Why backup is a business decision
Almost every business today depends on digital information: quotes, invoices, client documents, email and bookkeeping. When that information is unavailable, because of a fault, human error or an attack, work stops.
So the right question is not “which backup software should we buy”, but how long the business can be without its data (RTO), and how much data it can afford to lose (RPO). The answers to these two questions decide what the backup should look like.
Backup and restore are not the same thing
Backup is making copies of the data. Restore is the ability to bring the data back into use. You can back up every night for years and find out on the day it matters that the restore is too slow, partial or does not work at all.
That is why every backup plan also needs a recovery plan: what comes back first, how, and who is responsible for it.
What to back up
Start by mapping where the business data lives (data mapping). In most offices this includes:
- Shared folders and the file server: where most of the work is saved.
- Employee computers: especially laptops and files saved locally.
- Business email: including the mailboxes of employees who have left.
- Line-of-business systems: bookkeeping, CRM or professional software, following the vendor’s recommendations.
- Settings and documentation: network settings, licenses and the permission structure.
How many copies, and where
A common rule of thumb is 3-2-1: three copies of the data, on two different types of storage, with one of them outside the office (off-site). It is not a law, but it is a good starting point.
The key principle: a single event, whether fire, theft, flooding or ransomware, should not be able to hit both the source and every copy. A copy in the cloud or in a separate location gives you that separation.
Not sure where you stand? You can start with a review of your current backup: what is backed up, where, and when a restore was last tested.
Call +972-3-3801180Frequency and retention
Frequency decides how much work could be lost (RPO). If you back up once a day, a fault at four in the afternoon could cost a full day of work. For some data a daily backup is enough; for some, more frequent backups make sense.
Retention decides how far back you can go. You do not always notice right away that a file was damaged; sometimes it takes weeks. Keeping versions over time lets you go back to a point where the data was still intact.
Questions worth asking
- How many hours of work can the business afford to lose?
- How far back do we need to keep versions?
- Are there external requirements for keeping data in your field? (Worth checking with the relevant professional.)
Permissions and separation
A backup that any user on the network can reach and delete is exposed to the same risks as the original data. Limit access to the backup (least privilege), separate admin accounts from everyday work accounts, and use strong authentication (MFA) on the accounts that manage the backup.
Restore tests
The only way to know a backup works is to restore from it. Schedule regular restore tests, a single file, a folder and sometimes a whole system, and record what was tested, how long it took and what was found.
A test like this also gives you a realistic estimate of the time to get back to work (RTO), which is important for planning.
Common mistakes
- Backing up only to an external drive that is always connected: it is exposed to the same faults and attacks as the computer.
- Ignoring failure alerts: a backup that fails silently for months.
- Forgetting laptops and email: and backing up only the server.
- Never testing a restore: and discovering the problem on the day you need it.
- No one in charge: when everyone assumes someone else is handling the backup.
A closing checklist
- We mapped where the business data lives.
- We defined how much data we can lose (RPO) and how long we can be without it (RTO).
- There is at least one copy outside the office (off-site).
- Frequency and retention are defined.
- Access to the backup is limited and protected.
- Someone receives alerts and handles failures.
- A restore was recently tested and documented.
If the answer to some of these is “not sure”, that is exactly where to start.