Snapshot and backup cost optimization means keeping exactly the recovery points your recovery objectives require, on the cheapest storage that still meets them, and deleting everything else. The waste here is rarely a single big object; it is thousands of snapshots accumulating because deletion is nobody's job and retention policies were never tuned. Cleaning it up is low-risk and high-margin once you can see the full picture.
This article is part of our complete guide to cloud storage and data cost optimization, the cluster pillar it links up to. Pruning backup waste is a Cut move under our See, Cut, Lock, Run method, and the broader storage version is in how to audit your cloud storage footprint.
Every retained snapshot should trace to a recovery point objective and a recovery time objective. If a backup does not serve a defined RPO or RTO, you are paying to keep data you have already decided you will never need to restore.
Find the orphans first
The fastest win is orphaned snapshots: point-in-time copies of volumes that no longer exist. When an instance or disk is deleted, its snapshots often survive, billing forever with nothing to restore them to. Inventory all snapshots, match them against live volumes, and flag any whose source is gone. These are almost always safe to delete after a short hold. Orphaned disks and old backups are the related category covered in storage waste: snapshots, orphaned disks, and old backups.
Fix retention before it fixes you
Most backup sprawl comes from retention that was set generously and never revisited: daily snapshots kept for years, hourly copies retained as if they were monthly. Define a tiered retention schedule that matches risk, for example keep recent daily points for a short window, weekly for a medium window, and monthly for the long tail, then automate the pruning so it happens without a human. Manual retention always drifts toward keeping everything.
Want the backup waste cleaned up for you?
Our cloud cost audit inventories every snapshot and backup, flags the orphaned and over-retained copies, and hands you a safe deletion and retention plan tied to your recovery objectives. On the performance model, you pay only from realized savings. No savings, no fee.
Book a cloud cost audit →Use incremental snapshots and understand what they cost
Most cloud snapshot systems are incremental: the first snapshot is full, and later ones store only changed blocks. The cost trap is that deleting an old snapshot does not always free as much as you expect, because later snapshots may depend on its blocks. Understand your provider's snapshot chaining before assuming a deletion saves the full snapshot size, and prefer pruning whole chains over picking individual snapshots out of the middle. High-change-rate volumes generate larger incrementals, so the same retention policy costs more on a busy database disk than on a static one.
Tier old backups to archive storage
Backups you must keep for compliance but will almost never restore belong on archive-class storage, not on the default tier. Moving long-retention backups to a cold or archive tier can cut their storage cost dramatically, at the price of slower and metered retrieval, which is the right trade for data you keep only to satisfy a policy. The decision of when archive pays off is detailed in cold and archive storage: when it pays off, and automating the move by age is in how to build a storage lifecycle policy.
| Waste pattern | Why it costs | Fix |
|---|---|---|
| Orphaned snapshots | Source volume deleted, snapshot remains | Match to live volumes, delete after hold |
| Over-long retention | Daily copies kept for years | Tiered schedule, automated pruning |
| Compliance backups on hot tier | Paying premium for rarely-read data | Move to archive class |
| Duplicate backup tools | Native plus third-party both running | Consolidate on one |
Snapshot chaining behavior and archive-tier retrieval rules differ by provider and change. Verify how AWS, Azure, GCP or OCI bill incremental snapshots and archive retrieval before deleting or tiering, as of May 2026.
Watch the cross-region copy cost
Backups rarely cost only storage. Copying snapshots to a second region for disaster recovery generates cross-region data transfer on every copy, and a daily replication of a large, high-change dataset can make the egress line rival the storage line. Three levers keep that in check. First, match replication frequency to the recovery point objective rather than copying more often out of habit, since a four-hour RPO does not need hourly cross-region copies. Second, compress before the copy where the provider does not already, so fewer bytes cross the boundary. Third, scope what you replicate to the data that genuinely needs regional redundancy, rather than mirroring everything by default. The broader treatment of cross-region transfer cost is in reducing inter-region data transfer costs, and the reason these charges exist at all is explained in data egress charges explained.
The Cloud Storage and Egress Cost Playbook includes the snapshot inventory queries and the retention scoring model we use to rank backup waste by dollars. It is the downloadable companion to this method.
Consolidate the tooling
Many estates pay for backups twice. The cloud provider's native snapshot and backup service runs on a schedule, and a third-party backup product runs alongside it, often protecting the same resources because each was adopted by a different team at a different time. The result is double the storage, double the egress for any cross-region copies, and double the operational surface, for a single recovery requirement. Audit which tool protects which resource, decide on one per workload class based on what your recovery objectives genuinely need, and decommission the duplicate. Where a third-party tool adds real capability, such as application-consistent backups or cross-cloud portability that the native service cannot match, keep it for the workloads that need that capability and let the native service cover the rest, rather than running both everywhere. Consolidation here is pure margin: you remove cost without reducing the protection that anyone actually relies on.
The short version
Delete orphaned snapshots, tie retention to real recovery objectives and automate the pruning, understand incremental chaining before you assume a deletion saves money, and tier long-retention backups to archive. Tie it into the wider footprint review in how to audit your cloud storage footprint. When you want it run across the estate at once, that is what our rightsizing and waste elimination service delivers.