Store disk snapshots on standard storage

This is one of the cleanest savings on the Azure bill because the premium tier buys you nothing for a snapshot. A snapshot is cold storage of a backup. It is never in the hot path of a running workload, so the performance you pay for at the premium rate is performance the snapshot will never use. Moving the same data to standard storage cuts the per-GB cost by roughly 60%, which Microsoft states directly in the Advisor recommendation.

Frame it as recurring spend, not a one-off. Snapshot storage is billed monthly for as long as you keep it, so a fleet of premium-priced snapshots is a fixed cost that recurs every month until someone moves it. The saving compounds: the move is a one-time action, the lower rate applies every month thereafter, and switching new snapshots to incremental means you are billed only for changed data rather than a full copy each time.

The payback is immediate and the risk is close to zero. There is no performance trade because a snapshot is not serving a workload, and the restore you eventually run produces an identical disk whether the snapshot sat on standard or premium. The only thing the premium tier was doing here was costing more.

Priya is the finance partner sitting in on the spend review. The Advisor Cost recommendation 'Use Standard Storage to store Managed Disks snapshots' lands with an estimated monthly saving attached, and the instinct is to check what is actually behind it before counting it. The list shows a handful of OS and data-disk snapshots on Premium_LRS, each billed at the fast-tier per-GB rate for data that is never read except during a restore. There is no performance being bought here, so the saving is genuine: moving the same gigabytes to standard storage cuts the rate by roughly 60%, exactly as Microsoft states.

Priya reframes it from a one-off cleanup into recurring spend. Snapshot storage is billed every month for as long as the snapshot exists, so the premium-priced snapshots are a fixed monthly charge that will keep recurring until someone moves them. The saving is the difference between the premium and standard rate on those gigabytes, every month, indefinitely, plus a second, compounding saving once new snapshots default to incremental and are billed for changed data only. The finance note reads: 'Roughly 60% off this snapshot line, recurring monthly, no performance trade, no maintenance window. This is found money, and the only cost is the engineering time to move them and change the default.'

The cost model here is unusually clean: a roughly 60% saving with no performance trade and no risk to the backup itself. Each premium-priced snapshot is a fixed monthly charge that recurs until it is moved, so the saving is not a one-off rebate but a permanent reduction in run-rate on that line. On a fleet of OS and data-disk snapshots, the monthly difference between the premium and standard rate adds up quickly, and it keeps adding up every month it is left.

The asymmetry runs entirely in your favour. The remediation is engineering time to move the existing snapshots and change one default in the backup automation. Against that, leaving it costs the premium rate every month indefinitely, plus the full-disk billing of any non-incremental snapshots. The finance role is to count the saving as recurring, confirm there is no performance or protection trade being given up, and make sure the default is changed so the charge does not creep back the next time someone writes a backup script.

Azure Advisor has identified backups that are being billed at the premium rate when standard would do the identical job for roughly 60% less. This is not a performance question and not a resilience question. The snapshot protects the same data, restores to the same disk, at a fraction of the monthly cost.

The leadership point is that this is found money with no offsetting risk. The change does not touch any running system, does not reduce protection, and does not need a maintenance window. It is the rare optimisation where the only argument against acting is inertia.

The broader question it raises is governance: how many other cost recommendations is Advisor surfacing that nobody is actioning? A snapshot on the wrong tier is cheap to fix once and cheap to prevent from recurring. The signal worth tracking is whether Advisor's cost findings are being worked at all, not whether this one line is fixed.

After the budget review, the CFO asked the obvious question: are we acting on the cost recommendations Azure itself is giving us, or just collecting them? The snapshot finding was the easiest one to point at. Advisor had flagged that backups were being billed at premium rates when standard would do the identical job for roughly 60% less, and nobody had touched it.

The team treated it as a test of whether the Advisor Cost tab was a live queue or a graveyard. They moved every premium-priced snapshot to standard storage, changed the backup automation so new snapshots are incremental on standard by default, and then set a recurring review of the remaining Advisor cost recommendations. The next time the CFO asked, the answer was specific: 'This finding is closed, the saving is roughly 60% on that line and it recurs every month, and we now work the Advisor cost queue on a schedule rather than letting it pile up.' The shift, from a backlog of ignored recommendations to a worked queue, is the real signal, and the snapshot fix is the cheap, zero-risk place to start.

This is the rare optimisation with a saving and no trade. Roughly 60% off a recurring backup line, with no impact on any running system, no reduction in protection and no maintenance window. Azure Advisor has already done the work of finding it. The only thing standing between the estate and the saving is someone actioning the recommendation.

The wider point for leadership is governance, not this single line. If a saving this clean has been sitting unactioned in the Advisor Cost tab, the question is what else is. The fix is to work the Advisor cost queue on a schedule, and the snapshot recommendation is the cheap, zero-risk place to prove the routine works.