Skip to content

glossary

FinOps

The practice of bringing financial accountability to cloud spending: giving engineering teams visibility into costs, optimizing usage through rightsizing and commitments, and aligning cloud investment with business value.

In depth

Cloud computing replaced fixed datacenter budgets with on-demand spending that any engineer can increase with one API call, and FinOps is the discipline that manages the consequences. Formalized by the FinOps Foundation, it runs in a loop of three phases: inform, give teams accurate, allocated visibility into what they spend, which requires disciplined tagging and showback or chargeback; optimize, reduce waste by rightsizing over-provisioned resources, deleting idle infrastructure, choosing cheaper storage tiers, using spot instances, and purchasing reserved instances or savings plans for steady workloads; and operate, build continuous processes such as budgets, anomaly alerts, and unit-economics tracking like cost per customer or per transaction. FinOps is deliberately cross-functional, sitting between engineering, finance, and product, because engineers control consumption while finance controls budgets. The goal is not minimal spend but maximal value: sometimes the right decision is spending more on infrastructure that drives revenue, made consciously rather than by accident.

Why it matters

Cloud waste is enormous; industry surveys consistently estimate roughly a third of cloud spend is wasted, and cloud bills are often a company's largest infrastructure cost. FinOps skills let engineers tie technical decisions to business impact, a capability that distinguishes senior engineers and is increasingly a dedicated career path.

Real-world example

example.txt

A SaaS company's AWS bill grows 40% in a quarter with no matching user growth. A FinOps review finds untagged resources, development clusters running nights and weekends, and gp2 volumes never migrated to gp3. After enforcing tagging, scheduling non-production shutdowns, rightsizing, and buying a savings plan for the steady baseline, the bill drops 28% with zero performance impact.

Tools related to FinOps

AWS Cost ExplorerCloudHealthKubecostCloudabilityOpenCostInfracost

Interview questions

  1. Explain the FinOps lifecycle of inform, optimize, and operate.
  2. How would you investigate a sudden 30% increase in a cloud bill?
  3. Compare reserved instances, savings plans, and spot instances. When does each fit?
  4. Why is tagging strategy foundational to FinOps, and how do you enforce it?
  5. What are unit economics in cloud cost, and why are they better than raw spend?
  6. How do you make engineers care about cost without slowing them down?