item for SQL Server licenses running on cloud instances that nobody seems to own. Meanwhile, two floors up, the ITAM (IT Asset Management) team, the function responsible for tracking software licenses, entitlements, and compliance across the organization, is reconciling a separate spreadsheet that shows the same licenses as active on-premises entitlements that nobody is using.

Two teams. Two tools. Two spreadsheets. One problem.

Cloud waste and software license waste are not separate categories of inefficiency. They are the same underlying failure expressed through different billing systems: poor visibility into what is running, what is owned, and whether the two things match.

ITAM tracks technology assets across their full lifecycle, from purchase through deployment, use, and retirement. In practice that means software licenses, entitlements, maintenance contracts, and hardware inventory. The goal is compliance, cost control, and audit readiness. Vendors like Microsoft, Oracle, and SAP audit their customers regularly, and the penalties for being caught short are impactful. According to Flexera's 2025 State of ITAM Report, nearly half of all organizations have paid more than one million dollars in audit penalties over the last three years.

FinOps emerged from a different problem. As organizations scaled their use of AWS, Azure, and Google Cloud, spend became variable, distributed, and difficult to attribute. FinOps built the practices and tooling to bring financial accountability to that environment: cost allocation, tagging, commitment management, showback (showing teams what their cloud usage costs without billing them directly), and chargeback (allocating those costs back to the teams that incurred them).

Both functions are doing versions of the same job. Track what is being consumed, understand what was paid for, close the gap between the two. The reason they have not historically talked to each other is that they grew up in different parts of the organization, use different data sources, speak different vocabulary, and report to different stakeholders. That separation is starting to cost organizations in ways that are difficult to ignore.

Consider what happens when a workload moves to the cloud but the license question is never asked.

A team migrates a SQL Server database to an Azure virtual machine. From FinOps's perspective, this is a compute cost: instance hours, storage, network egress. From ITAM's perspective, this is a licensing event. The SQL Server instance running on a cloud VM may require a different entitlement than the on-premises version, depending on the number of virtual cores and whether the organization is eligible to apply a BYOL (bring your own license) benefit under its existing Microsoft agreement.

Miss that conversation and one of two things happens. Either the organization pays for cloud-native licensing on the Azure bill while the on-premises entitlement sits idle in the ITAM system, neither team aware the other holds half the answer. Or a scaling event quietly creates a compliance violation because BYOL deployments require manual intervention when core counts change, and nobody was watching.

Software running in the cloud can represent up to 25 percent of an organization's total cloud bill, according to Flexera. For an organization spending 20 million dollars annually on public cloud, that is potentially five million dollars in software licensing costs embedded inside the cloud invoice, largely invisible to ITAM because it looks like cloud spend, and only partially visible to FinOps because the licensing context is missing.

Now layer in AI. The same pattern is repeating itself at higher speed and larger scale. Teams are standing up AI workloads using licensed platforms like Databricks, Snowflake, and various foundation model APIs, often through cloud marketplaces, often outside formal procurement channels. ITAM has no visibility because the purchase looked like cloud infrastructure. FinOps has partial visibility but cannot attach licensing terms or entitlement context. The gap that existed for SQL Server now exists for GPU (graphics processing unit) compute, AI platform subscriptions, and inference (the process of running a trained model to generate outputs) costs that scale unpredictably. The financial exposure is the same structural problem, just moving faster.

The more durable version of ITAM and FinOps collaboration looks less like a meeting and more like a shared data model. When entitlement records and cloud billing data can be joined at the resource level, a few things become possible that were not before.

License optimization becomes concrete and specific. If FinOps can see that a cloud instance is running a pay-as-you-go SQL Server license, and ITAM can confirm unused BYOL entitlements are eligible for that workload, the saving is immediate. Flexera data suggests that 32 percent of organizations now report it is the FinOps team, not the ITAM team, generating savings on cloud software licensing. The boundary is already dissolving in practice even where the org chart has not caught up.

Audit defense improves. When cloud resource consumption is traceable to specific license entitlements, and that traceability is documented rather than reconstructed the week before an audit, the compliance posture changes from reactive to defensible.

Procurement gets better data before renewals. When the combined picture of what is owned and what is running is visible before a vendor conversation, the negotiation position is fundamentally different than when procurement is working from either dataset alone.

The organizational boundary between ITAM and FinOps made sense when on-premises software and cloud infrastructure were genuinely separate domains. That separation is eroding faster than most org charts are updating. Software migrates, scales, and gets deployed in ways that change its licensing terms every time it crosses a boundary. AI workloads are accelerating that pattern considerably.

Some teams are starting to notice that the most expensive inefficiencies in their technology spend are not hiding in the cloud bill or the ITAM spreadsheet. They are hiding in the gap between them.

The waste was always there, it just took two teams to make it invisible.