TL
The short answer

The monthly spend review is the operating rhythm of a FinOps program, and most of them fail in the same way: they present numbers, agree that costs went up, and adjourn without a single owned action. A review that works inverts that. It starts from cost allocated to the teams that own it, compares actual against forecast, and spends its time on the few movements that matter, ending each with a decision, an owner, and a due date. The output is not a report; it is a short list of commitments that the next meeting checks. Run this way, the review becomes the place where waste is caught and acted on every month, rather than a recurring update that nobody acts on.

Here is the agenda, who is in the room, and how each item closes.

Why do most spend reviews fail?

Three failures recur. The first is unallocated cost: if the bill cannot be read by team and workload, the meeting can only discuss a total, and a total has no owner. The second is reporting without deciding, where the meeting narrates variance but assigns no action, so the same variance returns next month. The third is the wrong room, where finance reviews numbers with no engineer present who can actually change the cost. Fix these and the review starts working: allocate first, decide always, and put the people who own the resources in the room. This is the operating discipline the FinOps operating model guide is built around.

Who should be in the room?

  • A FinOps lead who owns the agenda, the allocation, and the action log, and who chases last month's commitments.
  • Engineering or platform owners for the teams whose cost is being reviewed, because they are the only people who can act on a decision.
  • Finance to connect spend to budget and forecast and to hold the unit economics view.
  • A product or business owner when a cost movement reflects a product decision rather than an engineering one.

Keep it small. A review that works is a decision meeting, not a broadcast, so the people present are the ones who can commit to an action on the spot.

What is the agenda that produces decisions?

A tight, repeatable agenda is the whole trick. Open with allocated actual against forecast at the top line, then by team, so the room sees where reality diverged from plan. Walk the three or four largest variances only, not the whole bill, because attention is the scarce resource. For each, name the cause, decide the action, assign an owner, and set a date. Reserve a few minutes for commitment coverage and utilisation, so reservations and Savings Plans stay matched to demand. Close by reviewing last month's actions: done, in progress, or slipped. The slipped ones are the most important line in the meeting.

The agenda that works

1. Allocated actual vs forecast, top line then by team. 2. The three or four biggest variances, each closed with an owner and a date. 3. Commitment coverage and utilisation check. 4. Review of last month's actions: done, in progress, or slipped. Total time, 45 minutes.

How do you make decisions stick?

A decision sticks when it is written, owned, and revisited. Keep a single action log with the variance, the decision, the owner, and the date, and open every review with the previous month's log. An action that slips twice is escalated, not silently rolled forward. Tie the log to showback so teams see their own cost trend between meetings and arrive already aware of their movements. The discipline is unglamorous and it is exactly what separates a review that reduces the bill from one that documents it rising. Over a few cycles the action log becomes the record of a program that is actually changing behaviour.

A worked example

Worked example

A Fortune 500 retailer held a monthly cloud cost meeting that reviewed a single total with no allocation and no actions, and spend rose every quarter. We introduced cost allocated by team, a fixed agenda that walked only the largest variances, and an action log opened at the top of each meeting. Engineering owners joined finance in the room so decisions could be made on the spot, and commitment coverage became a standing item. Within a few cycles the review was closing each variance with an owner and a date, slipped actions were escalated rather than buried, and the cost trend turned. The cadence underpinned the wider program that left the company materially lighter on cloud spend. Figures are verified against billing data and anonymised.

Frequently asked questions

What makes a monthly cloud spend review effective?
Allocated cost so every movement has an owner, a tight agenda that walks only the few biggest variances, the engineers who can act in the room, and an action log that closes each variance with an owner and a date and is reviewed the following month.
Who should attend a cloud cost review?
A FinOps lead who owns the agenda and action log, engineering or platform owners for the teams under review, finance to connect cost to budget, and a product owner when a movement reflects a product decision. Keep it small enough to make decisions.
How long should a spend review meeting be?
About 45 minutes. Open with allocated actual against forecast, walk the three or four largest variances and close each with an owner and date, check commitment coverage, then review last month's actions. Anything longer usually means reporting instead of deciding.

Build a spend review that changes the bill

We help finance and platform teams stand up a monthly cloud spend review that produces owned decisions across AWS, Azure, GCP, and OCI, tied to clean allocation and a real action log. We take zero provider commissions and work on a Fixed Fee or a no risk Gainshare basis, guaranteed: we reduce your cloud spend or we reimburse our service fee. Book a strategy call to design your cadence.

Independent · buyer-side

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.

Buyer-side intelligence, monthly.

The Cloud Spend Navigator: what changed in cloud pricing, commitments, and FinOps — no vendor spin.