What interruptions cost, and how to reduce tool switching

What interruption research actually tells us

Interruptions do not have one universal recovery time. The cost depends on the task, the interruption and the person. A peer-reviewed CHI study by Gloria Mark, Daniela Gudith and Ulrich Klocke found that interrupted participants completed work faster, but reported more stress, frustration, time pressure and effort. That is a more useful finding than a single headline statistic: interruptions change both performance and how work feels.

Separate a notification from a necessary lookup

Many project questions span more than one system. A decision might begin in Gmail, continue in Slack and finish in a GitHub pull request or Linear issue. Searching each service separately adds steps and makes it easier to lose the original task.

Noet is designed to reduce those search steps. Its integration pages document what can be retrieved from Gmail, Slack, GitHub and Linear. From the Mac app, a user can ask a question with ⌥Space and review links back to the source material.

Reduce the steps in one recurring workflow

This does not remove interruptions, and it does not guarantee better focus. It gives you a shorter retrieval path when the information you need is distributed across connected tools. Keep notifications under control, batch non-urgent communication and use cross-tool search when a question genuinely spans systems.

Choose a repeatable task, such as checking the decision behind an onboarding PR before review. Write down the sources you normally inspect: the issue, the email setting the requirement, and the Slack discussion. Ask a question naming the project and timeframe, then compare the sources returned with your usual search. Check both the quality of the evidence and the number of steps needed to reach it.

Measure your own result

The practical test is simple: note how many separate searches a recurring workflow requires, then compare that with a single cross-tool query. Measure the steps and time for your own work rather than relying on a universal productivity claim.