← Blog

How New Zealand Businesses Should Test Backup and Disaster Recovery Plans

Test backups by restoring real data and confirming essential work can resume within an acceptable timeframe. Use risk-based schedules, clear roles and realistic scenarios.

Published 13/11/2026 · EdgePoint IT Solutions

A successful backup report is not proof that your business can recover. Test whether you can restore the right data, in the required order, within a data-loss and downtime window your business can accept. Start with small restore tests, schedule broader recovery exercises, and retest after significant technology or process changes.

Set testing frequency around business impact

There is no single testing interval that suits every New Zealand business. Your schedule should reflect the importance of each system, how quickly its data changes, how readily information can be recreated, and the impact if staff cannot use it.

The National Cyber Security Centre (NCSC) recommends identifying critical systems, setting recovery requirements and testing recovery processes regularly. Two measures help turn this into a practical plan:

  • Recovery point objective (RPO): the maximum amount of data loss the business can accept. If losing more than one day of work would cause material disruption, the recoverable copy needs to be no more than a day old.
  • Recovery time objective (RTO): the maximum time a service can be unavailable before the business is materially affected.

These targets may differ across your business. For example, active customer records or financial systems may need a faster recovery target than historical files. NCSC notes that recovery requirements should reflect data importance, change rate and whether it can be recreated. Read NCSC’s backup guidance.

A sensible starting framework is:

  • Monthly: review failed backup jobs, alerts, storage capacity, retention settings and whether recently introduced systems are protected.
  • Quarterly: restore representative files or records, and test a priority application or database where practical.
  • After a material change: test the affected restore process after a migration, major upgrade, new application, identity change or backup-platform change.
  • At least annually: run an end-to-end exercise for a critical business service, including recovery steps, staff access, communications and business checks.

This is a risk-based starting point, not a universal requirement. NCSC guidance includes differing indicative intervals, so businesses should not treat a single schedule as mandatory. Increase the scope or frequency where short outages or data loss would materially affect customers, cashflow, contractual commitments or operations.

What a practical backup test should include

A useful test does more than confirm that a file can be opened. It should show that restored information is sufficiently current and that staff can use the recovered service to complete an important business task.

NIST’s contingency-planning guidance distinguishes between validating recovered data and testing functionality. In practice, check that data is complete enough for your purpose, then confirm a user can sign in, run a report, process a transaction or complete another essential task. NIST SP 800-34 Rev. 1 is US guidance rather than a New Zealand legal requirement, but it provides a useful framework.

For each test, document the following:

  1. The scenario
    Choose a realistic event, such as accidental deletion, corrupted data, a failed application update, an unavailable server or a suspected ransomware incident.

  2. The service and restore point
    Record what will be restored and which backup copy will be used. This helps show whether the restore point meets your RPO.

  3. A safe test environment
    Avoid unintentionally overwriting production information. Where supported, use an isolated environment. For organisations using Azure Site Recovery, Microsoft recommends a separate test network because testing in a production recovery network can create downtime or duplicate-identity issues. This advice is specific to that product, not every backup environment. Microsoft’s Azure Site Recovery guidance explains the approach.

  4. Named technical and business roles
    Identify who can authorise recovery, begin technical work, validate the result, communicate with staff and make operational decisions. NCSC advises mapping recovery roles to individuals and testing them during disaster recovery and business continuity exercises. NCSC’s data recovery guidance also recommends maintaining an understanding of critical systems and responsibilities.

  5. Timed restoration and business validation
    Measure how long restoration takes. Then confirm the information is usable and that the required business task can be completed. A server starting successfully does not necessarily mean that applications, user access and dependencies are working.

  6. Results and corrective actions
    Record what worked, what failed, actual recovery times, dependencies and assigned actions. Do not assume an RTO is achievable until it has been tested and measured.

Test more than individual files

A single-file restore is valuable, but it tests only one recovery path. A broader disaster recovery exercise should consider the systems, people and decisions required to resume essential operations.

NCSC identifies restoration scenarios involving files, applications, databases, operating systems, and identity and access. Some may need to be tested together. Recovering an application, for example, may depend on user accounts, multi-factor authentication, network connectivity, DNS settings, databases or supplier access.

Keep an annual exercise manageable by focusing on one critical service. Walk through the scenario from incident declaration to business resumption:

  • Who decides that recovery is necessary?
  • Where are recovery plans and contact lists if email is unavailable?
  • Which service must be restored first?
  • How will staff access it?
  • What task demonstrates that the service is usable?
  • Who communicates with staff, customers or suppliers while normal systems are disrupted?

NCSC advises planning for communications when usual systems, including email, are unavailable, and keeping offline copies of important plans and contacts. It also warns that attackers may target backups, making an offline or disconnected copy an important resilience measure. This can reduce risk, but it does not prevent every attack or guarantee recovery. Read NCSC’s ransomware guidance.

Checklist for your next recovery test

Before closing the test, confirm that:

  • critical systems and data have been identified;
  • acceptable data-loss and downtime targets have been agreed;
  • the selected backup copy restored successfully;
  • recovered data was checked for completeness and usefulness;
  • a staff member could access the service and complete a key task;
  • actual recovery time was recorded;
  • recovery roles and communication methods were tested;
  • gaps have owners and follow-up dates; and
  • the schedule will be reviewed after material changes.

Testing does not guarantee uninterrupted operations, compliance or successful recovery from every event. It does provide evidence of whether current arrangements work in realistic conditions and where improvement is needed.

If you need help assessing recovery priorities or planning a controlled test, review EdgePoint IT Solutions’ IT services for New Zealand businesses and discuss your environment with a qualified provider.

Frequently asked questions

How often should a business test its backups?

There is no universal interval. Set the schedule according to data importance and change rate, acceptable data loss, acceptable downtime and technology changes. Regular smaller restore tests and broader recovery exercises provide stronger evidence than backup reports alone.

What is the difference between a backup test and a disaster recovery test?

A backup test may restore a file, record or application data. A disaster recovery test is broader: it checks whether systems, access, people, procedures and communications work together to resume an essential business service.

Should we test recovery after a major system change?

Yes. A migration, major upgrade, new application, identity change or backup-platform change can affect what is protected and how restoration works. Testing the affected recovery process can identify gaps before an incident.