Focus failure mode · Interrupt chasing
Your attention is interrupt-driven, and the real cost is not the 23 minutes everyone quotes
You do not lose your focus. Something takes it, and you hand it over, because handing it over is what a responsive colleague does.
Interrupt chasing is attention governed by incoming signals rather than by intention. The popular claim that an interruption costs 23 minutes traces to a 2006 interview rather than a peer-reviewed paper. Gloria Mark's published research found interrupted people work faster but with significantly more stress, and that interrupted work is typically picked up again about 25 minutes later. The reflex persists because responding is socially rewarded, immediately gratifying, and indistinguishable from doing your job.
What it looks like
The badge appears and you read it before deciding to. You know the message does not need an answer for four hours. You answer it in ninety seconds. Between 9am and lunch this happens eleven times and every single instance is defensible on its own. Nobody at your company will ever tell you to stop, because your interruptions are wearing a lanyard.
The mechanism
Almost everything written about interruptions quotes 23 minutes and 15 seconds. That number comes from a 2006 Gallup interview with Gloria Mark rather than from a peer-reviewed paper. The research behind it is real, but the peer-reviewed figure from the same body of work is 25 minutes 26 seconds, and the precise 23:15 does not appear in the study it usually gets attributed to. The published work is more interesting. In her controlled experiment with Daniela Gudith and Ulrich Klocke, interrupted people completed their work faster than the uninterrupted control group, with no drop in quality, but reported significantly higher stress, frustration, time pressure and effort. In her field observations with Victor Gonzalez and Justin Harris, interrupted work was resumed the same day after an average of 25 minutes 26 seconds, having passed through more than two other tasks first, though the spread was enormous. Underneath the arithmetic is a reward loop: each notification is unpredictable in content and timing, which is the schedule that produces the most persistent checking behaviour in behavioural research going back to Skinner. On top of that, answering quickly is visibly rewarded in every workplace dashboard, so the habit is reinforced socially as well as neurologically. It is a hard combination to argue with at 11am on a Tuesday.
The honest part
A significant share of this is fixable for free tonight, and it would be dishonest to sell you something before saying so. Turn off badge counts, which are the cheapest attention theft in software. Set macOS Focus modes to silence everything except the two people who genuinely need to reach you. Take email off your phone's notification list entirely. Most people who do only these three things report a real improvement within a week, and no product was purchased.
What the free fixes cannot reach
Notification settings govern what arrives. They do nothing about the interruptions you go and fetch: the quick check of the inbox tab, the glance at the Slack tab that is already open in your browser. Those carry the same resumption cost and no operating system setting sees them, because you initiated them. This is also why batching matters. Two or three defined windows for communication removes the reason to check between them, and the reflex loses its food supply within about a fortnight.
Where Hugo fits
During a session Hugo holds the tabs that are not serving the task you named, including the ones that ping. It is the browser-side half of the problem, which is the half Focus modes cannot cover. Start with the free settings first. If the self-fetched checks are still eating your afternoons after that, this is the layer that catches them.
Common questions
How long does it really take to get back to work after an interruption?
Less cleanly than the internet suggests. The famous 23 minutes 15 seconds comes from a 2006 Gallup interview rather than a peer-reviewed paper. The field study behind it is real; the published figure is 25 minutes 26 seconds. The published field figure is an average of 25 minutes 26 seconds to resume interrupted work, from Mark, Gonzalez and Harris (CHI 2005), with a standard deviation of nearly 55 minutes, so the average hides enormous variation. More useful is what the controlled study found: interrupted people worked faster but paid for it in stress, frustration and effort. The cost of a two minute reply is real, but it shows up in how the afternoon feels rather than on a stopwatch.
Do notification settings actually help?
Substantially, and they are free. Disabling badge counts, restricting macOS Focus modes to a small allowlist of people, and removing email notifications from your phone address most externally triggered interruptions. What they cannot address is self-initiated checking, which for many people becomes the larger source once the external ones are silenced.
What is batching and does it work?
Batching means handling communication in two or three defined windows a day rather than continuously. It works because it removes the ambiguity that justifies each check. Expect roughly two weeks of discomfort while colleagues recalibrate their expectations, and tell them the schedule rather than going quiet without explanation.