Cloud Cost Review Template (FinOps)

Cloud cost review by FinOps phase: spend by service and team, unit economics, anomalies, savings, commitments, tagging, actions

OperationsStandardFinOps Framework (phases)finops cost review mathUpdated Oct 20, 2026

Open in editorDownload .md

How to use: Use it monthly or quarterly to review cloud spend with engineering and finance. Reconcile the total with the invoice and fill the Inform section before discussing savings. The main pitfall is listing opportunities without owners, estimates and decisions.

Preview

Cloud Cost Review: [Organisation or business unit] [Period]

Periodic review of cloud spend organised around the three FinOps Framework phases: Inform (visibility and allocation), Optimize (rates and usage) and Operate (governance and continuous improvement). Covers spend summary, cost by service and team, unit economics, anomalies, optimisation opportunities, commitments, tagging compliance and actions. Replace every [bracketed] value and every row marked Example.

Field	Value
Review ID	[FCR-2026-10]
Period	[YYYY-MM-DD to YYYY-MM-DD]
Scope	[Accounts, subscriptions or projects included]
Cloud providers	[Provider A], [Provider B]
Currency	[USD], amounts excluding tax, after credits
Data source	[Billing export, cost tool], extracted YYYY-MM-DD
Prepared by	[@finops-analyst]
Reviewed with	[Engineering leads, finance, product]
Status	[Draft / Final]

1. Summary

Metric	This period	Previous period	Change	Budget	Variance to budget
Total cloud spend	Example: 182,400	168,900	+8.0%	175,000	+4.2%
Cost per 1,000 orders	Example: 3.42	3.61	-5.3%	3.50	-2.3%
Commitment coverage	Example: 64%	58%	+6 pts	70% target	-6 pts
Tagged (allocatable) spend	Example: 91%	87%	+4 pts	95% target	-4 pts
Identified savings (annualised)	Example: 245,000	n/a	n/a	n/a	n/a

Key messages: [Example: "Spend grew 8% while orders grew 14%, so unit cost fell 5%. Data platform storage is the main driver of the overrun. The four approved or pending opportunities in section 4 could save about 18,500 per month."]

2. Inform: visibility and allocation

2.1 Spend trend

Month	Compute	Storage	Database	Network	Other	Total
Example: [Month -2]	71,200	31,800	39,500	12,100	9,600	164,200
Example: [Month -1]	72,900	33,400	40,100	12,600	9,900	168,900
Example: [This month]	74,300	41,700	41,200	14,800	10,400	182,400

2.2 Cost by service

Service	This period	Previous	Change	Share of total	Comment
Example: Virtual machines	58,100	57,300	+1.4%	31.9%	Stable
Example: Managed Kubernetes	16,200	15,600	+3.8%	8.9%	Expected growth
Example: Object storage	29,900	22,100	+35.3%	16.4%	Analytics raw data retention
Example: Managed database	41,200	40,100	+2.7%	22.6%	Stable
Example: Data transfer	14,800	12,600	+17.5%	8.1%	Cross-region replication added
[Service]	[Amount]	[Amount]	[%]	[%]	[Comment]

2.3 Cost by team or product

Team / product	Owner	This period	Previous	Change	Budget	Status
Example: Checkout	[@lead]	48,300	45,900	+5.2%	50,000	On track
Example: Data platform	[@lead]	61,700	50,800	+21.5%	52,000	Over budget
Example: Shared platform	[@lead]	55,800	55,200	+1.1%	57,000	On track
Example: Unallocated (untagged)	n/a	16,600	17,000	-2.4%	n/a	Reduce
[Team]	[@lead]	[Amount]	[Amount]	[%]	[Amount]	[Status]

Shared costs are allocated by [method, for example "proportional to each team's direct compute spend"].

2.4 Unit economics

Unit cost links spend to business value:

c_{\text{unit}} = \frac{C_{\text{period}}}{Q_{\text{period}}}

where $C_{\text{period}}$ is allocated cloud cost and $Q_{\text{period}}$ the number of business units delivered (orders, active users, API calls).

Unit metric	Cost	Units	Unit cost	Previous	Change	Target
Example: Cost per 1,000 orders	182,400	53.3 M orders	3.42	3.61	-5.3%	3.50
Example: Cost per monthly active user	182,400	1.92 M	0.095	0.098	-3.1%	0.090
[Metric]	[Cost]	[Units]	[Value]	[Value]	[%]	[Target]

2.5 Anomalies

ID	Date detected	Service / account	Expected	Actual	Cause	Status
AN-01	YYYY-MM-DD	Example: Object storage, analytics account	800 per day	1,450 per day	Retention policy removed during migration	Fixed YYYY-MM-DD
AN-02	YYYY-MM-DD	Example: GPU instances, sandbox	0	2,300 total	Test cluster not deleted	Fixed; budget alert added
[AN-nn]	YYYY-MM-DD	[Service]	[Amount]	[Amount]	[Cause]	[Status]

3. Commitments and reservations

Commitment	Type	Term	Expiry	Monthly value	Utilisation	Coverage of eligible usage
Example: Compute savings commitment	Spend-based	1 year	YYYY-MM-DD	32,000	98%	61%
Example: Database reserved instances	Resource-based	3 years	YYYY-MM-DD	18,500	100%	72%
[Commitment]	[Type]	[Term]	YYYY-MM-DD	[Amount]	[%]	[%]

Effective savings rate across the estate:

\text{ESR} = 1 - \frac{\text{Actual cost of covered and on-demand usage}}{\text{Same usage at on-demand prices}}

Example: ESR this period = 23% (previous 20%). Expiring in the next 90 days: [list], decision required by YYYY-MM-DD.

4. Optimize: opportunities

ID	Opportunity	Lever	Est. monthly saving	Effort	Confidence	Owner	Decision
OP-01	Example: Restore 90-day lifecycle policy on analytics raw data	Usage	7,200	Low	High	[@name]	Approved
OP-02	Example: Rightsize 42 VMs below 15% peak CPU for 30 days	Usage	4,900	Medium	Medium	[@name]	Approved
OP-03	Example: Increase compute commitment coverage from 61% to 75%	Rate	3,800	Low	High	[@name]	Pending finance
OP-04	Example: Schedule non-production clusters off outside 07:00 to 20:00 weekdays	Usage	2,600	Low	High	[@name]	Approved
OP-05	Example: Move cross-region replication to compressed batches	Architecture	1,900	High	Low	[@name]	Investigate
[OP-nn]	[Opportunity]	[Rate / Usage / Architecture]	[Amount]	[L / M / H]	[L / M / H]	[@name]	[Decision]

Total approved savings: [amount] per month, [amount] annualised.

5. Operate: governance and tagging

5.1 Tagging compliance

Required tags: cost-center, team, environment, service.

Account / project	Spend	Spend with all required tags	Compliance	Previous	Owner
Example: prod-core	92,000	89,200	97.0%	95.1%	[@name]
Example: analytics	48,600	41,800	86.0%	79.4%	[@name]
Example: sandbox	9,800	6,100	62.2%	60.0%	[@name]
[Account]	[Amount]	[Amount]	[%]	[%]	[@name]

5.2 Governance controls

Control	Status	Notes
Budgets and alerts per team at 80% and 100%	Example: In place for 9 of 11 teams	Add for two remaining teams
Anomaly detection with daily notification	In place	Threshold 30% above 7-day baseline
Tag policy enforced at resource creation	Partial	Enforced in prod, warning only in sandbox
Sandbox auto-expiry	Not in place	See action A-04
[Control]	[Status]	[Notes]

6. Actions and owners

ID	Action	Related	Owner	Due	Expected impact	Status
A-01	Example: Reapply lifecycle policy and add policy-as-code check	OP-01, AN-01	[@name]	YYYY-MM-DD	-7,200 per month	In progress
A-02	Example: Rightsize VMs in two waves with performance check	OP-02	[@name]	YYYY-MM-DD	-4,900 per month	Not started
A-03	Example: Present commitment purchase to finance	OP-03	[@name]	YYYY-MM-DD	-3,800 per month	Not started
A-04	Example: Enforce tags and auto-expiry in sandbox	5.1, 5.2	[@name]	YYYY-MM-DD	Compliance > 90%	Not started
[A-nn]	[Action]	[Ref]	[@name]	YYYY-MM-DD	[Impact]	[Status]

6.1 Actions from previous review

ID	Action	Outcome	Realised saving
[A-nn]	[Action]	[Done / Partly / Dropped and why]	[Amount per month]

7. Forecast and next review

Item	Value
Forecast next period (before actions)	[Amount]
Forecast next period (after approved actions)	[Amount]
Main assumptions	[Example: "orders grow 4%; no new regions"]
Next review date	YYYY-MM-DD
Attendees	[Names or roles]

Based on the publicly described FinOps Framework phases (Inform, Optimize, Operate), summarised in original wording. This template is original text written for this collection and is not affiliated with the FinOps Foundation. Template licence: MIT.

Related templates

Operations