A backup job reporting “success” only tells you that a job ran. It does not tell you that the right data was copied, that the copy is readable, or that you could get your business running again from it before your customers notice. The only way to know those things is to restore from the backup — on purpose, before you need to.
How backups quietly fail
Most backup failures we see were invisible until the day they mattered:
- The wrong things were backed up. Someone set up the job years ago; since then the accounting data moved to a new folder that nobody added.
- Open databases were skipped, so the files were copied but the business system’s data was not.
- Ransomware encrypted the backups too, because the backup drive was always plugged in or reachable with the same password as everything else.
- The encryption key or password was lost along with the person who set it up.
- The restore works, but takes days, because nobody had ever timed it.
Start with two questions
Before choosing any product, agree on two numbers for each important system:
- How much work can you afford to lose? If the answer for your accounts is “no more than an hour”, a nightly backup is not enough.
- How long can you afford to be down? If the answer is “half a day”, you need a restore you have actually completed within half a day.
These answers are business decisions, not IT ones, and they are different for email, your shared files and your line-of-business system. Once they are written down, designing the backup becomes straightforward.
The 3-2-1 rule, plus one
A sensible baseline is three copies of your data, on two different types of storage, with one copy off-site. Today we add one more: at least one copy that cannot be changed or deleted from your network, so a ransomware attack or a compromised admin account cannot reach it.
Cloud apps need backing up too
Microsoft 365 and Google Workspace keep your data available, but their built-in retention is limited and designed for their resilience, not your mistakes. If a user deletes a mailbox folder, or a departing employee’s account is removed, the data can be gone for good once the retention window passes. A separate backup of your cloud email and files closes that gap.
Test restores on a schedule
- Monthly: restore a handful of files and one mailbox item, and check they open.
- Quarterly: restore a whole system to a test machine and time how long it takes.
- Every time: write down what was restored, how long it took and anything that went wrong.
That written record is what turns a backup into a recovery plan. If the timings do not meet the numbers you agreed, you find out on a quiet Tuesday rather than in the middle of a crisis.
Where to start this week
Ask whoever looks after your IT three questions: what exactly is being backed up, where the copies live, and when a restore was last tested. If the answers are vague, our data backup and recovery service starts with exactly that audit, then puts automated, tested backups in place.