Two kinds of focus failure
There are two ways to lose a workday, and they feel completely different from the inside.
The first is distraction. You open Twitter to check one thing and surface forty minutes later. You watch a YouTube video that autoplays into three more. This is the version of focus failure that gets all the attention. Every productivity article warns you about it. Every focus tool is built to stop it.
The second is fragmentation. You have five legitimate things competing for your attention. A presentation due Thursday. A client email that needs a thoughtful reply. A project plan that's half-finished. Code that needs to ship. None of these are distractions. All of them are real work. All of them feel urgent.
So you start the presentation, get two slides in, and remember the email. You open the email, draft half a response, and realize you need information from the project plan. You open the project plan, start updating it, and a Slack message pulls you to the document review. Ten minutes later, you go back to the presentation. But now you've lost your train of thought.
At no point during this sequence did you do anything wrong. You didn't open Reddit. You were working the entire time. And yet, by the end of the day, the presentation has two slides, the email is still a draft, and the code is untouched. You were genuinely busy. You have genuinely nothing to show for it.
The distinction that matters
Distraction pulls you to something that isn't work. Fragmentation bounces you between things that are all work. Most focus tools only solve the first problem. For most knowledge workers, the second one is worse.
What the research actually says
In 2005, Gloria Mark, Victor Gonzalez, and Justin Harris published "No Task Left Behind?" at CHI. They shadowed information workers and found that people switched activities roughly every three minutes. Not between work and distraction. Between work tasks. The average worker had 2.3 intervening tasks before returning to the original one.
Three years later, Mark, Gudith, and Klocke (2008) found something counterintuitive: interrupted workers actually completed tasks faster than uninterrupted workers. But at a cost. The interrupted workers reported significantly higher stress, frustration, time pressure, and effort. They were getting things done, technically, but at a cognitive price that compounded across the day.
From the outside
Looks productive
Faster task completion, more tasks touched, always responsive to messages and requests.
From the inside
Costs everything
Higher cortisol, burnout accumulation, lower quality output, less creative thinking, the feeling of running all day and going nowhere.
Then came the attention span data. Mark et al. (2016) tracked attention in real time and found the median time people spent on a single screen before switching was 40 seconds. Gloria Mark later documented the trajectory in her 2023 book Attention Span: average screen attention dropped from about 2.5 minutes in 2004 to 47 seconds. The direction is consistent and has not reversed.
But the finding that should genuinely concern you comes from Sophie Leroy's 2009 paper in Organizational Behavior and Human Decision Processes. Leroy introduced the concept of attention residue. When you switch from Task A to Task B, part of your attention stays stuck on Task A. Your brain doesn't context-switch cleanly. It leaves residue.
Attention residue in practice
Every time you bounce between tasks, you carry fragments of the previous task into the next one. The more you switch, the more residue accumulates. By the afternoon, you're operating at a fraction of your actual capability. Not because you're tired, but because your attention is smeared across a dozen half-finished things.
Why the obvious solution doesn't work
The focus tools market is built almost entirely around one idea: block the bad websites. Make a list of distractions, enforce that list during work hours, problem solved.
For pure distraction, this works. If your problem is that you open Twitter twelve times a day, putting it on a list is a reasonable approach.
But fragmentation is a different animal entirely. You can't put your second project on a blocklist. You can't label your inbox as a distraction when you genuinely need to respond to three of the twenty emails sitting there. You can't tell a tool to block Notion when Notion is both your project management system and the thing you're using to avoid harder work.
The entire blocklist model assumes your distractions are identifiable in advance, consistent across sessions, and clearly separable from legitimate work. For the fragmentation problem, none of those assumptions hold. Your "distractions" are your other obligations.
"Instead of a single flow of work, we live in a permanent state of partial attention."
RescueTime
The environment design principle
Cal Newport argues in Deep Work that the ability to perform deep, concentrated work is becoming simultaneously rarer and more valuable. But he goes further than naming the skill. He argues that deep work is not primarily a question of discipline. It's a question of environment.
There's a story about Victor Hugo. He was behind on a deadline for The Hunchback of Notre-Dame. So he reportedly had his valet lock away his formal clothes, leaving him with nothing suitable to wear outside. Unable to go out, Hugo had no option but to sit at his desk and write. He finished the novel ahead of schedule.
The insight
Hugo recognized that self-control was a finite resource that would fail him when he needed it most. So instead of relying on it, he changed his environment to make the desired behavior automatic and the undesired behavior structurally difficult. This is environment design. And it's the only approach to focus that scales.
Gazzaley and Rosen (2016) make a complementary argument in The Distracted Mind. We cannot actually multitask. What we call multitasking is rapid switching, and every switch carries a cost. The solution is not to fight the bottleneck with effort. It's to design your information environment so that the bottleneck is never triggered.
Kushlev and Dunn (2015) demonstrated this concretely with email. One group checked email freely. The other was limited to three times per day. The limited group reported significantly less stress. Not because they had fewer emails. Because the environment was structured so that the constant pull of "should I check?" was removed. The decision was made for them.
The internal trigger problem
If environment design is the answer, why hasn't it worked already? Everyone knows the advice. Turn off notifications. Close extra tabs. Put your phone in another room.
The trigger is almost never external. It's internal. We don't switch tasks because a notification pops up. We switch because of a feeling: boredom, anxiety, uncertainty, the low-grade discomfort of sitting with a hard problem that hasn't yielded to our efforts yet.
"Opening Reddit or Hacker News was the path to a surge of dopamine. Whereas getting down to work didn't have the same surge."
swalsh, Hacker News
You're two paragraphs into a difficult email and you feel a pull toward something easier. Not because something beeped. Because the email is hard, and your brain is looking for relief. Opening a new tab is the escape hatch. Not to Twitter, necessarily. Just to anything that provides a momentary hit of novelty, progress, or control.
Any real solution to fragmentation has to address both sides: the external environment (what's available to switch to) and the internal trigger (the moment of discomfort that initiates the switch). Blocking the external targets is necessary but not sufficient. You also need something that catches you in the moment of switching and makes you conscious of what you're doing and why.
The commitment problem
There's a deeper psychological dimension to fragmentation: choosing to focus on one thing means choosing to let everything else wait. And that feels dangerous.
When you have five things on your plate and you commit fully to one of them, the other four don't disappear. They sit in the background, accumulating urgency. The anxiety of uncommitted tasks is real, and it's often what drives the switching behavior in the first place.
The knowledge worker's environment makes this worse. Slack messages expect fast responses. Email norms reward quick replies. The culture of "responsiveness" treats delayed responses as a character flaw. Going heads-down for two hours feels like a risk.
"Building systems that require no motivation/discipline/willpower."
saulpw, Hacker News
The solution to the commitment problem is not a mindset shift. It's structural. You need an environment where committing to one thing is the default state, where the decision to focus is made once at the beginning of a session rather than re-made every forty seconds for the rest of the day.
What an intelligent environment looks like
If the solution is environment design, and if blocklists are too crude to address fragmentation, what does the right environment actually look like?
- It understands context. The same website can be on-task or off-task depending on what you're working on right now. The environment needs to know what you're trying to accomplish and evaluate everything else against that.
- It operates at the point of switching. Not before it. The environment needs to catch you at the moment you're about to switch and create a checkpoint where you have to consciously decide whether this switch is worth the cost.
- It addresses internal triggers. When you feel the pull toward something else, the environment should surface that pull to your conscious awareness. Not by lecturing you. By asking: is this what you're supposed to be doing right now?
- It makes commitment the default. You decide what you're working on once. From that point, staying on task is automatic. Leaving requires effort. The decision is inverted.
- It runs continuously. A 25-minute Pomodoro timer doesn't solve fragmentation. An effective environment adapts as your work shifts, so the space between sessions doesn't become a gap where fragmentation floods back.
What I built
I should be honest about why I wrote this essay. I'm the founder of Hugo, and Hugo is my attempt to build the kind of environment I just described.
Hugo is a macOS app. At the start of a session, you tell it what you're working on. "Writing the Q2 board deck." "Debugging the authentication flow." From that point, Hugo evaluates every new tab and every desktop app against that stated goal, in real time, using AI.
Stack Overflow while debugging? Allowed. Stack Overflow while browsing random questions during a writing session? Caught. YouTube tutorial on the API you're integrating? Allowed. YouTube rabbit hole on mechanical keyboards? Caught. The same site, different treatment, depending on context. No list to build.
When Hugo catches something off-task, it asks you to explain why you need it. That justification prompt is the intervention at the point of switching. Having to articulate why you need something to achieve what you said you were trying to do is usually enough to break the autopilot. You don't need a locked door. You need a mirror.
Hugo is free to try at tryhugo.app. The free tier includes two sessions per day. Pro is $12 a month or $99 a year.
The cost of not solving this
The fragmentation problem is not a personal failing. It's a structural consequence of how modern knowledge work is organized. We gave everyone an open internet, real-time communication tools, and five projects at once, and then acted surprised when they couldn't concentrate.
The people most affected by fragmentation are often the most conscientious. They bounce between tasks because they care about all of them. They check email constantly because they feel responsible. Fragmentation is not a symptom of laziness. It's a symptom of having too many legitimate demands and no structural support for prioritizing between them.
The answer is not more discipline. It is better environments. Whether you build that environment with a tool, a ritual, a physical space, or a valet who locks away your clothes is less important than the recognition that you need to build it at all.
The modern default is fragmentation. If you don't actively construct something different, fragmentation is what you'll get. And you'll be busy all day. And you'll have nothing to show for it.
References
Gazzaley, A., and Rosen, L. D. (2016). The Distracted Mind: Ancient Brains in a High-Tech World. MIT Press.
Kushlev, K., and Dunn, E. W. (2015). Checking email less frequently reduces stress. Computers in Human Behavior, 43, 220-228.
Leroy, S. (2009). Why is it so hard to do my work? Organizational Behavior and Human Decision Processes, 109(2), 168-181.
Mark, G. (2023). Attention Span. Hanover Square Press.
Mark, G., Gonzalez, V. M., and Harris, J. (2005). No task left behind? CHI 2005, 321-330.
Mark, G., Gudith, D., and Klocke, U. (2008). The cost of interrupted work. CHI 2008, 107-110.
Mark, G., Iqbal, S. T., Czerwinski, M., Johns, P., and Sano, A. (2016). Neurotics can't focus. CHI 2016, 1739-1744.
Newport, C. (2016). Deep Work. Grand Central Publishing.