How to Improve Focus (When Everything Is Designed to Break It)

If you've ever sat down to do focused work and found yourself checking your phone 12 minutes in — not because you decided to, but because you just did — the problem isn't your character. The problem is that your attention is being managed by systems designed by very smart people to capture and hold it, and you're fighting them without a counter-system.

Focus is not a fixed trait. It's a limited resource that depletes with use, takes time to build, and gets disrupted in specific, measurable ways. Understanding the mechanics of how focus works — and how it gets destroyed — is more useful than any generic advice about putting your phone in another room.

The Attention Residue Problem (Why Interruptions Cost More Than You Think)

In 2005, Gloria Mark, a professor of informatics at UC Irvine, published research that changed how organizational psychologists think about interruptions. Her finding: after being interrupted, it takes an average of 23 minutes and 15 seconds to fully return to the original task at the same depth of engagement. Not 5 minutes. Not "a moment." Twenty-three minutes.

This is what Mark calls attention residue — the cognitive traces of the interrupted task that persist in working memory and compete for attention even after you've "moved on." When you check your email in the middle of writing a report, the report doesn't disappear from your mental background. Your brain is now running two threads — the email and the report — and neither gets your full processing capacity.

The math on a typical workday is grim. Mark's later research found that office workers are interrupted or self-interrupt every 3 to 5 minutes on average. If each interruption costs 23 minutes of recovery, a workday structured around constant availability never generates any meaningful focused time at all — you're perpetually in the recovery gap.

The implication is significant: focus isn't just about starting a task. It's about creating a window long enough that the recovery period doesn't eat the whole session. A 20-minute work block isn't enough — the first 23 minutes of focus are essentially just getting back to where you were before the last interruption.

Shallow Work vs. Deep Work: The Distinction That Explains Most Productivity Problems

Cal Newport's framework from Deep Work (2016) is useful not because it's complicated but because it names something most people haven't explicitly articulated: not all work is the same, and treating it the same way is the source of most productivity failures.

Deep work is cognitively demanding, distraction-free work that pushes your intellectual capabilities — writing, analysis, coding, complex problem-solving, strategic thinking. It produces the output that creates real value, is hard to replicate, and requires sustained, uninterrupted attention to do well.

Shallow work is logistical, low-cognitive-demand tasks that can be done while distracted — answering emails, scheduling meetings, filling out forms, attending status-update calls. Shallow work feels productive because it's constant activity, but it's the kind of work that doesn't move the needle on your most important outcomes.

The problem is that most modern work environments are optimized for shallow work. Email is designed for constant availability. Open offices are designed for spontaneous collaboration. Notification systems are designed for immediate response. All of these design choices make deep work harder — and they do so on every person in the organization simultaneously.

Newport's prescription — designate specific, protected time for deep work and defend it rigorously — sounds obvious until you try to implement it in an environment where being available is socially and professionally expected. The implementation matters: a vague intention to "focus more" doesn't survive contact with your inbox. A scheduled, time-blocked deep work window with specific inputs blocked does.

The 90-Minute Ultradian Cycle: Your Natural Focus Clock

The human brain doesn't operate on a continuous availability model — it runs in cycles. Sleep researchers have documented 90-minute cycles during sleep (NREM/REM cycles), and physiologist Peretz Lavie identified analogous 90-minute ultradian rhythms in daytime alertness and performance. Cognitive neuroscientist Andrew Huberman has popularized this in productivity contexts: most people have a roughly 90-minute window of peak cognitive performance before their brain needs a recovery break.

This isn't a reason to only work 90 minutes a day — you can have multiple cycles. But it suggests that deep work sessions should be structured around 90-minute blocks followed by genuine downtime, rather than trying to sustain peak focus indefinitely through caffeine and willpower.

In practice: a single 90-minute deep work block in the morning — before checking email or social media — often produces more meaningful output than 4 hours of interrupted work scattered through the day. The first focus cycle of the day tends to be the strongest because decision fatigue and attentional depletion haven't accumulated yet.

The practical protocol: block 90 minutes in the morning for your most cognitively demanding work. No email, no Slack, no notifications. After the 90-minute block, take a 15 to 20-minute genuine break — walk outside, have a snack, don't look at a screen. Then return for a second block if needed. Three to four hours of deep work structured this way consistently outperforms eight hours of scattered, interrupted work on virtually every quality measure.

Why Open Offices and Notification Culture Destroy Focus (Specific Mechanisms)

It's common to hear that open offices and notifications are bad for focus, but the specific mechanisms are worth understanding — because the solution has to target the mechanism, not just the symptom.

Open offices: the mechanism. The core problem isn't noise — it's ambient social monitoring. In an open office, part of your brain's background processing is continuously tracking who is nearby, what they're doing, whether you're being watched, and whether someone might approach you. This background monitoring uses cognitive resources that should be available for focused work. Even with headphones in, the mere possibility of interruption maintains a state of partial vigilance that prevents the deepest levels of focus. Research by Ethan Bernstein and Stephen Turban at Harvard (2018) found that face-to-face interaction actually dropped by 70% in organizations that switched to fully open offices — people compensated for the forced proximity by withdrawing into headphones and digital communication. The office was open; the focus was worse.

Notification culture: the mechanism. Every notification — visual, auditory, or tactile — triggers an orienting response. The brain is wired to respond to novel stimuli as a survival mechanism; this isn't a willpower failure, it's automatic. But in a modern device environment, the orienting response fires hundreds of times per day. Each firing interrupts the current cognitive thread, even briefly, and initiates the attention residue process described above. A 2015 study by Stothart et al. found that merely receiving a notification — even one you don't look at — was enough to cause a significant increase in task errors. The notification doesn't have to be answered; its existence is enough to impair focus.

The implication for both: the solution isn't better willpower. It's structural changes that remove the stimulus before it triggers the response. Notifications disabled, not silenced. Deep work done in a different location from shallow work if possible. Phone physically out of the room — the Ward et al. (2017) study at UT Austin found that the mere presence of a smartphone on a desk reduced available cognitive capacity, even when turned face-down.

A 3-Tier Distraction Elimination Protocol

This protocol addresses distraction at three levels — environment, device, and workflow — in order of impact.

Tier 1: Environmental. Before a deep work session, remove or relocate every physical item that could trigger distraction. Phone goes in a different room or in a bag. Only the applications required for the current task are open on your computer. If you work in an open office, a "do not disturb" visual signal (headphones, a closed door if available, a table tent) signals to colleagues that you're unavailable. This isn't rudeness — it's the same contract as a closed-door meeting. If you have control over your schedule, guard your morning hours for deep work and consolidate meetings and shallow work in the afternoon.

Tier 2: Device and app level. Turn off all non-essential notifications at the OS level — not just on individual apps. Email, Slack, and social media should be checked on a schedule, not via push alerts. Set specific times (twice a day is the research-backed recommendation for email) and communicate them to colleagues. When doing deep work, use website blockers like Freedom, Cold Turkey, or built-in Screen Time/Focus modes to make distraction sites structurally inaccessible — not just inconvenient. Rely on commitment devices, not in-the-moment willpower.

Tier 3: Workflow design. Before each deep work session, write a single specific output goal: not "work on the report" but "write the methodology section, 600 words." This specificity reduces the cognitive load of deciding what to do and makes progress measurable. When a distracting thought arrives (and it will), have a capture system — a sticky note or open notepad — where you write it down and return to your task. The thought is acknowledged and stored; you don't have to hold it in working memory or act on it immediately.

Combine all three tiers, and a 90-minute session produces the kind of output that used to take a full day. The math isn't magic — it's simply protecting the cognitive resource that was being stolen by incremental interruptions before.

Building Your Personal Focus System

A focus system is a set of recurring commitments and environmental designs that make deep work the default rather than the exception. The following components, taken together, form a minimal effective version:

A fixed deep work window. Pick a time and defend it. Morning usually works best for most people before attentional resources are depleted, but the specific time matters less than its consistency. The brain learns to enter focus mode at the same time each day when the routine is reliable — similar to how sleepiness is conditioned to a regular bedtime.

A pre-work ritual. A brief, consistent sequence of actions before your deep work block signals to your brain that focus is about to be required. It doesn't need to be elaborate — five minutes of reviewing your task, making coffee, putting on specific music or silence. The ritual is a neurological trigger, not a productivity theater exercise.

An end ritual. Cal Newport's "shutdown complete" ritual — a specific end-of-day sequence that includes reviewing unfinished tasks and making a plan for them — reduces the attention residue that bleeds into evenings and weekends. When your brain trusts that unfinished tasks have a written plan, it stops cycling through them as a background process. You actually rest better.

Weekly calibration. Once a week, review how much deep work you actually completed versus planned. Over time, this data tells you whether your system is working and what's disrupting it. Most people find one or two recurring friction points — a recurring meeting that lands in their focus window, a colleague who habitually interrupts at 10am — that are easy to address once visible.

Focus is a resource, not a trait. You can't improve it through intention alone — you improve it by protecting the conditions that allow it to exist. Remove the stimuli before they trigger the response, work in windows aligned with your natural cognitive rhythms, and build the structural commitments that make deep work the default. The output that matters most comes from those protected windows. Build the architecture around them.