The drive-by

Bring up “the drive-by” in any IT team and people will know exactly what you mean. Someone’s deep in something: multiple windows open, logs on one screen and notes on another. They’ve got the whole thing in their head: dependencies, sequencing, what might break, how they’d roll it back if it does. Then someone appears.

“Quick question.”

It’s harmless. It takes 30 seconds. Everyone’s polite. Everyone moves on. Except they don’t.

The person asking carries on with their day. The person who answered turns back to their screen, and stops. Because the thing they were holding in their head has gone. Not the task. The context. And the next few minutes aren’t spent doing the work, they’re spent trying to rebuild where they were before the interruption landed.

It looks normal. People collaborating, asking for help, work moving forward. But if you watch it closely, something else is happening. The work isn’t being paused. It’s being reset.

In IT, most of the work isn’t typing, it’s thinking. It’s holding a mental model of how systems connect, what depends on what, what happens if something changes, where the failure points are, and what recovery looks like. When that model breaks, you don’t just pick up where you left off. You start again.

IT professionals are quite good at managing infrastructure. We think about capacity, monitor usage, and design systems so they don’t fall over under load. But attention doesn’t work the same way. People treat it as infinite, as if it resets instantly, as if interruptions are free. They’re not.

Every interruption has a cost. It’s just not visible in a dashboard. And over time, that cost adds up: work takes longer than it should, people feel slower than they actually are, more gets deferred, less gets finished properly. Not because the team isn’t capable, because their capacity is being chipped away, one “quick question” at a time.

This isn’t about stopping people from talking to each other. It’s about recognising that attention behaves like any other limited resource: interrupt it constantly, and you get less out of it. We design systems to avoid running out of CPU, memory, or storage. We should probably start thinking the same way about the people running them.

Attention isn’t infinite. It’s infrastructure. And in most environments, it’s the first thing that gets exhausted.