Disaster Recovery Plan for Small Business: One-Page Template
A disaster recovery plan for small business is a short written plan for getting your systems and data back after an outage, a ransomware attack or the loss of your office: who does what, in what order, how fast, and from which backup. Below is a one-page template you can fill in this week, then a guide to each part of it.
The plan is the document you need on the day your systems are down, so it has to work when they are: printed, with phone numbers in it, and with the restore it depends on already tested. This guide is for a 10 to 100 person business with no in-house IT staff, where a managed IT provider (MSP) or a part-time IT person runs the systems. It covers getting technology and data back. It is not a full incident response plan for a cyberattack; CISA's #StopRansomware Guide covers that.
Disaster recovery plan for small business: the one-page template
Print it or copy it into a document and fill in every blank, section A together with your IT provider if you have one. Time budget: under an hour (an estimate; each section shows its share). Filled in, it may run to two printed sides. Keep it that short.
Small Business Disaster Recovery Plan (one-page template)
| Business name | Plan owner | Version and date | Next review date | Where printed copies are kept |
|---|---|---|---|---|
A. Roles and contacts (about 10 minutes, estimate)
| Role | Name | Mobile | Backup person | Backup's mobile |
|---|---|---|---|---|
| Incident lead (declares the disaster, section D) | ||||
| Leadership sign-off (approves spending and customer messages) | ||||
| Internal IT contact, if any | ||||
| Communications lead (runs section E) | ||||
| Finance contact (payroll, payments, insurance claim) |
| Field | Fill in |
|---|---|
| Provider name and 24/7 emergency number | |
| Escalation path: first contact, next level, account manager, provider leadership | 1. ____ 2. ____ 3. ____ 4. ____ |
| Contract and SLA reference: agreement name, date, section or clause number | |
| Response time the SLA commits to for a critical (system-down) incident | |
| Is restoring systems and data in scope, and at what cost? | [ ] In scope [ ] Extra charge [ ] Not stated |
| Who on your side can approve after-hours work or extra charges | |
| Where the signed contract is kept (one copy outside company systems) |
If the contract states no response time for a critical incident, write "not stated" and raise it with the provider now, not during an outage.
B. Critical systems (about 15 minutes, estimate)
- RTO (recovery time objective): the longest a system can stay down before the damage to the business becomes unacceptable.
- RPO (recovery point objective): how much recent data you can afford to lose, measured in hours. It sets how often the system must be backed up.
| Priority (1 = restore first) | System or data | What stops if it is down | Owner | RTO (hours) | RPO (hours of data) | Where it runs (on-site, cloud, provider-hosted) | Who restores it (staff or provider) |
|---|---|---|---|---|---|---|---|
| Example (illustrative): 1 | Accounting and invoicing | Billing, payroll prep | Office manager | 8 | 24 | Cloud app | Provider |
Keep the list to the systems the business cannot run without. If everything is priority 1, nothing is.
C. Backups and restore tests (about 10 minutes, estimate)
| System | Where backup copies live (on-site / off-site or cloud / offline) | How often | Encrypted? | Who can delete or change the copies | Who checks the backup report, and how often |
|---|---|---|---|---|---|
| [ ] Yes [ ] No | |||||
| [ ] Yes [ ] No | |||||
| [ ] Yes [ ] No |
Keep at least one copy offline and encrypted. Ransomware attackers look for backups they can reach and try to delete or encrypt them.
| Date | System restored | What was restored (one file / full system) | Time to restore (actual) | Within RTO? | Age of restored data | Within RPO? | Tested by | Problems found and fix |
|---|---|---|---|---|---|---|---|---|
| [ ] Yes [ ] No | [ ] Yes [ ] No | |||||||
| [ ] Yes [ ] No | [ ] Yes [ ] No | |||||||
| [ ] Yes [ ] No | [ ] Yes [ ] No |
Until a restore has been tested, you do not know whether the backup works.
D. When to declare a disaster (about 5 minutes, estimate)
Declare a disaster recovery incident when any box is true. Whoever spots it calls the incident lead; the incident lead declares.
- [ ] A priority 1 system has been down longer than ____ hours. Set this well below its RTO so recovery starts in time.
- [ ] Signs of ransomware or a security incident: a ransom note, files renamed or unreadable, or the IT provider reports an active attack.
- [ ] Data is lost or corrupted further back than its RPO allows.
- [ ] The workplace is unusable: fire, flood, extended power loss, no building access.
- [ ] A cloud service or vendor reports an outage that will outlast a priority 1 RTO.
- [ ] The IT provider says it cannot meet the SLA response time for a critical incident.
Declared by: ____ Time declared: ____ Logged where: ____
First three moves after declaring
- Call the IT provider's emergency line (section A) and the incident lead.
- For a suspected cyberattack, switch to phone calls and texts instead of company email, and follow the provider's instructions for isolating affected devices.
- Notify the cyber insurance carrier (section F).
E. Communication tree (about 10 minutes, estimate)
| Audience | Who tells them | How (phone, text, email, website) | How soon after declaring | Where the message template is kept |
|---|---|---|---|---|
| Staff | ||||
| Leadership or owners | ||||
| IT provider | ||||
| Key customers | ||||
| Key vendors and suppliers | ||||
| Bank or payment processor | ||||
| Landlord or building manager | ||||
| Legal counsel, if customer or employee data may be exposed (counsel decides any notification duty) |
Holding statement (write it now, send it later): "We are working through a systems outage. [Orders / appointments / support] are [paused / running manually]. Next update by [time]."
Keep a printed copy of this plan and every phone number somewhere that does not depend on company systems.
F. Cyber insurance (about 1 minute, estimate)
| Claims or breach hotline | Policy number | Notice requirement in the policy | Carrier-approved responders required? |
|---|---|---|---|
| [ ] Yes [ ] No [ ] Check the policy |
G. Annual test schedule (about 5 minutes, estimate)
Test the plan at least once a year, and again after any big change: a new system, a new IT provider, an office move. A quarterly rhythm (illustrative) spreads the work out:
| Quarter | Test | Owner | Date done | Result and fixes |
|---|---|---|---|---|
| Q1 | Restore test: one file and one full priority 1 system. Log both in section C. | |||
| Q2 | Call-tree drill: send the staff message from section E, time the replies, fix stale numbers. | |||
| Q3 | Tabletop exercise with the IT provider: talk through a ransomware scenario from first alert to full restore. | |||
| Q4 | Full review: RTO and RPO still right, contacts current, SLA clause re-read, plan re-approved and reprinted. |
| Approved by (owner or CEO) | Date | Next review | Copies kept |
|---|---|---|---|
| [ ] Printed at ____ [ ] Offline copy at ____ [ ] IT provider has a copy |
A. Roles and contacts: a backup for every name, and your provider's escalation path
Give every role a second person, because the disaster may take out the first one. Write mobile numbers, not desk extensions: the desk phones may run through the system that is down.
If an MSP runs your systems, copy its terms from the signed contract, not the sales proposal: the response time for a critical, system-down incident, the escalation path past the first technician, and the clause number. Our IT support SLA guide covers what a good response commitment looks like. Then check whether a restore is in scope and what it costs, and make the plan match the contract. NIST's contingency planning guide lists "Disaster declaration (i.e., circumstances constituting a disaster, notification procedures)" among the terms an agreement with a recovery-services provider should cover. For the rest of the contract, see our guide to MSP contract red flags.
The mistake to avoid: a contact sheet that lives only in email, or on the server you are trying to recover.
B. Critical systems: set an RTO and RPO for each one
Two numbers drive the plan, and NIST SP 800-34, the federal contingency planning guide, defines both. The recovery time objective (RTO) is "the maximum amount of time that a system resource can remain unavailable before there is an unacceptable impact" on the work it supports. The recovery point objective (RPO) is "the point in time, prior to a disruption or system outage, to which mission/business process data can be recovered" from the latest backup. In plain terms: how long can this system be down, and how many hours of work can you afford to redo?
In the template's illustrative example, an accounting system with an RTO of 8 hours and an RPO of 24 hours can wait a working day, and the business can re-enter a day of invoices. If it cannot, the backups need to run more often.
Set the numbers from the business side, then check them against reality. Ready.gov's IT disaster recovery plan guidance says: "Data restoration times should be confirmed and compared with the IT and business function recovery time objectives." A four-hour RTO means little if a full restore takes two days.
The mistake to avoid: marking every system priority 1. The ranking is your provider's restore order on the day.
C. Backups and restore tests: log restores, not backup jobs
Section C asks three things about each backup: where is the copy, who can delete it, and when was it last restored? Choosing the backup tools is a separate job; see our guide to cybersecurity for small business, and for where backup sits in the budget, how much a small business should spend on IT.
The #StopRansomware Guide from CISA and its federal partners says: "Maintain offline, encrypted backups of critical data, and regularly test the availability and integrity of backups in a disaster recovery scenario." Offline matters because ransomware actors look for backups they can reach and try to delete or encrypt them. An account that can delete the copies is worth stealing, which is why the template asks who holds one.
A backup report that says "success" tells you a copy was written. A restore test tells you the copy can be read, how long it takes to get back and how old the data is when it arrives. A disaster recovery plan you have never tested is a guess. One restore test teaches you more than the rest of the plan put together.
The mistake to avoid: a checkbox that says "backups are running".
D. When to declare a disaster: write the triggers down now
A disaster recovery plan does nothing until someone says it is in use. NIST's guide says: "Effective escalation and notification procedures should define and describe the events, thresholds, or other types of triggers that are necessary for additional action." Section D is that list. Set the downtime trigger well below each RTO. For example, if accounting can be down for 8 hours, declaring at hour 2 or 3 leaves the restore time to finish. Name one person who declares, plus a backup, so the decision does not wait for a meeting.
A suspected cyberattack changes the first moves. CISA's guide advises isolating systems and using "out-of-band communication methods such as phone calls" so that attackers watching your systems are not tipped off. This is where the disaster recovery plan meets the incident response plan: once someone declares, both start.
The mistake to avoid: triggers with no name next to them.
E. Communication tree: who tells whom when email is down
List each audience, who tells them, how and how soon: staff first, then customers, vendors, your bank and your landlord. Write the two-line holding statement now, not during the outage.
Keep a copy the outage cannot reach. CISA's guide says of an incident response plan: "Ensure a hard copy of the plan and an offline version is available." A phone list on the file server is no use when the file server is the problem. If customer or employee data may be exposed, call legal counsel before you say anything about the data; counsel decides any notification duty. If you hold health records or card data, see our guide to MSP compliance for HIPAA, PCI and SOC 2 for what to ask a provider.
The mistake to avoid: a call tree that depends on company email.
F. Cyber insurance: one line, filled in before you need it
If you carry cyber insurance, the policy may set a deadline for notifying the carrier and may name the vendors you are expected to call. CISA's guide lists your cyber insurance company among the stakeholders to notify during an incident. Copy the hotline, the policy number and the notice requirement from the policy itself. For what the application asks and the evidence to keep ready, see our guide to cyber insurance requirements for small business.
G. Annual test schedule: test the plan at least once a year
Ready.gov says: "Test the plan periodically to make sure that it works." NIST's guide, written for federal systems, is more specific: its plan template has the plan owner attest that the plan "will be tested at least annually". Treat once a year as the floor, and test again after a new system, a new IT provider or an office move.
The quarterly rhythm in the template is illustrative: a restore test, a call-tree drill, a tabletop exercise and a full review. A tabletop exercise is a meeting where the team and the provider talk through a scenario from first alert to full restore. CISA's Tabletop Exercise Packages are "a comprehensive set of resources designed to assist stakeholders in conducting their own exercises", with ransomware among the cybersecurity scenarios. The SBA's emergency preparedness guide puts it simply: "Practice your plan with your staff so you're ready when a disaster occurs."
Then have the owner sign it. The #StopRansomware Guide asks for the incident response and communications plans to be "reviewed and approved by the CEO, or equivalent, in writing". A signed plan is one the business has agreed to fund and follow.
The mistake to avoid: a plan still dated the year it was written.
What to ask your IT provider about disaster recovery
Take section A to your provider and ask these questions. Starting with a new provider? Add the plan to the onboarding checklist. Leaving one? Make sure the backups and documentation come with you; see how to switch MSP providers.
- "What response time does our contract commit to for a system-down incident, and who answers after hours?"
- "Is a full restore of our systems and data in scope, and if not, what does it cost?"
- "If the first technician cannot fix it, who do we call next, and who above them?"
- "Where do our backup copies live, is one offline and encrypted, and who can delete them?"
- "When did you last test a restore for us, how long did it take, and can we see the log?"
- "Will you join a tabletop exercise with us once a year, and keep a copy of our plan?"
Then compare providers on the same five signals you would use for any MSP:
- Response time. A commitment for system-down incidents written into the contract, with after-hours coverage spelled out.
- Ticket tracking. Each restore request, test and declared disaster logged as a ticket with a time stamp.
- Proactive monitoring. Someone reads the backup reports and fixes a failed job before you need the backup. Ask who, and how fast.
- Experience with businesses your size. Asked, not assumed: "Have you run a full restore for a client our size, and how long did it take?"
- Security posture evidence. Restore-test logs, an offline copy and MFA on the backup console, shown rather than promised.
Red flags in a disaster recovery plan
- The plan lives only on the server it is meant to recover. If the server is down, so is the plan.
- No one has run a restore. A backup job that reports success is not a restore test.
- The provider cannot say how long a full restore takes. Then your RTO is a hope, not a number.
- No one is named to declare a disaster. The decision waits while the clock runs.
- The phone list is in email. An email outage takes it away when you need it.
Compare backup and disaster recovery providers
Compare backup and disaster recovery providers in your state, then ask each one for its last restore-test result. The backup and disaster recovery directory links every state; for full IT support, see the managed IT directory. Earlier in the search? Start with the MSP buyer's guide.
Frequently asked questions
What should be in a disaster recovery plan?
For a small business, seven parts: roles and contacts, including your IT provider's escalation path; a ranked list of critical systems with an RTO and RPO for each; where the backups live and a restore-test log; the triggers for declaring a disaster; a communication tree; your cyber insurance notice details; and a test schedule. Ready.gov's IT disaster recovery guidance starts the same way, with an inventory of hardware, software applications and data.
What are the 5 steps of disaster recovery planning?
There is no single standard list. Here is one way to group the work, mapped to the template above: (1) inventory your systems and data and rank them; (2) set an RTO and RPO for each; (3) protect the backups and test a restore; (4) decide when you declare a disaster and who you tell; (5) test the plan on a schedule and update it after changes.
How much does a disaster recovery plan cost?
Writing the plan costs staff time: under an hour for the template (an estimate), plus a meeting with your IT provider for section A. The real cost sits in the backup and recovery capability the plan documents: the backup service, the offline copy and the provider's time to run a restore. Ask your provider to state restore scope and the cost of a full restore in the contract, so you know the number before the day you need it.
How often should we test a disaster recovery plan?
At least once a year, and again after any big change such as a new system, a new IT provider or an office move. NIST SP 800-34, written for federal systems, has the plan owner attest that the plan "will be tested at least annually". Smaller tests spread across the year, such as a quarterly restore test or call-tree drill, give you more chances to catch a stale phone number or a backup that will not restore.
What is the difference between RTO and RPO?
RTO (recovery time objective) is how long a system can be down before the damage becomes unacceptable. RPO (recovery point objective) is how much recent data you can afford to lose, which depends on when the last good backup ran. RTO tells your provider how fast to restore; RPO tells it how often to back up.
Is a disaster recovery plan the same as a business continuity plan?
No. A business continuity plan covers how the whole business keeps working through a disruption: people, premises, suppliers and processes. An IT disaster recovery plan covers getting technology and data back. Ready.gov says the IT disaster recovery plan "should be developed in conjunction with the business continuity plan", so the restore order in section B should follow the business's own priorities.
Free Buyer's Guide
The complete guide to finding and hiring the right MSP for your business.
Download Free Guide