Philippines staffing blog ·

Write a customer notice for a portal outage

A good outage notice says what is down, what still works, and when the next update arrives, so customers do not discover the failure one call at a time.

Write a customer notice for a portal outage help desk operations illustration

Direct answer

Operational guidance

Say what is down and what still works

A portal outage notice should answer two questions before anything else: which service is unavailable and which channels still work. Customers read the first sentence to decide whether their problem is covered, so put the scope there. Say whether the outage affects sign-in, ticket submission, attachments, or status updates, because a vague notice sends people to test every feature. A precise scope also reduces duplicate contacts from customers who assume everything is broken.

Name the fallback channel the desk has approved, and describe what it can handle. Email may accept a request but not a large attachment, while phone may verify identity but not process a refund. Tell the customer what to expect from the fallback so they do not repeat the same failure. If no fallback is available, say so plainly and give the next update time instead of implying one exists.

Give the next update time

Every outage notice needs a time for the next update, even when the desk has no new information. The time is a promise the desk can keep, while a repair estimate is a promise that depends on engineers. Choose an interval that matches the impact, such as hourly for a full outage and daily for a degraded feature. Put the time in the notice and set a reminder that reaches the person who will send it.

Send the update even when nothing has changed. A short note that says the team is still working and the next update is at three o'clock is more useful than silence. Silence invites customers to open a second ticket or call to ask whether anyone is there. Keep the wording stable across updates so a reader can scan for the one fact that changed.

Avoid promising a fix time

Do not convert an engineering estimate into a customer promise unless the engineer has confirmed it for publication. A repair may depend on a vendor, a deployment window, or a test that has not run. When the estimate changes, the desk inherits the anger created by the earlier promise. Say what is known and what is still being investigated, and keep the uncertain part visibly uncertain.

If the client insists on a target, label it as a target rather than a commitment. Explain that the time may move and that the desk will send an update before it passes. Record who approved the wording and when. The approval protects the agent who has to repeat the message to an upset customer at the end of a long shift.

Keep one notice across channels

Use the same core notice on the portal, the help page, the phone greeting, and the social account. Customers move between channels and notice when the stories differ. A single approved paragraph can be shortened for a phone greeting without changing the facts. Assign one person to update every location, because a stale page is worse than no page during an outage.

Keep the notice somewhere the whole desk can find it, not only in the inbox of the person who wrote it. A shared document with a version and a timestamp lets a new shift continue without reconstructing the message. Store the approved fallback steps beside it. When the outage ends, archive the notice with the incident record rather than deleting it.

Prepare the notice before the outage

Draft the notice while the systems are working, when nobody is under pressure. Fill in the placeholders for scope, fallback, update interval, and approval contact. Ask the client to approve the structure once, so an incident only requires the facts to be inserted. A prepared template is the difference between a calm first message and a paragraph written in a hurry.

Test the fallback channel before you name it in a notice. Send a request, confirm it reaches the queue, and check that the customer receives an acknowledgement. A fallback that has never been tested may fail at the moment it is advertised. Record the test date with the template and repeat it after a platform change or a mailbox migration.

Close the loop after recovery

Publish a recovery notice that says which services are back and which are still limited. Customers who waited need confirmation that the problem is over, not just the absence of new updates. If the desk expects a backlog, say when first responses will return to normal. That sentence prevents a second wave of contacts from customers who assume the queue is already clear.

Review the outage week with the client and count the duplicate contacts, the missed update times, and the tickets that arrived through the fallback. Each number points to a different correction in the notice or the process. Keep the review short and factual, and record the date. The next outage will reuse the same notice faster if the lessons are written down.

Limits and a practical standard

A notice cannot restore a service, and it cannot make a customer read the page before calling. It can reduce confusion, duplicate work, and the number of people who discover the outage one at a time. The desk also cannot control the engineering estimate or the vendor's schedule. Say that in the client review so the notice is judged on clarity rather than on the length of the outage.

A practical standard is a scoped opening, an approved fallback, a next-update time, a stable message across channels, and a recovery note. Rehearse the template once with the client before an outage, not during one. Keep the template with the incident contacts and the approval rule. A desk that communicates well during an outage is judged on the notice as much as on the repair.

Authoritative sources

Related planning pages