ManageTime

Article

Interrupting yourself at work: the distraction nobody else caused

You interrupt yourself when the task stops holding you, and it does that in two opposite directions: it turns hard enough to feel stuck, or easy enough to feel dull. A 2019 study of 151 employees found people broke off more often when they were bored, overstretched, or short of energy for the task in front of them. This article covers what the research says about why you leave a task, what it honestly cannot tell you about how often that happens, and how to catch your own trigger before it fires.

What makes you break off a task halfway through?

Halfway through the difficult paragraph, your hand goes to the inbox. Nobody messaged you. No alert fired. The work got harder, you left, and at the time the leaving registered as nothing at all. That moment has a research literature behind it, and the useful part of that literature is that the pull has more than one cause.

Adler and Benbunan-Fich set out a typology of self-interruptions in Computers in Human Behavior in 2013, built on flow theory and self-regulation theory. Their argument is that immersion breaks when what a task demands and what you can bring to it fall out of balance, and that the imbalance runs both ways. Negative feelings about progress, the sense of being stuck in something too hard, push you off. Positive feelings do it too: a task easy enough to feel already handled leaves your attention with nothing to grip.

Both ends of that scale send you to the same inbox.

Seipp went at the situational side directly in Performance Enhancement and Health in 2019. Using the critical incident technique, 151 employees each described one recent situation: either a moment they interrupted themselves, or a stretch of focused work. Self-interruption came up more often when energy was low, when the person was bored or overstretched, and when the situation carried a social character. Focused work ran the other way, appearing in situations people called dutiful and intellectually stimulating.

That social finding is the one most likely to describe your afternoon. A turn toward a colleague or a chat window counts as a self-interruption in this research, and it does not feel like avoidance while it is happening. It feels like being sociable. Which is why it survives every rule you have ever written about notifications.

Read the method before you lean on the result. Seipp's employees were recalling one incident each, and Adler and Benbunan-Fich watched students switch tasks inside a purpose-built program rather than anyone getting through a Tuesday. Neither study followed a real workday from the outside. What survives that caveat is the shape rather than any size: the two triggers are opposites, so a single fix aimed at willpower is aimed at the wrong thing. Boredom and difficulty do not answer to the same move.

Is interrupting yourself worse than being interrupted by someone else?

Nobody can tell you yet. Puranik, Koopman and Vough reviewed 247 publications on work interruptions for the Journal of Management in 2020, and the problem they open with is definitional. The research sits across disciplines that do not define an interruption the same way, and by the authors' own account it has stayed scattered, with little integration between those fields. Their review exists to propose one integrative definition, not to settle a scoreboard between types.

So treat any confident ranking you meet as running ahead of its evidence. That includes the figures that circulate about what share of your interruptions you cause yourself. This article is not going to hand you one, because the studies behind it do not carry a number anyone measured on a workday like yours.

One difference is practical rather than measured. An external interruption arrives at a moment somebody else picked; a self-interruption arrives at a moment you picked, and that is the whole of its advantage. You cannot schedule your colleague walking over. You can notice, twice in an afternoon, that the pull came exactly as the work turned hard.

There is also a case for leaving some of them alone. Seipp's conclusion from those 151 employees was that internal interruptions can work as an energy management strategy: when a task is straining, stepping away may help you replenish and come back to the goal rather than pulling you off it. That is an interpretation drawn from the framework, not a tested effect, so hold it loosely. Hold it, though, because the opposite reading, that every departure is a failure of discipline, is not what this evidence says either.

Some of those departures are the recovery.

What none of this settles is what coming back costs you, which is a separate question with its own literature. The focus pillar takes the environment side of the problem first, on the argument that a default you change once beats a decision you have to make every morning. This article is about the half no default reaches.

How do you stop switching away from a task before it is finished?

Not with a better blocker. A self-interruption starts inside the task rather than outside it, so a setting never gets a look at it; by the time your hand moves, the decision has already been made. What is available to you is what happens in the two seconds after you notice.

Find out which trigger is yours first, because it is probably not both in equal measure. For the next fortnight, each time you catch yourself leaving a task, write one word: stuck, or dull. Nothing else. No timestamp, no analysis. Fifteen of those words will tell you more about your own pattern than any figure in this article, and each one costs about a second.

If most of them say stuck, the trouble is that the next step has nothing to take hold of. A task written down as finish the proposal gives you nowhere to start; the same task written as draft the two sentences summarising the budget section does. Getting Things Done is built on that move, defining the next action as the concrete physical thing you would do next, and it is the part of the method worth borrowing even if you never assemble the rest of it.

If most say dull, the pull is toward stimulation rather than away from difficulty, and a container handles that better than a rule does. The Pomodoro Technique earns its place here for an unglamorous reason: a 25-minute sprint has a visible end, so holding out for it is a promise sized to be kept on a dull afternoon. The focus pillar already covers what to do with whatever surfaces mid-sprint, so that move does not need a second system built around it here.

One warning covers all three paragraphs above. None of the studies behind this article tested an intervention. They described where self-interruptions come from; nobody randomised anyone into writing a word a day, or into a sprint, and then measured what changed. These are readings of the evidence applied to a desk, which is a weaker thing than a result, and they are worth two weeks of your own data before they are worth anything more.

Write the word. Read the words on Friday.

How does your working style change the answer?

None of these studies asked how people organise a week, so what follows is this site's reading rather than a research finding. The four working styles describe how you already operate when nobody is supervising, and the moment you leave a task looks different inside each of them.

If you build the week in advance, the Architect's pattern, your self-interruptions tend to be the productive-looking kind. You leave the hard task to tidy the plan, reorder the list, or set up tomorrow's block. It reads as work, and it is work; it is just not the work that was in front of you. Expect your one-word log to say stuck more often than you would have guessed.

Burst workers, closest to the Sprinter, mostly leave at the top of the curve rather than the bottom. The piece turns easy, the sense of having got somewhere arrives, and attention goes looking for the next thing while the last fifth of the job sits there unfinished. That is the positive-feelings trigger in Adler and Benbunan-Fich's typology, and what it asks of you is finishing the sprint you started instead of starting a better one.

The Visionary leaves toward a more interesting problem. Nothing about that is idle: the thought that pulls you away is often the good one, and losing it costs something real. Write it down, then stay.

An Improviser's day rarely offers a clean stretch to protect, so the trigger to watch is the social one Seipp's study picked up. When the room is the source of most of what is good about your day, stepping toward it is not a lapse, and treating it as one only adds guilt to a day that was already in pieces. Pick one task a day to finish without leaving. One is enough to learn from.

So the next step is a small one. Catch two departures tomorrow, write stuck or dull beside each, and leave the rest of your setup exactly as it is until Friday. If you are not sure which of these styles is yours, the working style assessment takes a few minutes.

Frequently asked questions

Why do I keep interrupting myself at work?

Because the task stopped holding you, in one of two opposite directions: it turned hard enough to feel stuck, or easy enough to feel dull. A 2013 typology traces both to the same imbalance: what the task demands against what you can bring to it. A 2019 study of 151 employees added the situational half. People broke off more often when energy was low, when they were bored or overstretched, and when something social was available.

Is interrupting yourself worse than being interrupted by someone else?

The research has not settled it, and a 2020 review of 247 publications explains why: the field defines an interruption inconsistently enough that findings conflict. What is practical rather than proven: a self-interruption happens at a moment you chose, which makes it the one you can move. Your colleague walking over is not on your schedule. Your own departure is.

How do you stop switching away from a task before it is finished?

Name the trigger before you change anything: for two weeks, write one word, stuck or dull, each time you catch yourself leaving. The two want opposite answers. Stuck wants a smaller next step, written concretely enough to begin. Dull wants a container with an end close enough to sit still for. A blocker catches neither, because the decision was made inside the task.

How long does it take to break a self-interruption habit?

No study behind this article measured that, so treat any specific timeline you are offered as unsupported. This site works to two weeks, which is long enough to hold one ordinary busy week and one quieter one. The point of the fortnight is the log rather than the streak. You are looking for which trigger repeats, not for a run of clean days.

Sources