Most organisations have a monthly Azure cost review. Someone from IT presents the bill. A few line items get questioned. Nobody makes a decision. The meeting ends, the bill gets paid, and the same conversation happens next month.
This is not a cost review — it is a cost report. A report produces acknowledgement. A review produces decisions. The distinction matters enormously, because Azure spend does not self-correct. Only decisions change it.
What a Bad Cost Review Looks Like
A cost review structured around line items is structured around the wrong thing. Line items — compute, storage, networking, licences — are Azure’s categories, not the organisation’s. They tell you what Azure charged, not whether it was worth charging.
The other markers of a review that will not produce decisions: it is IT-led rather than architecture-led, so the person presenting cannot speak to business context; there is no baseline or trend, so a number cannot be judged good or bad; and no decisions are made — only observations, which are noted and forgotten.
The output of this kind of review is a shared awareness that money was spent. That awareness does not reduce spend by a single pound.
What a Good Cost Review Looks Like
A cost review that produces decisions is structured around business outcomes, not Azure service categories.
The difference in practice: instead of “virtual machines cost £8,400 this month, up 4% on last month,” the framing is “the customer portal cost £12,000 to run last month, serving 84,000 active sessions — that is £0.14 per session.” One of those statements invites a question about a service category that most people in the room cannot evaluate. The other invites a question about business value that everyone in the room can evaluate.
Three metrics belong in every cost review. Total spend versus budget — not just the number, but whether the variance is expected and explained. Cost per unit of business output — transactions, sessions, customers served, API calls — whatever makes spend legible to a non-technical stakeholder. And reservation coverage percentage, which surfaces how much of stable compute is still on pay-as-you-go pricing when it should not be.
Anomalies should be flagged with context, not just flagged. A 20% spike in storage costs is not useful information without knowing whether a new data pipeline went live, a retention policy expired, or something is accumulating unexpectedly.
And every review should end with one decision. Not a list of things to investigate. One decision — to commit, to scale down, to sunset something, to approve a reservation purchase. A review that ends without a decision has produced nothing.
Who Should Be in the Room
The people in the room determine what decisions are possible. The wrong room makes the right decisions impossible.
The architect who owns the platform needs to be present — not an IT manager, not a delivery lead, but the person who understands why the workload is designed the way it is and what it would take to change. Without that person, cost observations cannot be connected to architecture changes, and architecture changes are how costs actually move.
A business stakeholder who can connect cost to business value. This person’s role is not to approve spending — it is to say whether the business value the workload is delivering justifies the cost being reported. That judgement cannot be made from inside IT.
Someone with budget authority who can approve a change in the room, not in a follow-up email. The single most common reason cost reviews produce no decisions is that the person who could approve a change is not present to make it.
Finance is advisory in this meeting, not the chair. Finance provides budget context and variance analysis. The decisions are made by the people who own the architecture and the business outcomes.
In my experience running cost reviews that produced decisions, the change that made the most difference was replacing IT operations in the chair with the platform architect. Operations can tell you what the bill says. Architecture can tell you why it says that, what the options are, and what a change would cost in both directions — financially and technically. That combination is what makes a decision possible.
The monthly cost review is the mechanism by which Azure spend becomes a managed business asset rather than an unmanaged business liability. The structure of that meeting determines which one it is.
Structuring a cost governance process that actually produces decisions is something I help organisations do. Get in touch to talk through your current approach.
