Why Interruptions Cost More Than Their Length

An interruption that lasts half a minute can cost far more than half a minute of working time. The reason lies in what has to be rebuilt afterwards rather than in the interruption itself.

What is actually lost is the arrangement in your head

Complex work involves holding several things in mind at once: where you are, what you tried already, and what the next two steps depend on. None of that is written down anywhere.

An interruption forces that arrangement out. The information still exists in your notes and your file, but the assembled version, the one you were actually working from, has to be built again from parts.

Rebuilding takes longer than reading, because you are not retrieving a fact but reconstructing relationships between facts. That is why returning to a task feels slow even when you remember exactly where you stopped.

Resumption is slower when the task is unfinished mid-thought

Interruptions that arrive at a natural boundary are cheap. If you have just finished a section and are about to start another, there is little state to lose and little to rebuild.

Interruptions that arrive mid-sentence or mid-calculation are expensive, because the partial structure was entirely in working memory and nowhere else. The cost depends heavily on timing, not just duration.

This is why people who are interrupted often develop the habit of finishing a thought aloud or writing a fragment before turning away. They are converting a fragile mental state into a durable external one.

Anticipated interruptions damage the stretch before them

An interruption you are expecting affects the period leading up to it. A meeting in twenty minutes tends to make the preceding twenty minutes shallow, because you avoid starting anything you cannot finish.

The result is that a single scheduled event can degrade a much larger block of time. The visible cost is the meeting; the invisible cost is the fragment of morning it renders unusable.

People often describe such days as busy without being able to say what consumed them. The consumption happened in the gaps, which are too short to log and too shallow to produce anything.

Self-interruption behaves the same way

Checking a message you were not sent is structurally identical to being interrupted by it. The state is displaced whether the trigger came from outside or from your own restlessness.

Self-interruption tends to rise as a task becomes difficult, which is precisely when the mental arrangement is most elaborate and most expensive to rebuild. The impulse arrives at the worst moment by design.

Reducing the number matters more than shortening each one

Because the cost is dominated by resumption rather than duration, making interruptions briefer achieves little. Two quick questions cost roughly twice as much as one long one.

Batching is effective for the same reason. Answering four messages in one pass involves a single rebuild, while answering them as they arrive involves four, spread across the work they interrupted.