The finance team sees a cloud bill that grew 40% year over year and asks IT to explain it. IT pulls up a cost dashboard and can attribute maybe 60% of the spend to a specific team or product. The other 40% sits in a shared account, tagged inconsistently or not at all. That gap — not the per-hour compute rate — is where most mid-market cloud cost problems actually live.
It's a Governance Problem Wearing a Pricing Costume
Teams ask us to help "reduce cloud costs" and expect a rightsizing exercise. Rightsizing helps, but it's a one-time fix on top of a structural problem: without consistent tagging and ownership, nobody can see which team, product, or environment is driving spend, so nobody is accountable for it. You can't govern what you can't attribute.
The FinOps Fundamentals That Actually Move the Number
- Mandatory tagging enforced at the policy level — not a wiki page asking engineers nicely. AWS Service Control Policies and Azure Policy can both reject untagged resource creation outright.
- Showback before chargeback — give teams visibility into what they're spending before you try to bill them for it internally. Chargeback without prior visibility just breeds resentment and gaming.
- Reserved instances and savings plans reviewed quarterly, not set once — usage patterns shift, and commitment coverage that was optimal a year ago frequently isn't anymore.
- Automated rightsizing recommendations reviewed on a cadence, not a one-time consulting exercise that decays within two quarters.
- Non-production environment schedules — dev/test environments running 24/7 when they're used 40 hours a week is one of the most common and easiest-to-fix waste sources we find.
The dashboard tells you what you spent. Governance is what determines whether next quarter's number is different.
Where This Intersects With Your AWS-vs-Azure Decision
Cost governance tooling looks different on each platform — AWS Cost Explorer and Azure Cost Management + Billing solve the same problem with different depth in different areas — but the discipline is identical: attribution, accountability, and a recurring review cadence, not a platform feature. See our AWS vs Azure comparison for how the platforms differ structurally.
Before commissioning a rightsizing audit, fix tagging enforcement — an audit's recommendations decay within a quarter if there's no ownership structure to sustain them. Establish showback reporting first, get teams looking at their own spend regularly, and treat commitment coverage (reserved instances/savings plans) as a quarterly review, not a set-and-forget purchase.