We get the same question from almost every prospect: "What does the assessment actually involve?" Fair question. You're sharing cost data for your Azure estate and you want to know what happens with it. Here's the straightforward answer.
More Than an Azure Advisor Screenshot
If all we did was pull up Azure Advisor and forward the recommendations, we wouldn't be worth your time. Advisor is a useful starting point, but it only scratches the surface. It misses pricing model mismatches, over-provisioned security tooling, redundancy overkill, and dozens of other patterns that only show up when someone with FinOps experience looks at the numbers in context.
Our assessment is a manual analysis of your Azure estate. It uses real expertise and produces recommendations specific to your environment — not generic best practices from a Microsoft Learn article.
What We Need From You
Read-only access to your Azure cost data. That's it.
Either a Cost Management export (CSV or via the portal) or a Reader role on the relevant subscriptions. We don't need write access or contributor permissions. We don't make any changes to your environment. Zero risk to your running infrastructure.
We typically ask for three to six months of data — enough to spot trends, seasonal patterns, and anomalies.
What We Cover
Every assessment covers ten areas across your Azure estate:
-
Cost concentration — where 80% of your spend actually sits, and where concentration risk could wipe out savings made elsewhere.
-
Compute — VM sizing versus actual utilisation, instances running 24/7 that are only used during business hours, missing licence benefits, and candidates for reservations or savings plans.
-
Storage — tier assignments, redundancy levels that don't match the data's criticality, and orphaned disks and snapshots that nobody remembers creating.
-
Backup & DR — retention policies kept longer than needed, forgotten replicas for decommissioned workloads, and overlapping backup policies protecting the same data twice.
-
Security tooling — whether protection levels are proportionate to what they're guarding. Non-production environments rarely need the same security posture as production, but they often have identical configurations.
-
Networking — circuit utilisation, firewall tier appropriateness, peering traffic, and egress patterns. Networking costs are often the least well-understood part of the bill.
-
PaaS services — tier selection, pricing model mismatches, and plan sizing checked against actual utilisation.
-
Monitoring & logging — ingestion volumes that grow without anyone noticing, sampling rates, and workspace sprawl.
-
Governance — tagging coverage, orphaned resources with no clear owner, and whether budget alerts are actually configured.
-
Licensing — hybrid benefits applied where they should be, dev/test pricing for non-production workloads, and licence entitlements that could offset Azure costs.
What You Get Back
The scope of the report depends on the engagement, but typically includes:
- Prioritised optimisation opportunities with estimated annual savings. Biggest items first.
- Confidence levels for each recommendation — high, medium, or low — so you know what's certain versus what needs further validation.
- Implementation complexity — some changes take five minutes in the portal, others need a maintenance window.
- Quick wins separated from strategic changes — what you can action this week versus what needs planning.
What We Typically Find
The typical result is 20 to 40 percent in optimisation opportunities. That's consistent with industry data — Flexera's State of the Cloud report puts average cloud waste at around 32 percent.
The exact number depends on how mature your existing cost management is. Organisations that have never done a formal review tend to be at the higher end. Those that have done some work but not systematically are usually in the 20 to 30 percent range.
Want to find out where your Azure spend is actually going? Get a free FinOps assessment — read-only access, no obligation.