Great Circle Waypoints Blog

Having a documented incident management process and having an incident management capability are not the same thing. Many companies have only the first.

A senior leader drops “let’s try to get this resolved in the next 10 minutes, please!” into the incident channel, meaning it as encouragement. It usually backfires: exhortation adds anxiety, not speed. Urgency helps a response only when it arrives as information.

Most of the time, simultaneous incidents are no big deal. But once in a while, they collide, and the individual incident commanders can’t sort it out among themselves. A coordination pattern called Area Command, adapted from the Incident Command System, fills the gap.

When an incident pulls in too many active responders, your experienced people quietly break into smaller groups. Here’s how to build your process around that instinct.

Modern software architectures are designed so nobody has to understand the whole system. The hardest incidents are when that strategy fails, and generating that understanding on demand takes structure: incident roles, a dedicated tech lead, and communication practices built for synthesis.

AI coding tools are producing real velocity gains, but they’re also thinning the shared understanding teams have of their codebases. That gap is manageable during normal development. During an incident at 3am, it’s where the cost shows up.