
If you run Azure as an administrator, or sit on a FinOps, ITOps, or DevOps team, the practical question is this: what does the workbook reveal, and which cost, governance, and operational problems need more than a workbook to solve? Cloud waste is a real stake. Flexera's 2026 State of the Cloud report puts estimated wasted cloud spend at 29%.
Below, we cover the workbook's overview, its rate and usage optimization areas, and the gaps around automation, workload context, governance, cross-cloud visibility, and live block storage.
Key Takeaways
- The workbook centralizes Azure cost and recommendation data in one interactive view.
- Rate and usage findings are candidates, not approved savings.
- It works best as a diagnostic and prioritization tool, not a replacement for FinOps ownership.
- Dynamic block storage environments may need continuous optimization on top.
What the Azure Cost Optimization Workbook Shows
The workbook lives in Azure Advisor under Workbooks > Cost Optimization (Preview). Microsoft describes it as a source of cost-relevant insights aligned with the Well-Architected Framework's Cost Optimization pillar. A separate, editable edition ships with the FinOps Toolkit as the Optimization workbook. Confirm the current name, navigation path, and preview status in Microsoft's documentation before you rely on either.
Overview and environment visibility
Microsoft documents filters for subscription, resource group, and tag. Its 2024 launch article also describes location and service as analysis dimensions. That lets teams move from an account-wide view down to a specific cost center or resource group.
The overview covers goals and prerequisites, resources by region, and security and reliability recommendation queries. Don't assume it includes budgets, forecasts, or anomaly detection. Those aren't documented as overview features, and older Toolkit releases differ from the current Advisor template.
Rate optimization
Rate optimization lowers the price you pay for something you already use. Usage optimization changes what you consume. The workbook surfaces three main levers:
- Azure Hybrid Benefit: Eligible Windows Server and SQL Server licenses with Software Assurance, plus eligible RHEL and SLES subscriptions—an entitlement, not a term commitment
- Azure Reservations: 1- or 3-year commitments that suit stable, continuously running workloads.
- Savings Plan for Compute: a fixed hourly commitment for 1 or 3 years that flexes across eligible services and regions.
Before you commit, check workload stability, expected duration, eligibility, and utilization. Unused savings plan commitment expires, and purchases can't be canceled or refunded. The workbook offers reservation and savings plan lookbacks of 7, 30, or 60 days, with 1- or 3-year terms. You may still need to complete the purchase in another Azure experience.
Usage and workload optimization
The usage side covers compute, storage, networking, databases, sustainability, and an "Other popular Azure services" category. Examples include:
- Idle resources and improperly deallocated VMs (a stopped VM is still allocated and billed)
- Unattached managed disks and backup storage redundancy
- Unattached public IPs and virtual network gateways with no connections
- SQL pools with zero databases (Toolkit v0.5)
Drill-downs, filters, and links to recommendations help you build a shortlist. Treat it as a shortlist. VM recommendations default to a 7-day utilization window (configurable up to 90 days). Microsoft's own caveats cover upcoming traffic, testing, disaster recovery, and metrics outside the model. Rightsizing considers CPU, memory, and outbound network, while shutdown recommendations omit memory.

Version and data-quality caveat
Tabs, queries, recommendation logic, and preview labels change. Toolkit v0.4 added reservation break-even recommendations and idle ExpressRoute identification. v0.5 added processor-architecture insights and powered-off VMs with Premium Disks. Before you act on a finding, confirm the workbook version, query definitions, refresh behavior, and pricing assumptions.
What the Azure Cost Optimization Workbook Misses
The workbook shows evidence of an opportunity. It doesn't know your application, business process, performance objective, or risk tolerance.
Point-in-time visibility versus continuous optimization
A recommendation reflects data from a particular window, not a control loop that keeps enforcing a decision. Timing matters:
- Cost Management data typically arrives within 8-24 hours for EA and MCA, and up to 72 hours for pay-as-you-go.
- Advisor checks roughly every 24 hours.
- VM recommendations can take up to 48 hours to update after a lookback change.
- Resource Graph is not strongly consistent, and results depend on viewer permissions.

Seasonality, new deployments, AI workloads, and disaster-recovery events all change the right answer. For fast-moving environments, validate freshness before treating a finding as current.
Limited workload and business context
Resizing a VM, shutting it down, or changing redundancy can break an SLA, a production peak, a recovery objective, or a regulated workload. A commitment can strand cost if an expansion is planned.
Pair workbook findings with utilization metrics, application ownership, performance telemetry, data-residency requirements, and business criticality. Cost cuts that cause an outage aren't savings.
Governance, accountability, and unit economics
A workbook view doesn't give you showback or chargeback, cost per customer or transaction, policy enforcement, approval routing, or escalation for overdue actions. Those come from a FinOps practice.
The FinOps Foundation frames examples such as cost per tenant, transaction, and case resolved. Native calculation of those denominators isn't a documented workbook capability.
What turns a list into governance:
- Accounting tags for cost centers and ownership tags for responsible teams
- Resource-group budget alerts assigned to workload teams
- Azure Policy enforcing mandatory tags
- A documented exception process
Recommendation does not equal remediation
Viewing, accepting, executing, and validating are four different steps. Some queries offer Quick Fix. The workbook doesn't include an end-to-end workflow for ownership, approval, change management, and realized savings. Dismissing a recommendation in Advisor removes it from view without touching the resource.

Production compute, networking, databases, and storage need guardrails, testing, rollback plans, and post-change measurement.
Azure scope and multi-cloud limitations
The workbook analyzes your Azure environment. It isn't a unified view of AWS, Google Cloud, or on-premises infrastructure.
Enterprises on several providers need a separate allocation and normalization layer, often built on the FinOps FOCUS specification, which standardizes billing data across sources. FOCUS ingestion isn't a supported workbook capability.
Comparing equivalent services, commitments, regions, and cost categories across providers falls apart without consistent tagging, ownership, and unit-cost definitions.
The storage gap: identifying waste versus continuously optimizing live block storage
Finding an unattached disk is cleanup. Continuously matching live block storage to changing demand is a different job. Azure doesn't support shrinking an existing managed disk, and attempting it risks data loss. Provisioned-disk charges don't vary with space used, and detaching a disk doesn't delete it, so charges continue until deletion.
That's where Lucidity fits, as an optional complement rather than a replacement. Its AutoScaler is an autonomous, application-agnostic layer that expands and shrinks block storage in real time with zero downtime and no code changes.
Lucidity states it can cut block storage costs by up to 70%. It also monitors every volume across AWS, Azure, and Google Cloud. For enterprises with dynamic or AI-heavy storage demand, that covers what a periodic workbook review can't.

How to Use the Workbook Effectively
- Define scope. Pick subscriptions and a time period, confirm tagging and ownership data, and write down the outcome the review should improve.
- Triage by more than savings. Rank by impact, confidence, risk, reversibility, and effort. Separate quick cleanup from architecture or procurement decisions. Microsoft's VM guidance notes that estimates use retail rates and exclude discounts, so real savings may differ.
- Validate each action. Check utilization telemetry, availability objectives, application calendars, licensing eligibility, commitment assumptions, and DR requirements.
- Assign an owner. Record the action, approval, date, expected result, exception reason, and rollback plan in your FinOps or ITSM workflow.
- Run a recurring loop. Review refreshed findings, execute approved changes, measure actual impact, close or defer items, and update policy when the same waste returns.
- Add specialized tooling where needed. Use monitoring or automation when a recommendation depends on continuous behavior, cross-cloud normalization, or safe changes to live storage.
Here's how layering works in practice. The workbook flags a set of oversized disks during the monthly review. Left alone, someone files tickets and the disks stay oversized for weeks.
With a tool like Lucidity Lumen, each disk's tier is scored continuously on IOPS, throughput, latency, and cost. Tier changes can be scheduled, run in one click, or rolled back. The workbook finds candidates. The automation keeps the result from decaying.
Also remember the FinOps Foundation's warning: rate and usage savings aren't automatically additive. Commitments can cover resources that usage optimization later changes, so sequence the two deliberately.
Conclusion
The Azure Cost Optimization Workbook is valuable for finding Azure cost patterns, rate opportunities, usage inefficiencies, and candidate actions. It gives teams a shared starting point.
Treat it as a visibility and prioritization layer inside a broader FinOps process:
- Business context
- Governance
- Execution controls
- Measurement
- Specialized automation where the workbook's coverage ends
Frequently Asked Questions
What is the Azure Cost Optimization Workbook?
It's an Azure Advisor workbook that presents cost information and optimization recommendations across your Azure environment. Exact tabs and queries vary by version.
Where can I find the Azure Cost Optimization Workbook?
Look in Azure Advisor under Workbooks, where it's labeled Cost Optimization (Preview). An editable version ships with the FinOps Toolkit. Verify the current path in Microsoft's documentation.
What is the difference between rate optimization and usage optimization in Azure?
Rate optimization lowers the price paid through commitments, licensing benefits, and discounts. Usage optimization improves how efficiently resources are deployed and consumed.
Can the Azure Cost Optimization Workbook automatically apply every recommendation?
No. It mainly surfaces and helps investigate opportunities. Some queries offer Quick Fix, but most actions need separate workflows, permissions, approvals, scripts, or infrastructure changes.
Does the workbook provide real-time Azure cost optimization?
Don't assume so. Data refresh depends on the underlying sources, with delays from hours to days. A workbook view doesn't continuously change resources or enforce savings.
Is the Azure Cost Optimization Workbook enough for optimizing Azure storage costs?
It highlights common storage opportunities such as unattached disks. Complex environments also need workload telemetry, governance, and continuous block storage optimization to balance cost, performance, and availability.