What is backup?
A backup is a separate, recoverable copy of a company's data and systems, kept outside the production environment so everything can be restored after a failure, a human error, a deletion or an attack. It is not a mirrored disk or file sync: it is the copy that lets you return to an earlier point in time when the original is lost or encrypted.
How backup works
A good backup is not just copying files. It is a continuous process that makes sure the copy exists, is isolated from the risk and actually restores when you need it.
Scheduled copy
The system copies the data and, ideally, the whole system at set intervals, several times a day for what changes constantly.
Sent outside the environment
The copy leaves the source server for a separate destination, in the cloud or another location, so it does not go down with the environment if it is compromised.
Retention over time
Several points in time are kept for days, months or years, so you can go back to before the problem began, not just to the last state.
Recovery testing
The copy is restored and verified automatically, turning “I think I have a backup” into proof that it comes back.
Sources: NIST SP 800-34 (continuity).
What a real backup has to cover
- The whole system, not just files Recovering loose folders does not bring the server back. Image backup restores the system, the settings and the applications, ready to boot.
- Servers, computers and cloud Critical data is spread out: on the server, on the workstation and increasingly in Microsoft 365. All of it needs a copy.
- Microsoft 365 and other cloud services Microsoft keeps the platform running, but does not guarantee recovering what you deleted, what was encrypted or what an attacker destroyed. Native retention is short and removable.
- Databases and business systems The ERP, the finance system and the database need a consistent copy that can restore without corrupting the transaction.
Why having a backup is not the same as being able to recover
Most companies learn too late that the difference between a scare and a shutdown is the quality of the backup, not its existence. After a ransomware attack, more than half of victims recover less than half of their data, and only one in ten brings back more than 90% (Veeam, Ransomware Trends 2025). The reasons are well known: backups that were never tested, that sat on the same network the attacker reached, or that kept data for far too short a time. A backup that does not restore on the day of the crisis is a false sense of security, and the cost shows up as lost data, days of downtime and the decision to pay a ransom or not.
How to build a backup you can trust
What separates a decorative backup from one that saves the company is not the software, it is discipline. Five points define it:
- A copy isolated from the environmentThe backup cannot live on the same network the attack reaches. Isolation by architecture, not by procedure.
- Frequency that matches the dataWhat changes every hour needs a copy every hour. Frequency defines how much you can lose.
- Enough retentionSeveral points in time, to go back to before the infection began, not just to the last state, already contaminated.
- Automated recovery testingRestore regularly and with evidence. An untested backup is an unconfirmed backup.
- Full coverageServers, workstations and Microsoft 365. The one piece of data without a copy is exactly the one that will be missing.
In practice
The question that reveals the truth is not “do you have a backup?”, it is “when did someone last restore it and prove it works?”.
How Zamak handles backup
Zamak Technologies delivers managed backup with copies isolated from the environment, proper retention and automated recovery testing, so restoring is a documented certainty, not a hope. A good starting point is the Microsoft 365 backup check, which shows what native retention does not protect. It is part of Continuity in the Zamak Method.