By Touchbase Technologies · Published
An email-provider outage response plan for a small business
When email stops working, start by finding out what is affected and assigning someone to coordinate the response. Keep a record of important customer work while you investigate. Repeatedly pressing Send or changing unrelated account settings can make it harder to understand what happened.
This is an operational preparation guide, not a live outage report or a claim that a particular provider is currently unavailable. The example business and response log are fictional. Use the provider's current information and your own observed symptoms to decide the next step.

Identify the symptom before calling it an outage
Write down which account and action failed, the approximate time, the displayed message and whether the problem affects the browser, a mail app or a connected application. Keep private message content and credentials out of any public report.
A failed sign-in, an undelivered message and a missing notification are different symptoms. A working website does not establish that mailbox access is healthy. A coworker's working account does not establish that your account is fine. Use these observations to narrow the issue rather than to announce a cause you have not confirmed.
Ask the person responsible for your email service to check its official health information. Microsoft's service-health documentation describes the Microsoft 365 admin center's Service health view and a status page for known issues that prevent admin sign-in. The Google Workspace Status Dashboard reports on listed Workspace services and points users to support for issues not listed there. Match the product and account involved; a broad public dashboard may not answer every account-specific question.
Give one person responsibility for the response
Choose a coordinator who keeps the facts, open questions and next review time in one place. They do not have to solve the technical problem personally. Their job is to prevent five people from independently changing settings or sending inconsistent customer updates.
Keep a short incident note with the affected service, observed symptoms, official status link, responsible administrator and the next check time. Record what you tried and what happened. If a provider gives an incident identifier, keep it with the note so later updates refer to the same event.
Decide what work needs immediate attention. A customer waiting for information about today's appointment may need a different response from a promotional campaign scheduled for next week. Assign the customer-facing work separately from the technical investigation.
Pause messages you cannot safely account for
If your campaign tool or workflow has a supported pause control, have its owner review whether scheduled work should pause while the issue is investigated. Record the time and the affected work. Do not assume that pausing one application stops every queued message in another system.
Keep a ledger of uncertain sends. “I clicked Send” is not enough to decide that a customer received a message. Likewise, the absence of a reply is not evidence that the message failed. Preserve the available application or provider status rather than replacing it with a guess.
| Item | What to record | What to decide later |
|---|---|---|
| Appointment update | Customer record reference, deadline and current message status | Whether a permitted alternative contact method is needed |
| Campaign | Audience version, owner and pause time | Whether the offer remains accurate when sending resumes |
| Individual reply | Conversation reference and uncertain send time | Whether the provider confirms acceptance or an error |
| Notification | System that should create it and observed missing step | Whether the underlying work was received despite the missing alert |
Keep this ledger in an appropriate private system. Use references instead of copying customer conversations into a public document. The contact-backup guide covers a related continuity task, but a backup alone does not restore an unavailable mailbox service.
Tell affected customers what they can do now
Use an existing, appropriate channel that actually works and that the customer can reasonably use. That might be a phone number already used for their appointment or a status message on a business website you control. Do not move customer records into an unfamiliar service simply to improvise a workaround.
A useful update states the observed service limitation, what it means for the customer, the available next step and when you will review the situation. Avoid a restoration deadline unless the provider has confirmed it and it is relevant to your service.
For a fictional repair shop, the update might read:
We are having difficulty accessing our email today. If you need to change a repair pickup arranged for this afternoon, please call our usual shop number. We are keeping a record of pending requests and will review them again at 2 p.m. local time.
Replace those details with verified information. The note does not claim all messages were lost or promise that the issue will end at the review time. A check-in time is a commitment your own team can manage.
Preserve security while investigating
Use your provider's normal support and recovery process. An outage does not create a reason to share passwords, disable account protections or trust unsolicited “repair” messages. Verify the destination before entering account details.
If you can still access the administrator's existing tools, keep their findings with the incident note. If you cannot, use official support channels available for your account. Avoid repeatedly changing DNS, mailbox connections or application permissions without a diagnosis and an accountable owner.
A connected application may show its own error while the provider is healthy, or it may need reconnection after an authorization problem. Treat that as a separate hypothesis to verify. Do not attribute every error to the broad incident being discussed online.
Recover in a controlled order
After the provider reports recovery, check the task your business actually needs. With permission and a suitable test address, verify account access and the relevant send/receive workflow. A general status update does not establish that every integration or queued job has recovered.
Review uncertain messages before resending them. Use the provider or application's available message status and conversation record. If the status remains unclear, keep that uncertainty visible and decide on a customer-appropriate follow-up instead of replaying a whole campaign blindly.
Check whether the message still makes sense. An offer may have expired, stock may have changed or a customer may already have contacted you by phone. Recheck exclusions and recent replies before resuming promotional work.
A suggested recovery order is to resolve time-sensitive customer requests, reconcile uncertain individual messages, verify required integrations and then review paused campaigns. Adapt that order to the business impact you actually observed.
A worked review after service returns
Imagine the coordinator recorded four pending appointment replies and one paused campaign. Two customers called during the interruption, one reply later shows as successfully sent and one remains unclear. The team updates the two phone conversations, avoids duplicating the confirmed reply and assigns someone to investigate the remaining record. The campaign owner separately checks the offer and audience before resuming it.
These are fictional counts, not an outage case study. The useful principle is to reconcile each kind of work rather than label everything “fixed” when the dashboard changes color.
Afterward, record what was delayed, what alternative contact method helped and what made the response harder. Add a coordinator and review slot to your business marketing calendar. Keep the runbook available to the people who will need it when normal communication is inconvenient.
Keep the plan usable across tools
This response process works with an existing mailbox, a private work ledger and accountable owners. If you evaluate Touchbase alongside your other systems, compare its current plans and allowances with the work you need to coordinate. A software purchase does not replace an incident owner or the provider's recovery process.
Check uncertain sends before reopening the queue
Keep messages with an unknown result separate from messages that definitely failed before submission. For an uncertain send, record the provider reference when available, the intended recipient reference and the last attempted time. Have the owner check the sending service's supported delivery or activity records before deciding whether another attempt is needed. A missing copy in one local view is not enough evidence to resend the whole queue.
Questions before the next interruption
Does a public status page prove my account is healthy?
No. Match the service and reported scope to your symptoms. An account-specific or connected-application problem may require its own investigation even when a broad dashboard shows no listed issue.
Should we resend everything afterward?
Review the recorded status and customer context first. Some messages may already have been sent or become irrelevant. Assign an owner to uncertain items and keep the decision visible.
What should we prepare today?
Write down the official service-health links, the responsible administrator, the customer-response coordinator and the private place where pending work will be recorded. Agree an appropriate alternative contact route before it is urgently needed.
Prepared with AI assistance and checked against the sources linked in this guide.
Your next step
Get started with Touchbase
Find relevant customers, prepare outreach for review and keep contacts, replies and follow-up connected.
Start for freeStart with a Free account. Upgrade or add credits when you need more capacity. Compare plans