Frequency of communication

With each notification, set an expectation for communication frequency. Tell your stakeholders the maximum allotted time before which they’ll receive another update. For example, saying “we will post an update within 30 minutes,” means that they might receive an update sooner if something about the situation changes. But if nothing changes, they will at least receive another update 30 minutes from now. In this example, it isn’t necessary to append text to say something like “or whenever new developments occur.” Make your updates as succinct as possible.

Updates should occur as often as necessary to communicate critical information to internal stakeholders. We recommend using 30 minute intervals, by default. The only consideration to watch out for here is making updates too frequently. Incidents take time to troubleshoot and resolve. Initial suspicions often turn out to be incorrect. Updates that occur too frequently can potentially undermine confidence rather than boost it.

If no updates are available, avoid reissuing the same generic message every 30 minutes if possible. For example, you might post a message that says:

We are still working to restore service X and do not have any additional updates at this time.

In 30 minutes, if nothing has changed, rather than posting that same message again you should consider what other details might be helpful for internal stakeholders. The next update might look something like:

We are continuing to work on restoring service X. In the meantime, advise any customers to use workaround Y until this issue is resolved.

You may not have identified a workaround during this time, nor are we suggesting you prioritize doing so instead of working to resolve the incident. But throughout the course of an ongoing incident, typically some helpful bit of additional information is discovered that can be useful if passed along to stakeholders. These updates, when the incident status still hasn’t changed, are good opportunities to do so.