Direct naar content

The hidden costs of Azure Backup

Part 2: The policy nobody revisits, and what it costs

In part 1 of this series, Edco Wallet described the technical cost pitfalls of Azure Backup: storage tiers that cost more than they need to, growing snapshots, and forgotten vault items. In this second part, he goes one layer deeper. Because behind every one of those technical choices sits a policy layer. Or rather: the absence of one.

Edco Wallet

Co-Founder & eigenaar
Edco Wallet - Co-Founder & eigenaar
Verborgen kosten bij Azure Backup deel 2

A retention policy nobody has touched since

At some point in the past, someone decided: “We’ll keep backups for 365 days.” Maybe that was a compliance requirement. Maybe a precaution. Maybe just the default the vendor had set.

That decision is now costing money every single month. 365 days of retention versus 90 days is a factor of four in storage cost, for the exact same data.

What I see in practice: organizations keep backups far longer than their own compliance or recovery needs actually require. The reason? Nobody ever asked the question: “When was the last time we actually needed a backup older than 90 days?”

For most SME+ organizations, the answer is: never. Or, at most, once in the past five years.

Tiered retention: the fix few organizations know about

Azure Backup supports tiered retention: you can keep recent backups on fast storage and automatically let older backups cool down into cheaper tiers (from Vault-Standard to Archive).

The Archive tier costs roughly 80% less than standard vault storage. For backups older than 180 days that you keep for compliance but rarely or never actually need, that’s a serious saving.

And yet, we rarely find this configuration actively in use at the organizations we audit. Not because they don’t want it, but because nobody has ever consciously thought about it.

The backup that’s never been tested

Then there’s the topic that comes up in every FinOps analysis, but rarely gets treated as a cost item: never testing your backups.

At first glance, this sounds like a risk management question. The two are, in fact, inseparable.

A backup that doesn’t work costs you twice. First, you pay for the storage every month. Then, when an incident hits, you pay for the manual recovery effort, the downtime, the consultants flown in to help, and possibly the data that’s simply gone.

There’s also a direct cost to the testing itself: restore operations generate egress costs. When you restore a large VM or database from a backup, especially cross-region, those costs can run into the hundreds of euros per test.

The result? Organizations test less than they should. They avoid the cost of testing, and in doing so, increase their exposure to the far higher cost of a real restore gone wrong.

The quiet double-up: backup and disaster recovery for the same workloads

One specific pattern we see regularly: organizations running both Azure Backup and a separate Disaster Recovery solution (such as Azure Site Recovery) for the same workloads, without that ever having been a deliberate choice.

Backup and DR are different things with different goals. Backup is for restoring data. DR is for keeping services running through a full regional outage. Some systems genuinely need both. For others, one of the two is more than enough.

When both are active for the same VMs (and we find this at multiple clients) you’re paying twice. Not necessarily for nothing, but without any deliberate trade-off ever having been made.

Adding it all up

A recap of the cost pitfalls from this two-part series: the wrong storage tier (GRS versus LRS) accounts for 40-50% of unnecessary backup cost. Excessive snapshot retention: 30-40%. Forgotten vault items: 10-20%. Retention policies kept too long with no tiered archive: an extra 20-40%. Not using the Archive tier for backups older than 180 days: 30-60%. Unwanted overlap between backup and DR: situation-dependent.

In our experience, organizations spending €5,000 to €15,000 a month on Azure Backup save 25 to 40% structurally by working through these points systematically. Without giving up anything on availability or compliance.

Getting control starts with an honest picture

Most of the problems described across this series aren’t the result of bad decisions. They’re decisions that were right at the time, but that got lost in blind spots.

Azure grows. Your organization grows. Your backup configuration grows along with it, but rarely in the right direction. Getting a grip on it starts with an honest picture of what’s running, what it costs, and whether it still matches what your organization actually needs today.

That’s the problem OptimaSure solves. Within 30 days, you’ll know where you stand. Within 31, you’ll start saving.

Ready for an honest picture of your Azure Backup costs? Schedule an Azure Cost Scan. We’ll look at your environment together and present concrete savings.

Other blogs

FinOps blog
Secret Link