The FinOps Foundation describes cloud financial management in three phases: inform, optimize, and operate. Inform is about visibility and allocation, making cloud cost understood and attributed to the teams that drive it. Optimize is about action, reducing spend through rightsizing, waste elimination, storage tiering, commitment coverage, and architecture decisions. Operate is about durability, embedding the practice so savings hold and the cycle repeats as the estate grows. The phases are iterative, not a single pass: a given workload moves from inform to optimize to operate while new workloads keep entering at inform, so most organisations are in all three phases at once. Understanding which phase a workload is in tells you what to do next with it.
Here is what each phase delivers, how to know a workload is ready to move forward, and where teams get stuck.
What does the inform phase deliver?
Inform answers a simple question that is surprisingly hard in practice: what are we spending, and who is responsible for it. The work is visibility and allocation. You bring billing data into one place, ideally standardised so it reads the same across providers, since the FinOps Foundation FOCUS specification standardises billing data across AWS, Azure, GCP, and OCI. You build a tagging strategy that survives real teams, so resources carry the team, environment, and product they belong to. And you allocate cost cleanly, through tags, accounts, or projects, so a leader can see their own number rather than a single undifferentiated bill.
A workload is ready to leave inform when its cost is accurate, attributed, and trusted by the people who own it. The common trap is to rush past this phase. Optimization built on bad allocation chases the wrong targets, because you cannot reduce what you cannot see, and you cannot hold a team accountable for a number they do not believe.
What happens in the optimize phase?
Optimize is where the visible savings come from. With trustworthy data, you act on the levers in order of effort and risk. Rightsizing matches instances and services to real demand. Waste elimination removes idle resources, unattached storage, and forgotten environments. Storage tiering moves cold data to cheaper classes. Commitment coverage applies the biggest lever and the biggest risk: AWS Savings Plans and Reserved Instances, Azure Reservations and the Azure Savings Plan, GCP Committed Use Discounts, and OCI Universal Credits discount roughly 20 to 72 percent against on demand pricing in exchange for utilization risk the buyer carries. Architecture decisions, such as choosing the right service or moving to a more efficient processor family, are the deepest changes and the slowest to land.
The native advisors help here. AWS Compute Optimizer, Azure Advisor, GCP Recommender, and the OCI Cost Analysis console all surface candidates, but they recommend and do not decide, so each recommendation is validated against engineering reality before it is applied. Optimization that breaks production is not optimization.
What does the operate phase add?
Operate is the phase teams skip and then wonder why their savings evaporate. Optimization is a one time event unless something keeps it in place, and cloud estates drift back to waste as new workloads launch, teams change, and yesterday's right size becomes today's oversize. Operate embeds the practice: a regular cost cadence where teams review their spend, budgets and forecasts that finance trusts, anomaly detection that catches a spike within days rather than at month end, and clear ownership so someone is accountable for each cost centre. It is the difference between a project that cut the bill once and a function that keeps it low.
A European SaaS company ran a successful optimization project that cut its bill substantially, then watched the savings erode over the following two quarters as new services launched untagged and idle resources crept back. The missing phase was operate. Reinstating clean allocation from inform, adding a monthly cadence where each team reviewed its own spend, and wiring anomaly detection to catch new waste early held the gains and surfaced fresh savings continuously rather than in one burst. The lesson was that inform, optimize, and operate are a cycle, and skipping operate turns a durable reduction into a temporary one. Figures are verified against billing data and anonymised.
How do the phases work together?
Treat them as a loop, not a ladder. New workloads enter at inform, get attributed and understood, move into optimize where the levers are applied, and then into operate where ownership and cadence keep them efficient. Meanwhile operate feeds back into inform, because a workload that drifts shows up in the cadence and re enters optimization. A mature organisation is running all three phases simultaneously across a portfolio, and its maturity is measured less by how far any one workload has progressed than by how reliably every workload moves through the cycle. That is the practical meaning of a FinOps operating model: the machinery that keeps the loop turning.
Frequently asked questions
What are the three phases of FinOps?
Are the FinOps phases sequential?
Which phase delivers the savings?
Take the full FinOps playbook with you
Our cloud cost optimization playbook lays out the inform, optimize, and operate phases as a working program across AWS, Azure, GCP, and OCI, with the mechanics and decision frameworks for each. We are an independent buyer side advisory with zero provider commissions, and our guarantee is simple: we reduce your cloud spend or we reimburse our service fee.
Put a defensible number on your cloud spend.
No provider in the room, no published price list. Tell us your footprint and we will scope the savings against your billing data — we reduce your cloud spend or we reimburse our service fee.
The Cloud Spend Navigator: what changed in cloud pricing, commitments, and FinOps — no vendor spin.