Deep Work for People Who Can't Focus for 90 Minutes

Somewhere in the last decade, a specific number attached itself to focus: 90 minutes. Block off 90 uninterrupted minutes, the advice goes, and you'll drop into deep work — the implication being that anything short of that doesn't really count. For a lot of people, this advice produces a strange result: not more focused work, but a new reason to feel like they're failing at focus, since a 90-minute uninterrupted block is, for many actual jobs and actual lives, close to impossible to find.

It's worth separating the popularized rule from the research it's loosely based on, because the real findings are both less tidy and more usable than "90 minutes or it doesn't count."

This follows the same pattern as a few other popularized numbers already examined on this blog — the 21-day habit myth traced back to an unrelated clinical observation, the specific claim built on a much more nuanced original finding. The 90-minute deep work rule is a close cousin of that pattern: real research at the root, a much tidier and more demanding number by the time it reaches a productivity article.

Where the 90-minute number actually comes from

The number traces back to research on ultradian rhythms — cycles shorter than a day — and specifically to physiologist Nathaniel Kleitman's mid-20th-century work on what he called the Basic Rest-Activity Cycle, originally observed in sleep research as roughly 90-minute cycles between different sleep stages. Later work extended the idea to waking alertness, suggesting attention and energy might follow a similar rhythm during the day.

This is real, legitimate research — but notice how much distance exists between "sleep researchers observed roughly 90-minute cycles in some physiological measures" and "block off exactly 90 minutes for deep work." The original research describes an approximate, individually variable biological rhythm. The popularized version turned it into a fixed productivity rule, the same way the real habit-formation research got flattened into "21 days" — a specific, memorable number doing more work than the underlying science actually supports.

A rhythm observed in sleep research became a rule imposed on a calendar. The number survived the trip. The nuance didn't.

The research that actually explains why focus breaks

A more directly useful finding comes from organizational psychologist Sophie Leroy's 2009 research on what she termed attention residue. Leroy's studies found that when people switch from one task to another before finishing the first, a portion of their attention stays stuck on the unfinished task — measurably impairing performance on the new one. The effect was strongest specifically when the first task was left incomplete and open-ended, and notably reduced when people had a clear stopping point or had made a specific plan for returning to it.

This reframes the actual problem. It's not that a focus session shorter than 90 minutes is inherently insufficient. It's that a focus session interrupted by task-switching — a notification, a quick email check, a context switch to answer a message — carries a real, measurable cost that has nothing to do with the total minutes logged. Someone who protects 25 genuinely uninterrupted minutes is, by this research, likely better off than someone who "blocks" 90 minutes while actually switching tasks four times within it.

A concrete comparison makes this vivid. Person A blocks 90 minutes, but checks a notification at minute 20, answers a quick message at minute 45, and glances at email at minute 70 — three switches inside one long block, each one leaving attention residue behind. Person B protects a genuine 25-minute block with the phone in another room, no notifications, one task, start to finish. Person A logged more total minutes. Leroy's research suggests Person B likely did more actual focused work, because the cost isn't measured in minutes — it's measured in how many times attention got split and how much of it stayed behind each time.

Why this matters more than the length of the block

Attention residue research suggests the actual lever worth pulling is protection from switching, not duration for its own sake. A shorter block with zero task- switching produces less residue than a longer block with several switches inside it — which means someone who genuinely can't clear 90 uninterrupted minutes isn't locked out of deep work. They just need to be more precise about protecting whatever length of block they can actually get.

What this means if 90 minutes isn't realistic

Protect the block you can actually get, completely. Twenty-five genuinely uninterrupted minutes — phone in another room, notifications off, one task — is doing more, by the attention residue mechanism, than a 90-minute block that gets checked twice.

Give an interrupted task a specific stopping point before switching away from it. Leroy's research found that residue was reduced when people had a clear plan for returning to the unfinished task. A one-line note — "next: fix the header, then test" — measurably lightens the cognitive load an interruption leaves behind, compared to simply abandoning the task mid-thought.

This is a small habit with an outsized effect precisely because it targets the actual mechanism the research identified, rather than the symptom. The instinct when interrupted is usually to just switch and deal with the original task "whenever" — but "whenever" is exactly the open-ended, unresolved state Leroy's studies found carried the most residue. A specific next step, written down before switching, closes that open loop just enough to reduce what follows you into the next task.

Treat 90 minutes as a ceiling to work toward, not a floor to start from. If the only version of focus that "counts" requires conditions most people's actual days don't have, the likely outcome isn't more deep work — it's more days where deep work feels unreachable and gets skipped entirely.

There's a longer-term version of this worth naming too. Someone who consistently protects a real, uninterrupted 25 minutes is, over weeks, building a genuine track record of focused work — the exact kind of repeated, observable evidence that self-perception research suggests actually shapes identity, discussed at more length elsewhere on this blog. Someone who keeps failing to hit an unrealistic 90-minute bar is building a different, less useful kind of evidence — a repeated experience of falling short of a number that was never well-suited to their actual day in the first place.

A note on individual variation

It's also worth being honest that ultradian rhythms, to the extent they do apply to daytime attention, vary significantly between individuals — the original sleep research itself found meaningful variation in cycle length from person to person. A fixed 90-minute rule applied universally was always going to fit some people better than others, independent of anything about their discipline or capability. Noticing your own actual pattern — when focus reliably holds and when it reliably breaks — is more useful than importing someone else's number.

Actionable takeaways

• The 90-minute rule is a popularized simplification, not a precise finding — the real research describes an approximate, individually variable rhythm.
• Task-switching, not block length, is what the strongest evidence points to as the real cost — attention residue research shows interruption impairs performance independent of how long the original block was.
• Protect a shorter block completely rather than loosely holding a longer one — full protection beats raw duration.
• Give interrupted work a specific stopping point before switching away — this measurably reduces the residue left behind.
• Notice your own actual rhythm instead of importing a number that was never meant to be universal.

FAQ

Is the 90-minute focus rule based on real research?

Loosely — it traces back to Nathaniel Kleitman's research on ultradian rhythms in sleep, later extended to waking attention. The original findings describe an approximate, variable biological rhythm, not the fixed productivity rule the number became popularized as.

What is attention residue?

A concept from Sophie Leroy's 2009 research describing how switching tasks before finishing the first one leaves a portion of attention stuck on the unfinished task, measurably impairing performance on the next one — strongest when the first task was left open-ended with no clear stopping point.

Is a 25-minute focus session actually useful, or just a consolation prize?

Genuinely useful, according to the attention residue mechanism — a fully protected 25 minutes with zero task-switching produces less cognitive residue than a longer block interrupted several times, regardless of which one has more total minutes logged.

How do I reduce attention residue when I do have to switch tasks?

Leroy's research found that having a specific plan for returning to the unfinished task — even a one-line note on the exact next step — measurably reduced the residue left behind, compared to switching away with no defined stopping point.

Should I track my focus sessions to find my own actual rhythm?

It's a reasonable approach — self-monitoring research discussed elsewhere on this blog shows that simply observing your own behavior tends to shift it, and logging when focus reliably holds versus breaks gives you real, personal data rather than a borrowed number that was never guaranteed to fit your day.