Cloud makes it easy to provision resources and just as easy to overspend. FinOps is an operating model that brings engineering, finance and business teams together to make informed trade-offs between speed, cost and quality.
The three phases
The FinOps Foundation describes a cycle of inform (visibility and allocation), optimise (reducing waste and improving rates) and operate (embedding cost into everyday decisions). Teams move through the cycle repeatedly as they mature.
Inform: make cost visible
- Tag or label every resource with owner, application and environment.
- Allocate shared costs fairly so teams see their true spend.
- Publish dashboards and forecasts that engineers actually use.
Optimise: reduce waste
- Shut down idle and orphaned resources; schedule non-production environments.
- Rightsize oversized compute and databases.
- Use commitment-based discounts for steady workloads.
- Review storage tiers and data transfer costs.
Operate: build accountability
Set budgets and anomaly alerts per team, review unit costs such as cost per customer or per transaction, and include cost in architecture reviews.
Tip: unit economics — cost per business outcome — are more meaningful than total spend, which naturally rises as the business grows.
6 proven FinOps best practices
- Enforce tagging. Require tags for owner, application, environment and cost centre, and use policies to block or flag untagged resources.
- Show costs to engineers. Dashboards that show teams their own spending in near real time encourage better decisions than monthly finance reports.
- Rightsize continuously. Review utilisation and downsize or switch instance families for over-provisioned resources.
- Schedule non-production. Shut down development and test environments outside working hours where possible.
- Use commitments wisely. Savings plans and reserved capacity reduce costs for steady workloads; cover a safe baseline and review coverage regularly.
- Track unit economics. Measure cost per customer, transaction or product feature so spending can be compared to business value.
Building a FinOps team
A small central FinOps function typically sets standards, manages commitments and provides tooling, while engineering teams remain responsible for their own spending. Finance, procurement and product leaders all play a role in forecasting and budgeting.
Common mistakes to avoid
- Treating cost optimisation as a one-off project after a large bill.
- Buying long-term commitments before rightsizing.
- Ignoring storage, data transfer and managed service costs.
- Focusing only on savings rather than value delivered.
Frequently asked questions
Does FinOps apply beyond public cloud?
Yes. Many organisations now apply FinOps principles to SaaS, licensing, data platforms and AI workloads.
How quickly can FinOps reduce costs?
Quick wins such as deleting idle resources and scheduling environments often show results within weeks.
A 90-day action plan
Days 1 to 30: enable detailed billing exports, agree a tagging policy and build a dashboard showing spending by team, application and environment.
Days 31 to 60: remove idle resources, schedule non-production environments, rightsize the largest over-provisioned services and set budget alerts.
Days 61 to 90: analyse steady usage for commitment purchases, introduce monthly cost reviews with engineering leads and define one unit cost metric per major product.
Questions engineering teams should ask
- What does this service cost per month, and what drives the bill?
- Which resources run around the clock but are used only during working hours?
- Are we paying for storage, snapshots or logs that nobody uses?
- Would a different instance family, region or storage class be cheaper?
- How will a new feature change our spending at scale?
Key terms explained
- Showback: reporting spending to teams without charging their budgets.
- Chargeback: allocating costs to the budgets of the teams that incur them.
- Rightsizing: matching resource size to actual demand.
- Unit economics: cost related to a business measure, such as cost per order.
- Commitment coverage: the share of usage covered by discounted commitments.
The bottom line
Cloud spending is driven by thousands of daily engineering decisions, so cost control must be shared across the organisation. Visibility through tagging and dashboards, continuous optimisation, smart use of commitments and unit cost metrics help connect spending to business value. Build a small central team to set standards and provide tools, while engineering teams own their consumption. The result is a culture where speed and cost efficiency support each other.
Further reading on FinOps
For authoritative, vendor-neutral guidance on FinOps, see the FinOps Foundation framework. You can also browse our free whitepapers.
