An IT disaster recovery policy is the short, management-approved document that sets the rules for recovering IT after a disruption: its scope, who is responsible, the recovery and testing requirements, and when it is reviewed. The step-by-step recovery procedures live in the separate disaster recovery plan, which the policy requires and governs.
Table of Contents
ToggleWhat is an IT disaster recovery policy?
An IT disaster recovery policy is a formal statement from management that gives IT recovery its authority, scope and rules.
An IT disaster recovery policy is the management-approved document that says which systems must be recoverable, who is responsible, how recovery is tested and when the arrangement is reviewed. It holds rules, not recovery steps.
NIST SP 800-34 Rev. 1 calls it the contingency planning policy statement and makes it step one of its seven-step process, because “a formal policy provides the authority and guidance necessary to develop an effective contingency plan”. The guide also says the plan must be based on a clearly defined policy.
Real policies show the shape. The Florida Tech IT Disaster Recovery Policy opens with a purpose clause: to ensure “preparedness for and recovery from any event that disrupts Florida Tech’s business continuity”. A policy is not a runbook, a downloadable template or the business impact analysis that feeds it.
Policy and plan: how the two documents differ
The policy sets the rules and the plan holds the procedures: they differ in purpose, audience, detail and how often they change.
NIST describes the policy as defining the overall objectives and the organisational framework and responsibilities, and the plan as the documented procedures that carry them out. We follow NIST here: the policy governs the plan, not the other way round.
| Aspect | Policy | Plan |
|---|---|---|
| Purpose | Sets objectives, rules and responsibilities | Gives the procedures to recover systems and services |
| Audience | Executives, auditors and system owners | The people who carry out recovery |
| What it contains | Scope, roles, testing and review requirements, exceptions | Contact lists, recovery steps, restore order, alternate site details |
| Level of detail | High level, no system-by-system steps | Detailed, per system or service |
| How often it changes | Rarely: annual review and after major change | Often: as systems, people and suppliers change |
| Usual owner | Senior management, usually through the CIO | The IT disaster recovery coordinator and system owners |
The IT recovery plan sits inside wider continuity work. Ready.gov says an IT disaster recovery plan should be “developed in conjunction with the business continuity plan”, and our guide to business continuity and disaster recovery covers how the two fit together.
What an IT disaster recovery policy should contain
A sound policy covers purpose and authority, scope, roles, recovery objectives, backup rules, testing, training, communication, review and exceptions.
NIST SP 800-34 Rev. 1 lists seven key policy elements: roles and responsibilities, scope, resource requirements, training requirements, exercise and testing schedules, a plan maintenance schedule, and the minimum backup frequency with storage of backup media. Florida Tech and Weill Cornell Medicine add exceptions, enforcement, governance and recovery tiers. Together they give ten elements:
- Purpose and authority: why the policy exists and who approves it.
- Scope: the systems and sites covered, including cloud and third-party hosted systems.
- Roles, responsibilities and resources: the named owner, the coordinator, the system owners and the budget or staff the programme needs.
- Recovery objectives and tiers: a requirement that every system has an objective and a tier, with the numbers held in a schedule.
- Backup and offsite storage: the minimum backup frequency and where copies are kept.
- Testing and exercises: how often recovery is tested and what evidence is kept.
- Training: who is trained and how often.
- Communication: who tells staff, customers and regulators what, and when.
- Review and maintenance: the triggers and the frequency.
- Exceptions and enforcement: how to request a waiver and what happens when the policy is ignored.
The policy sets only the backup rule. How backups are taken and restored is explained in our guide to backup and disaster recovery.
Three short example clauses show the register (illustrative wording, adapt to your organisation):
Scope: This policy applies to all IT systems that support the organisation’s operations, including systems hosted in the cloud or by third parties.
Objectives: Every in-scope system is assigned a recovery tier. The recovery time objective and recovery point objective for each tier are set in the tier schedule approved by the policy owner.
Testing: Recovery of each tier is tested at the frequency set in the tier schedule. Results are recorded and reported to the policy owner, and open follow-up actions are tracked to closure.
NIST prints a full sample contingency policy statement in section 3.1 of its guide. To create your own policy, work through five steps:
- Decide the owner and the approver.
- Run a business impact analysis, or reuse a current one.
- Set the recovery tiers and the testing rules.
- Draft the scope and the roles.
- Approve and publish the policy, and set the first review date.
Contact lists, runbooks and restore steps belong in the plan, not the policy. Our IT disaster recovery plan template covers that side.
Who owns and approves the policy
A senior executive approves the policy, a named owner, usually the CIO or an equivalent, keeps it current, and system owners carry it out.
| Role | What it does with the policy | Seen in |
|---|---|---|
| Approver | Signs off the policy and its changes | Florida Tech: approved by the university president |
| Owner | Keeps the policy current and enforces it | Florida Tech names the CIO as policy owner; NIST says senior management, most likely the CIO, must support the programme |
| Governance forum | Receives status reports and ensures resources are available | Weill Cornell: a governance committee that receives reports at least twice a year |
| DR coordinator | Coordinates planning, testing and auditing activity | Weill Cornell’s disaster recovery programme |
| System owners | Review and update plans each year and approve test results | Florida Tech: the Information System Owner |
Where a regulator applies, accountability moves up. The MAS Guidelines on Business Continuity Management, written for financial institutions, say the Board and senior management are ultimately responsible for business continuity (section 10.1, published 6 June 2022).
A provider can run recovery but, in our view, cannot own your risk decisions. Even when a provider such as us operates the recovery, the policy should still name an internal owner.
Recovery objectives and testing: what the policy fixes
The policy requires recovery objectives for every system and sets the testing rules, while the actual RTO and RPO values sit in a tier schedule.
The Canadian Centre for Cyber Security defines three measures. Maximum tolerable downtime is the total time a process can be unavailable without significant harm. The recovery point objective (RPO) is the data loss the organisation can tolerate. The recovery time objective (RTO) is the planned time and level of service needed to meet the system owner’s minimum expectations.
A recovery tier is a criticality rating that fixes the RTO and RPO for a group of systems, so each system inherits its targets from its tier.
Weill Cornell’s recovery tier chart is a worked example: it defines the RTO and RPO of every system its IT department manages. Objectives come from the business impact analysis. Ready.gov puts it plainly: “The recovery time for an IT resource should match the recovery time objective for the business function or process that depends on the IT resource.”
We suggest keeping the numbers in a tier schedule attached to the policy, so a changed target does not reopen the policy itself. The Singapore Government ICT&SS catalogue uses the same pattern. Its Backup and Recovery controls BR-1, BR-2 and BR-3 leave backup frequency, recovery testing frequency and backup retention as parameters that each agency sets.
Testing rules should match what actually runs. At Accrets, DR rehearsal is available to customers who subscribe to our IT disaster recovery as a service and is usually run once a year, so the testing rule in the policy and the rehearsal calendar should say the same thing.
When a customer joins that service, we help produce a documented disaster recovery plan, and the RTO and RPO are determined during that step. Depending on the scale of the IT estate, a customer may choose to rehearse only selected systems, for example a single application or the network failover on its own. The testing rule in the policy should therefore state which scope each test covers and how often every system is covered.
How to review an IT disaster recovery policy
Review the policy at least once a year and after any major change or real incident, using a fixed checklist.
Weill Cornell requires its programme and plans to be updated after a significant organisational change or after the plans are used in response to a disruptive event, and otherwise reviewed annually. For financial institutions, MAS section 6.7 says to update business continuity policies, plans and procedures when the operating environment changes, and to review critical services and their recovery objectives at least annually.
A policy changes slowly, so the review is a set of checks rather than a rewrite:
- The scope still matches the estate, including cloud, SaaS and third-party hosting.
- The named owner and approver are still in post.
- Every in-scope system has a tier, and the tiers still match the business impact analysis.
- The backup and offsite rules match what actually runs.
- There is test evidence since the last review and the follow-up actions are closed. MAS section 7.3 describes recording the test objective, scope, scenario, participants, results and follow-ups, which is a good-practice example for any organisation.
- The exceptions register has been reviewed.
- The version, approval date and next review date are recorded.
Standards and Singapore references to check your policy against
No single standard is mandatory for every organisation, so check your policy against the references that apply to your sector.
Each row below states who the reference is written for, so nobody reads a financial-institution or government reference as a general legal duty. Our overview of cloud backup and disaster recovery in Singapore connects these references to day-to-day operations.
| Reference | What it is | Written for |
|---|---|---|
| NIST SP 800-34 Rev. 1 | Contingency planning guide, published May 2010 and updated 11 November 2010 | US federal information systems |
| ISO 22301:2019 | Requirements for business continuity management systems, published October 2019; ISO lists it as due for revision | Any organisation |
| ISO/IEC 27031:2025 | ICT readiness for business continuity, edition 2, published May 2025, replacing the withdrawn 2011 edition | Any organisation |
| MAS Guidelines on Business Continuity Management | Guidelines published 6 June 2022 | Financial institutions |
| Singapore Government ICT&SS Backup and Recovery controls | Controls BR-1 to BR-6, page last updated 24 March 2026 | Government agencies and industry partners |
The table does not say that any of these is required of you. Check with your regulator, auditor or counsel which ones apply.
Where to go next: from approved policy to tested plan
Once your IT disaster recovery policy is approved, the next job is the plan that carries it out.
- Write or update the plan, starting from our IT disaster recovery plan template.
- Run a recovery test that follows the testing rule in the policy, and keep the evidence.
- Review the policy annually and after any major change.
If you would like a second view on whether your IT disaster recovery policy and your recovery set-up match, talk to an infrastructure specialist.
Disaster recovery
Want a second view on your recovery set-up?
If you are unsure whether your policy and your recovery set-up match, our team can talk it through with you.
Contact UsDandy Pradana is an Digital Marketer and tech enthusiast focused on driving digital growth through smart infrastructure and automation. Aligned with Accrets’ mission, he bridges marketing strategy and cloud technology to help businesses scale securely and efficiently.




