Outage Communication
E-mail went down at 10:15. The desk knew at 10:16, because forty people told them. By 10:40 the desk had found the cause, restarted the right thing, and watched the queue drain. At 10:45 e-mail was back. Nobody had been told anything at any point, including that it was back. People kept phoning until noon to ask whether it was safe to send. The outage lasted thirty minutes. The not-knowing lasted two hours, and the not-knowing is what everyone remembered. Communication during an outage is not a courtesy added to the fix. For most of the people affected, it is the fix.
This series is about saying something, early, to everyone, and then saying it again. It covers the first notice, sent inside five minutes with almost no information. It covers choosing the channels, including the problem of announcing an e-mail outage by e-mail. It covers a status page for a small organisation that has never had one. Then comes update cadence during an incident, with the interval by severity and the update that says there is no news. It covers the all-clear and the follow-up note the next day. And it covers planned-maintenance notices that people actually read. The incident itself, and the roles behind it, have a series of their own. This one is about the messages.
Articles in This Series
- The First Notice: Say Something in Five Minutes
What to say when you know almost nothing: what is affected, that you are on it, when the next update comes. Covers the three-line notice, who sends it, and the first notice to copy. - Choosing the Channels for Outage Messages
E-mail and what to do when e-mail is the outage, chat, the status page, a phone tree, signage, the receptionist; reach versus noise; and a channel matrix by outage type. - A Status Page for a Small Organisation: One Page, Kept Current
What a status page is, cheap ways to have one, and what goes on it and what does not. Covers keeping it current, the page that lies by being stale, and the status-page entry. - Update Cadence During an Incident: The Interval by Severity
The interval by severity, the no-news update, what each update contains, for every audience; and the cadence table with the update message. - The All-Clear and the Follow-Up: Declare It Fixed Only Once
Declaring it fixed only when it is, what users should do now, the short explanation, the apology, the after-note for management a day later, and both messages to copy. - Planned-Maintenance Notices People Actually Read
Advance notice and the reminder, timing, what to expect in plain words, the subject line, the maintenance that overruns, and the maintenance notice to copy.
Explore More Topics
This series is part of the QuerySolv helpdesk topic library. It covers the foundations, processes, communication, knowledge, tooling, metrics, classic ticket domains, people and strategy behind a support desk that works. The library pairs well with the software we build: QuerySolv. It is a self-hosted help desk that runs as one Windows service on your own network. It is licensed once per installation, not per agent per month.