A backup is a copy of your data; disaster recovery is the tested plan and capability to get the business running again within a set time. RTO is how long you can be down and RPO is how much data you can afford to lose, and both come out of a business impact analysis. This guide explains the terms, the 3-2-1-1-0 rule, Microsoft 365's shared responsibility model and how to set realistic targets for a small firm.
Backup vs disaster recovery comes down to one question: when something breaks, do you just have a copy of your data, or do you have a tested way to get the business working again by a specific time? A backup is the copy. Disaster recovery (DR) is the plan, tools and practice that turn that copy into running systems within your recovery time objective (RTO) and recovery point objective (RPO).
Most small firms we meet have backups. Far fewer know how long a full restore would take, or have ever tried one. That gap is what this guide is about.
Backup vs disaster recovery: the short version
| Backup | Disaster recovery | |
|---|---|---|
| What it is | A copy of data stored separately from the original | A plan and capability to restore operations after an outage |
| Answers the question | "Do we still have the file?" | "How fast are we working again, and how much did we lose?" |
| Measured by | Coverage, frequency, retention | RTO and RPO, proven by testing |
| Typical scope | Files, databases, mailboxes, system images | Servers, apps, network, identity, people, communications |
| Failure mode | Missing data, corrupted or encrypted copies | Long downtime even when data survives |
A useful way to think about it: backup protects data, DR protects the business. You need both, and DR depends entirely on backups that work.
RTO, RPO and the terms behind them
These terms get used loosely, so here are the definitions from NIST, which most frameworks and insurers follow.
- Recovery time objective (RTO): NIST SP 800-34 defines it as the overall length of time a system's components can be in the recovery phase before negatively impacting the organization's mission or business processes. NIST glossary.
- Recovery point objective (RPO): "The point in time to which data must be recovered after an outage." NIST glossary. In plain terms, how much recent work you can afford to lose.
- Maximum tolerable downtime (MTD): the amount of time a business process can be disrupted without causing significant harm. NIST glossary. Your RTO for the systems behind a process has to fit inside its MTD.
- Business impact analysis (BIA): the process of analyzing operational functions and the effect a disruption might have on them. NIST glossary.
- Contingency plan: management policy and procedures that guide the response to a loss of capability. NIST's glossary also lists the disaster recovery plan alongside it. NIST glossary.
RPO drives how often you back up. If your RPO for the accounting database is one hour, a nightly backup does not meet it. RTO drives how you back up and where you restore to. A 4-hour RTO for a file server is hard to hit if the only copy is in a cloud bucket that takes 20 hours to download.
The 3-2-1 rule and the 3-2-1-1-0 update
The 3-2-1 rule has been the baseline for years. Veeam, which popularized the extended version, describes it as three copies of your data, stored on two different media, with one copy offsite.
The 3-2-1-1-0 version adds two requirements:
- 1 copy that is immutable, offline or air-gapped.
- 0 errors, meaning you verify that backups restore cleanly.
Here is what that can look like for a 25-person office:
| Copy | Where it lives | Purpose |
|---|---|---|
| Production data | Server, NAS or Microsoft 365 | The live copy |
| Local backup | Backup appliance on site | Fast restores of files and servers |
| Cloud backup, immutable | Provider storage with retention lock | Survives ransomware, theft, fire or flood at the office |
| Verification | Automated restore checks plus manual tests | Proves the "0 errors" part |
"Two different media" trips people up. Two folders in the same cloud account, or two drives in the same RAID array, share a failure point. The goal is that one event cannot take out every copy.
Immutability and why ransomware changed backup design
Ten years ago the main threats to backups were hardware failure and accidental deletion. Now attackers go after backups on purpose. CISA's #StopRansomware Guide warns that many ransomware variants try to find and then delete or encrypt accessible backups, and it recommends offline, encrypted backups of critical data, with regular tests of their availability and integrity.
Immutable storage means a backup cannot be changed or deleted for a set retention period, even by someone holding admin credentials. That matters because attackers who get into a network often get into the backup console too. A few practical rules:
- Use separate credentials, with MFA, for the backup system. Do not join backup servers to the same domain as everything else.
- Turn on retention lock or object lock where your provider supports it.
- Keep at least one copy that an attacker with your domain admin password still cannot touch.
- Alert on unusual backup activity, such as retention changes or mass deletions.
Cyber insurers ask about this directly. See our guide to cyber insurance requirements for the backup questions that show up on applications.
Microsoft 365 is not a backup plan by itself
A common assumption is that moving to Microsoft 365 means Microsoft handles backup. It does not, at least not in the way most owners mean. Microsoft's shared responsibility model lists customer data, configurations and identities as the customer's responsibility in every cloud model, including SaaS.
Microsoft does keep the service resilient, and there are built-in recovery features:
- SharePoint: deleted items stay in the site recycle bins for a total of 93 days from deletion, per Microsoft Support.
- Exchange Online: soft-deleted mail is kept for 14 days by default, adjustable up to 30 days, per Microsoft Learn.
Those features help with an accidental delete. They are not a point-in-time backup you control, and they will not help much if a compromised account quietly deletes or encrypts content and nobody notices for two months.
Microsoft also sells Microsoft 365 Backup, a separate pay-as-you-go product that backs up SharePoint sites, OneDrive accounts and Exchange mailboxes with configurable recovery windows. Microsoft describes its backup storage as append-only, and partner tools can be built on the same platform. Third-party Microsoft 365 backup tools are another option. Whichever you choose, it should be in your plan explicitly, with its own RPO, RTO and restore test. Our cloud services team can help you compare the options for your tenant.
Example RTO and RPO tiers for a small firm
Every business is different, so treat the table below as illustrative only. It shows how a 20 to 40 person professional services firm might tier its systems after a BIA. Your numbers should come from your own analysis.
| Tier | Example systems | Example RTO | Example RPO | Typical approach |
|---|---|---|---|---|
| Tier 1: run the business | Email and identity (Microsoft 365), practice management or line-of-business app, phones | 4 hours | 1 hour | Cloud services with separate backup, local image backups with fast restore, documented failover |
| Tier 2: needed this week | File shares, document management, accounting system | 24 hours | 4 hours | Frequent snapshots, immutable cloud copy, tested server restore |
| Tier 3: can wait | Archives, old project data, reporting tools | 3 to 5 days | 24 hours | Nightly backup, cloud copy |
| Tier 4: rebuild if needed | Individual workstations | 1 to 2 days per device | Not applicable if data lives in the cloud | Standard image, spare laptops, data kept off local drives |
Two things stand out once firms do this. First, Tier 1 is usually smaller than people expect. Second, the hardest part of a 4-hour RTO is rarely the data. It is having someone available who knows the restore steps, plus somewhere to restore to.
How to run a business impact analysis
NIST SP 800-34 Rev. 1 describes the BIA as three steps: determine mission and business processes and their recovery criticality, identify resource requirements, and identify recovery priorities for system resources. For a small firm, that translates into a half-day exercise.
- List the business processes, not the systems. Billing clients, filing returns, seeing patients, submitting bids, payroll.
- For each process, ask how long you could go without it before real harm occurs: lost revenue, missed legal deadlines, clients leaving, regulatory exposure. That is your MTD.
- Ask how much recent work you could afford to re-create. That is the basis for RPO.
- Map each process to the systems, data, people and vendors it depends on, including internet access, phones and identity.
- Set RTO and RPO for each system so that every process can recover inside its MTD.
- Compare targets to what your current backups can actually do. The gaps become your project list.
- Get sign-off from leadership. These targets are business decisions with costs attached, not IT preferences.
Testing: the part everyone skips
NIST's contingency planning process includes testing, training and exercises as its own step, and for good reason. A backup that has never been restored is a guess.
A reasonable testing rhythm for a small business:
- Monthly: restore a few random files and a mailbox item, and confirm backup job reports are clean.
- Quarterly: restore a full server or VM to an isolated environment and time it. Compare the time to your RTO.
- Annually: run a tabletop exercise. Walk through a ransomware scenario with leadership and decide who calls the insurer, who talks to clients, and who authorizes spending.
- After major changes: new server, new line-of-business app, cloud migration, office move.
Write down the results, including the uncomfortable ones. A restore that took 11 hours against a 4-hour target is useful information.
What a disaster recovery plan should contain
A DR plan does not need to be long. It does need to be usable at 2 a.m. by someone stressed. Keep a printed copy and an offline copy, because the plan is useless if it lives only on the server that just went down.
- Contact list: staff, IT provider, insurer, key vendors, ISP, landlord
- Decision authority: who can declare a disaster and approve spending
- System inventory with RTO and RPO for each
- Step-by-step restore procedures for Tier 1 and Tier 2 systems
- Credentials storage and access method for emergencies, kept separate from production
- Communication templates for staff and clients
- Test log and date of last review
If you want help setting targets or testing what you have, our data backup and disaster recovery service starts with a BIA and a timed restore test, so you know where you stand before something forces the question. For firms that want this handled continuously, backup monitoring and restore testing are part of our managed IT services.
Frequently asked questions
What is the difference between backup and disaster recovery?
A backup is a copy of data stored somewhere other than the original. Disaster recovery is the broader plan and set of tools that restore your systems, applications and people to working order after an outage, within agreed time and data-loss targets. You can have backups without disaster recovery, but you cannot have disaster recovery without good backups. The difference shows up as hours or days of downtime.
What is the difference between RTO and RPO?
RTO (recovery time objective) is how long a system can be down before the outage seriously harms the business. RPO (recovery point objective) is the point in time you must be able to restore data to, which sets how much recent work you can afford to lose. A four-hour RTO with a one-hour RPO means the system is back within four hours and you lose at most an hour of data.
What is the 3-2-1-1-0 backup rule?
The classic 3-2-1 rule means three copies of your data, on two different types of media, with one copy offsite. The 3-2-1-1-0 version, popularized by backup vendor Veeam, adds one copy that is immutable, offline or air-gapped so ransomware cannot alter it, and zero errors, meaning you verify that backups actually restore. The extra digits reflect how ransomware now targets backups directly.
Does Microsoft back up my Microsoft 365 data?
Microsoft keeps its service resilient and replicates data across its infrastructure, but under its shared responsibility model you are responsible for your data, accounts and configuration. Built-in features like the SharePoint recycle bin (93 days) and Exchange deleted item retention (14 days by default, up to 30) help with accidents but are not a full backup. Microsoft sells a separate, paid Microsoft 365 Backup product, and many third-party tools exist.
How often should we test our backups?
Test file-level restores monthly and a full system or server restore at least quarterly, and run a tabletop exercise of the whole disaster recovery plan once a year. Also test after any major change, such as a new server or a cloud migration. NIST's contingency planning guidance treats testing, training and exercises as a required step, because an untested backup is an assumption, not a recovery capability.
Sources and references
- NIST CSRC Glossary: Recovery Time Objective
- NIST CSRC Glossary: Recovery Point Objective
- NIST CSRC Glossary: Maximum Tolerable Downtime
- NIST CSRC Glossary: Business Impact Analysis
- NIST CSRC Glossary: Contingency Plan
- NIST SP 800-34 Rev. 1: Contingency Planning Guide for Federal Information Systems
- Veeam: What is the 3-2-1 backup rule?
- CISA: #StopRansomware Guide
- Microsoft Learn: Shared responsibility in the cloud
- Microsoft Learn: Overview of Microsoft 365 Backup
- Microsoft Learn: Recoverable Items folder in Exchange Online
- Microsoft Support: Restore items in the Recycle Bin of a SharePoint site