


A cloud service becoming available again does not tell you whether yesterday's customer enquiry arrived, an approval completed or a scheduled task ran. For business leaders, recovery needs a second question: what work still needs attention?
The Microsoft 365 disruption reported on 12 November is a timely reason to ask it. In a public update on 13 November, Cochise County in the United States said email was flowing again while warning that some messages could still be delayed. It directed urgent enquiries to telephone contact. That is a specific organisation's experience, not evidence that every Australian business was affected.
The useful lesson is broader than one incident. Technical recovery and business recovery need separate checks. This article is a planning guide, not a live status notice or a conclusion about the cause of the November disruption.
Ask your IT team or provider to confirm the affected services and time period for your organisation. Microsoft's service health guidance explains how administrators can check active incidents and recent history in the Microsoft 365 admin centre. If they cannot sign in, Microsoft points to its public service status page for relevant known issues.
Keep that technical record alongside a business record: which teams could not complete which tasks, what deadlines matter and who owns the follow-up. Avoid assuming that a colleague's working inbox proves that everyone has recovered. Equally, do not attribute every remaining fault to the supplier incident without checking.
Start with a short, prioritised list. Customer commitments, payroll approvals, order processing and time-sensitive bookings may deserve attention before lower-impact internal messages. The right order depends on your business, rather than the number of error notifications.
A failed screen or timeout can leave the outcome unclear. An action may have failed, be waiting to run, or have completed without returning a confirmation. Repeating it immediately can create duplicate work.
For example, consider an illustrative order-approval workflow that stopped responding during an interruption. Before approving the same order again, the owner should check the order's recorded state, any downstream dispatch instruction and the approval history. This is a hypothetical example, not a report of a customer incident.
Use a simple reconciliation list with four fields: the item, the last confirmed outcome, the next action and the responsible person. For each affected workflow:
A blanket instruction to resend everything may be appropriate for a particular recipient's situation, but it should not become your default for payments, orders or automated actions. Follow relevant recipient guidance and check the consequences of duplication.
Your email and collaboration tools may also support approval notifications, reporting jobs and connected applications. Ask the owners of those workflows what happened during the affected period. A person being able to sign in again does not establish that a scheduled job completed.
Where an AI assistant or automation relies on those services, check its actual task history. Did a request fail, remain queued or produce an incomplete result? Review material decisions or external actions before resuming unattended processing. Do not assume a generated summary contains information that was unavailable when it was created.
This does not mean every business needs a second copy of every application. Additional platforms bring cost, administration and their own dependencies. Begin with the workflows where an interruption has a material consequence, then decide what practical fallback is justified.
Use the recovery review to improve a small number of concrete arrangements. Keep an approved contact route available outside the service it is meant to replace. Identify who can authorise a temporary process, where essential instructions can be accessed and how temporary records will be brought back into the normal system.
Keep privacy and access controls in that plan. An outage is not a reason to move customer records into personal email or an unapproved file-sharing account. A limited telephone process or a controlled pause may be more appropriate than an improvised workaround.
Zarbtech's Cloud Services support the implementation and maintenance of business cloud environments, while Managed IT provides ongoing technical support and guidance. Our practical starting point is to connect the technical incident with the work your organisation needs to finish. No support provider can guarantee that a third-party cloud service will never fail, but clear responsibilities can make your response more organised.
Choose one important workflow this week. Confirm its fallback, its recovery checks and the person who can declare the business task complete. Bring that workflow to a discussion with Zarbtech about your cloud support arrangements.