We manage roughly 40 Azure subscriptions with annual spend in the mid-seven figures. The environment is mostly PaaS, with two AKS clusters, and tagging is already enforced through policy. Business-unit and environment tags propagate from resource groups, and chargeback data is centralized through an Azure Data Explorer-backed hub. Finance can now see who owns each cost, but ownership alone is not enough.
Azure Advisor keeps reporting the same three underused VMs, even though one has already been resized twice. Meanwhile, more meaningful waste has been found manually: Premium SSDs attached to development systems that shut down every evening, a barely used ExpressRoute circuit, and elastic pools still sized for an old load test. Finding these issues usually takes a couple of people several days, by which point the spend has already happened.
The other problem is follow-through. Someone posts a screenshot of an issue, it gets discussed briefly, and the same item appears in the next monthly review. What tools, dashboards, pipeline checks, or operating processes have you added once tagging and chargeback were already working to identify material savings earlier and make sure they stay fixed?
2 Answers
The biggest improvement was turning findings into tracked work instead of screenshots. Every anomaly gets an owner, estimated monthly impact, a decision deadline, and a follow-up check. The outcome can be resize, schedule, delete, retain for a documented business reason, or investigate further.
Also be careful not to label low utilization as waste automatically. A lightly used ExpressRoute circuit or reserved capacity may still be required for resilience, latency, or an expected workload. Cost tooling can surface the question, but engineers and product owners still have to connect the usage to the business need. Some of our best savings came from reviewing detailed cost lines and finding obscure settings that had been enabled for years without anyone using them.
We added Infracost to our OpenTofu pipeline. It gives developers an estimate for new deployments and makes cost changes visible during review, before anything reaches Azure. It is especially useful for catching oversized disks, databases, and networking changes while the context is still fresh.
It does not solve every runtime issue, though. Savings Plans, reservations, utilization, and resources created outside the pipeline still need separate reporting and ownership workflows.
Related Questions
Can't Load PhpMyadmin On After Server Update
Redirect www to non-www in Apache Conf
How To Check If Your SSL Cert Is SHA 1
Windows TrackPad Gestures