Encrypted on a Friday night: how long until you reopen on Monday?
Cybersecurity
In its 2026 assessment, Interpol reports that cyber-related financial losses reported in Africa rose by 152% in a single year, and ranks ransomware and digital extortion among the continent's leading threats, alongside business email compromise. The same year, an Internal Audit Foundation survey put cyber incidents at the top of the risks cited by 62% of the African professionals questioned.
And it is no longer distant: the African specialist press reported that BGFIBank, Central Africa's leading banking group, was allegedly targeted by the BianLian criminal group. The financial sector is a prime target — value of the data, dependence on systems, ability to pay.
We are not going to sell you a product in this article. We are going to suggest an exercise.
The only question that matters
Not "are we protected" — nobody can honestly answer yes. The useful question is: if everything is encrypted on Friday at 7 p.m., what time on Monday are the teams working again? That figure already exists in your organisation. It is simply unknown, because it has never been measured.
The ninety-minute exercise
Block a morning, bring together the system administrator, the business owner of the chosen application and a stopwatch. Pick a server that genuinely matters — not the test server. And restore it, for real, into an isolated environment.
Write down these six figures, to the minute:
- Time to find the backup. How many minutes between "we start" and "we have identified the right restore point". For many, that is already where it stalls.
- The real age of the restored point. Your contract may say "daily backup". When exactly was the last successful backup?
- The duration of the restore itself, from launch to system boot.
- Time to usable service: database up, application reachable, a test user logging in successfully. That is the only delay senior management cares about.
- The number of human steps required, and above all: how many needed one specific person. If the answer is "only so-and-so knows how", you have found your main risk.
- What was missing. A password in an unreachable vault, a licence key, an addressing plan, a support number. Write them down: that list is worth more than the rest of the exercise.
Then multiply the result by the number of servers to bring back in a real scenario. That is your true recovery time. Compare it with the one you had announced.
What the exercise almost always reveals
- Backups run, but nobody reads the failure reports. A job that has been failing for eleven weeks is a classic.
- Backups are reachable from the domain. Yet ransomware that gains domain administration rights encrypts first what it finds — backups included. Hence the 3-2-1-1-0 rule: three copies, two media types, one off site, one immutable or disconnected, zero verification errors.
- The procedure exists, but it describes an earlier version of the system.
- Detection reported nothing because alerts go to a mailbox nobody reads.
What to put in place next, in this order
The order matters, because a security budget is badly spent when it starts with the most visible tool.
- An immutable or offline copy, verified automatically. Without it, everything else is decorative. Our deployments rely on Veeam, Veritas NetBackup or ExaGrid depending on the case.
- Endpoint and server monitoring — EDR or XDR depending on size, with an alert that reaches someone who reads it. CrowdStrike, SentinelOne, Sophos and Fortinet cover this need.
- Monitoring of the backup itself in your operations tool, just like a full disk. PRTG, ManageEngine or SolarWinds raise a failed job as an incident, not as one more email.
- The exercise, repeated every six months, with a different person at the controls each time.
A word on what not to do
Do not pay before measuring your alternatives: paying guarantees neither the key nor that the data will not be published. Do not restart servers before collecting the evidence, or you lose what would explain how they got in. And communicate early: reporting to the authorities and informing affected customers are prepared before the incident, not during it.
We run this restore exercise with our clients, in half a day, on a server they choose. The deliverable fits on one page: your six figures, and the list of what was missing.