I'm a DevOps engineer and our team handles incident response communication through Slack. It's fantastic for real-time interaction when things go wrong, but it falls flat when it comes to follow-up tasks once the incident is resolved. We typically create a Slack channel during an incident, gather five team members, and after fixing the issue in a couple of hours, we list follow-up tasks like updating runbooks or addressing root causes. Everyone agrees on who will tackle these tasks, but when the dust settles, only about one out of five tasks gets completed within two weeks. While team members are well-intentioned, they often forget these tasks as the incident channel gets buried under a pile of new conversations, and there's no reminder system in place. We've tried using Jira for these follow-ups, but creating tickets while trying to resolve an issue at 3 AM is impractical. We end up saying we'll create them later, but those intentions often fall through when sleep-deprived. For our on-call reliability, actually completing these follow-ups is crucial, and we need a better way to hold ourselves accountable without adding too much complexity to our response process.
5 Answers
It sounds like you could really benefit from a structured post-mortem process. After resolving an incident, gather the channel discussions and write a post-mortem report with the team involved. This should include an action items section that pulls directly from the Slack chat. Assign Jira tasks in this documentation, and it's then the manager’s responsibility to ensure these items are tracked and completed. Don't rush the documentation right after the incident; it can be written the next day when the details are fresh. Keep the Slack channel open until all the follow-up tasks are finished to maintain visibility.
We're using a tool called incident.io, which helps summarize Slack discussions after the fact and suggests action items, assigning them to team members. It can feel like a hassle at times, but it actually does a decent job if configured right. Just remember, beyond the tools, cultivating a culture where everyone takes responsibility for follow-ups is key. It's not an easy shift, but it's necessary for accountability.
Absolutely! It’s all about finding the right balance between tools and team buy-in. You can have the best tech, but if the team doesn’t engage with it, it won’t work.
Have you thought about scheduling a post-mortem meeting at a time when the team is alert? It can be really beneficial to discuss the incident thoroughly when everyone's fresh. Also, I’d recommend not closing the channel too quickly. Slack channels can stay open indefinitely. Keeping it alive lets team members track progress on tasks and communicate updates, plus it makes it easier for someone to organize all follow-ups, like creating Jira stories.
You might want to check out [go.dev](https://go.dev/) as a starting point to gather tools that can help streamline your process. There's a ton of potential out there to improve incident management.
We’ve started using a tool called Chaser specifically for incident follow-ups. It lets us create tasks directly within the Slack channel during the incident, which means they stick around and we get reminders afterward. Our completion rates are way better than what we experienced with the old system. Definitely worth looking into!
I get where you're coming from about AI tools. It can definitely feel clunky sometimes, but it's helpful to know that many features can be set up to fit your team’s style. If something feels off, definitely reach out for support—they usually have options to customize how reminders work, which could ease some of the frustrations.