What a cloud backup is designed to do
A backup preserves a recoverable copy of information according to a retention policy. Depending on the service, it may help restore an accidentally deleted file, a previous version or a lost mailbox. The practical questions are what it contains, how independent it is and whether restoration has been tested.
Recovery can depend on a destination, credentials, software and available bandwidth. If the normal destination is the cloud service that is unavailable, having a backup alone may not give staff a working environment during that interruption.
What continuous synchronisation is designed to do
Synchronisation keeps an agreed copy close to the current source. For a local continuity appliance, that copy feeds the applications your team will use on site.
Changes arrive with a measurable delay. A continuous process is not a promise of zero data loss or instant mirroring: cloud APIs, rate limits, account permissions and connectivity all influence timing.
Synchronisation can also copy a deletion or an unwanted edit. That is why it needs an appropriate independent backup or snapshot policy, with protection designed around the risks you are addressing.
What a continuation environment adds
An operational environment combines the data with usable local applications, access rules, sign-in and a procedure. The test is concrete: can the right people open the right information and complete an agreed task while the primary service is unavailable?
| Capability | Question to test |
|---|---|
| Independent backup | Can we restore a selected earlier version after a bad change? |
| Continuous local sync | How recently did the last successful copy arrive? |
| Local working environment | Can staff use the agreed mail and files without the primary cloud service? |
| Continuation plan | Does the team know when to switch, what can continue and how to return? |
One supplier may provide several capabilities, but the labels are not interchangeable. Ask for a demonstration against the interruption your business needs to withstand.
Agree recovery and synchronisation targets
Decide how much recent work the business can tolerate being unavailable and how long it can wait before the essential tasks resume. These are often discussed as recovery point and recovery time objectives, but they should become measured acceptance tests rather than unexplained acronyms in a proposal.
For cloud continuity, also agree the normal sync gap, what happens when a sync job fails and which cloud-only features remain outside scope. Read the Microsoft 365 or Google Workspace guide for examples.
Build the layers around the business
An on-site appliance introduces its own dependencies: storage, power, the local network and the people operating it. Consider physical loss of the premises separately, alongside independent off-site recovery copies. A local working option is one part of the wider resilience design.
Orwell’s continuity appliance combines agreed cloud data with local services and a rehearsed working procedure. Use our planning checklist or discuss the requirements after your next incident review.