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.
- Production and non-production workloads were not consistently separated.
- Active and decommissioned workloads were retained similarly.
- Historical data requiring long-term retention stayed in premium backup storage.
- Lower-value retained data was not routinely reassessed for archival suitability.
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 state | Appropriate question |
|---|---|
| Active production VM | Does it have the right backup policy and recovery objective? |
| Active non-production VM | Does it need backup, and at what retention level? |
| Decommissioned production VM | Is long-term retention required and what recovery speed is needed? |
| Decommissioned non-production VM | Is retention required at all? |
| Historical retained data | Can 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.
- VMs still connected to Recovery Services vaults
- Decommissioned VMs with retained recovery points
- Inconsistent backup policies and retention settings
- Production workloads requiring genuine long-term retention
- Non-production workloads that did not justify expensive long-term retention
- Historical backup data suitable for archival storage
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.
| Issue | Impact |
|---|---|
| Decommissioned VMs retained in vaults | High long-term storage cost |
| Inconsistent policies | Unclear retention standards |
| Production and non-production treated too similarly | Over-retention of lower-value workloads |
| No repeatable archive migration process | Accumulating cost over time |
| Backup coverage gaps | Some 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.
| Area | Action |
|---|---|
| Policy design | Refined and standardised backup retention policies |
| Workload classification | Separated production, non-production, active, and decommissioned workloads |
| Retention validation | Confirmed which workloads genuinely required long-term retention |
| Archive migration | Developed a controlled process to move eligible historical VM data into archive storage |
| Lifecycle controls | Implemented retention-aligned lifecycle deletion policies |
| Recovery assurance | Tested archive rehydration and confirmed business acceptance of up to 24-hour restore timing |
| Security hardening | Restricted access and added accidental deletion safeguards |
| Governance | Introduced controls and alerting for production VMs not connected to backup |
| Team enablement | Trained operations teams on future decommissioned workload handling |
Business Outcome
The remediation delivered a six-figure annual saving while preserving retention obligations and improving governance.
| Area | Outcome |
|---|---|
| Annual saving | Six-figure reduction in backup storage cost |
| Data migrated | Approximately 200 TB of eligible historical VM backup data |
| Cost impact | Approximately 95% storage cost reduction for migrated data |
| Retention | Seven-year requirements preserved where required |
| Recovery | Archive rehydration tested and accepted for long-term retention use cases |
| Governance | Production backup coverage gaps identified and addressed |
| Operational maturity | Repeatable 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 mindset | Better 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.