Case Study: 95% Saving Through Azure Backup Retention Redesign

At a glance

  • Client type: Large New Zealand council
  • Problem: Long-term historical VM backup retention in expensive storage tiers.
  • Finding: Retention redesign identified around 200 TB suitable for archive migration.
  • Outcome: 95% storage cost reduction for migrated data and six-figure annual savings.
  • Related service: Cloud cost optimisation / Azure Backup governance

Overview

A large council organisation was using Azure Backup and Recovery Services vaults to protect virtual machine workloads across its Azure environment.

The organisation had a legitimate requirement to retain certain data for up to seven years. Over time, this requirement had been applied too broadly, and decommissioned virtual machines were being retained in backup vaults even when they no longer required expensive vault retention.

CloudQbit reviewed backup policies, separated production and non-production retention requirements, identified eligible historical VM data, and developed a controlled process to migrate retained data into secure Azure archive storage.

The result was a six-figure annual saving, with approximately 200 TB of eligible historical backup data migrated and around a 95% storage cost reduction for the migrated data.

The Challenge

The council's long-term data retention obligation had led to a conservative model where decommissioned VMs were often left in Azure Backup vaults for up to seven years.

That approach felt safe but treated very different workloads as if they had the same recovery and retention needs.

The challenge was not to remove retention. The challenge was to preserve required retention in a controlled and more cost-effective way.

Why the Existing Backup Model Became Expensive

Azure Backup is an excellent fit for active workloads that require managed backup and fast operational recovery. But long-term retention of decommissioned workloads in backup vault storage can become very expensive.

In this case, historical VM data was being retained in a model designed for active recovery scenarios, rather than archival retention needs.

Workload stateAppropriate question
Active production VMDoes it have the right backup policy and recovery objective?
Active non-production VMDoes it need backup, and at what retention level?
Decommissioned production VMIs long-term retention required and what recovery speed is needed?
Decommissioned non-production VMIs retention required at all?
Historical retained dataCan it be archived securely at lower cost?

Investigation Approach

CloudQbit reviewed the Azure Backup estate to identify which workloads were contributing to storage growth and cost.

The review also uncovered governance gaps, including production workloads not fully covered by backup policy, which expanded the work into governance improvement as well as cost optimisation.

Root Cause

The root cause was inconsistent retention design and no repeatable lifecycle process for decommissioned workloads.

IssueImpact
Decommissioned VMs retained in vaultsHigh long-term storage cost
Inconsistent policiesUnclear retention standards
Production and non-production treated too similarlyOver-retention of lower-value workloads
No repeatable archive migration processAccumulating cost over time
Backup coverage gapsSome production workloads lacked expected protection

Solution Approach

CloudQbit delivered a practical backup optimisation and governance redesign focused on two goals: reduce unnecessary storage cost and strengthen backup assurance for production systems.

AreaAction
Policy designRefined and standardised backup retention policies
Workload classificationSeparated production, non-production, active, and decommissioned workloads
Retention validationConfirmed which workloads genuinely required long-term retention
Archive migrationDeveloped a controlled process to move eligible historical VM data into archive storage
Lifecycle controlsImplemented retention-aligned lifecycle deletion policies
Recovery assuranceTested archive rehydration and confirmed business acceptance of up to 24-hour restore timing
Security hardeningRestricted access and added accidental deletion safeguards
GovernanceIntroduced controls and alerting for production VMs not connected to backup
Team enablementTrained operations teams on future decommissioned workload handling

Business Outcome

The remediation delivered a six-figure annual saving while preserving retention obligations and improving governance.

AreaOutcome
Annual savingSix-figure reduction in backup storage cost
Data migratedApproximately 200 TB of eligible historical VM backup data
Cost impactApproximately 95% storage cost reduction for migrated data
RetentionSeven-year requirements preserved where required
RecoveryArchive rehydration tested and accepted for long-term retention use cases
GovernanceProduction backup coverage gaps identified and addressed
Operational maturityRepeatable decommissioning and archival process introduced

The Mindset Shift

The key lesson was to move from one-size-fits-all retention to lifecycle-aware retention design.

Old mindsetBetter mindset
"Keep decommissioned VMs in backup vaults for seven years.""Retain required data in the right tier with the right recovery expectation."
"Long retention means premium backup storage.""Long retention can be met with secure archive when rapid restore is not needed."
"Backup setup is a one-time activity.""Backup coverage needs policy, monitoring, and governance."

Conclusion

This case study demonstrates how backup cost optimisation can deliver substantial financial outcomes while improving operational controls.

By redesigning retention policies, migrating eligible historical data to archive, hardening storage, validating recovery expectations, and introducing governance controls, CloudQbit helped the organisation reduce cost, preserve obligations, and strengthen backup discipline at the same time.

Long-term retention does not have to mean long-term premium storage cost.