FinOps crash course for engineers who hate FinOps
FinOps has an image problem. For many engineers, it conjures images of spreadsheet reviews, finger-pointing, and meetings about who owes whom $400 in S3 bills. It doesn't have to be that. The core of FinOps is three things, and if you do them, you'll capture 80% of the value.
1. Tagging that actually holds up
Every resource gets at least three tags: Owner, Environment, and CostCenter (or Application). Tag coverage below 80% means your cost allocation is fiction. If engineers hand-tag resources, compliance tops out around 40% — the only way to get to 80%+ is enforcement at the policy level (SCPs in AWS, Azure Policy, Org Policy in GCP).
2. Showback, not chargeback
Showback means every team sees their cost attribution in a dashboard. Chargeback means finance actually moves money between budgets based on cloud usage. Showback is a behavior change lever; chargeback is a political quagmire. Start with showback. Most teams will self-correct within two months once they see their own spend.
3. Three recurring rituals
- Monthly review (30 min): top 5 cost drivers, top 3 savings opportunities, one committed action.
- Quarterly commitment review (60 min): what to reserve, what to keep on-demand, what to unwind.
- Annual policy review: tagging rules, approved services, budget guardrails.
That's it. Do those and you have a FinOps practice. Nothing else is optional, but nothing else is strictly required either.
Tools that help
- Monthly Cloud Forecast — project next month from your current run rate.
- RI Optimizer — decide what fraction to commit.
- Idle Resource Finder — quick-win waste detection.
You don't need a FinOps team to start. You need one engineer with 2 hours a week and executive air cover.