Problem
Cloud bills grow faster than revenue. No visibility into per-feature cost. Engineers don't see cost impact of decisions.
economics-evolutionemerging · high operational burden
Cloud Cost Optimization
Also known as: finops, cloud-cost, cost-management, unit-economics
Intent
Continuously reduce cloud spend while maintaining performance and reliability, using unit economics and architectural choices.
Forces
- •Cloud spend is 2nd largest OpEx (after people)
- •Engineers make architectural decisions daily that affect cost
- •Finance sees bill, engineering makes decisions — disconnect
- •Optimization must not compromise reliability or velocity
Solution
✓ When to Use
- •Cloud spend > $10k/month
- •Finance asking for cost accountability
- •Architectural decisions need cost input
✗ When Not to Use
- •Pre-revenue (optimize for velocity)
- •Tiny spend (<$1k/month)
- •Team not ready for cost visibility
Pros
- +30-50% savings typical without architecture change
- +Alignment: engineering ↔ finance ↔ product
- +Architectural discipline: cost as design constraint
- +Predictability: budget vs actual, no surprises
Cons
- −Operational overhead: tagging, dashboards, reviews
- −Over-optimization: reliability risk, velocity loss
- −Tagging discipline: garbage in, garbage out
- −Shared costs: data transfer, support, enterprise agreements
Cost Profile
Infrastructure
Low — tagging, dashboards, alerts
Operational
High — monthly reviews, optimization sprints
Cognitive
Medium — unit economics, commitment planning
Failure Modes
Right-sizing too aggressive → OOM, latency
Spot interruption → batch job fails, no fallback
Tagging gaps → unallocated spend → no accountability
Savings Plans unused → wasted commitment
Cost focus → reliability neglected → outage
Real-World Examples
Alternatives
- over-provisioning
- reactive-cost-cutting
- finance-only
Related Patterns
- capacity-planning
- unit-economics
- right-sizing
- spot-instances
- serverless
- finops
Competency Domains
economics evolutionreliability opsscalingdeploymentdata state