The Claim That Everyone Repeats
Attention spans have collapsed. Smartphones broke our brains. The average person now focuses for eight seconds, less than a goldfish. You have read this, probably more than once, possibly in an article that took four minutes to finish.
The goldfish statistic is worth examining because of what it reveals about the genre. It spread from a 2015 marketing report, attributed to research that nobody has been able to locate, describing a species whose attention has never been measured that way. It is a fabrication that persists because it confirms something people already feel.
The feeling is not fabricated. Sustained attention genuinely is harder in an environment engineered to interrupt it, and the research on that environment is more specific and more useful than the panic built on top of it. This article covers what is actually established: how long focus lasts, what interruption costs, what flow requires, and which interventions have support.
Attention Is Not a Fixed Quantity
The premise behind "my attention span is eight seconds" is that attention is a personal constant, like height. It is not. The same person who cannot get through two paragraphs of a work document will watch a three-hour film, play a video game until 2 am, or read a novel through a flight delay.
What varies is not capacity but the interaction between the task, its reward structure, the person's interest, and the environment. This is why "improve your attention span" is a poorly formed goal, and why the interventions that work address specific obstacles rather than a general faculty.
The research distinguishes several things that get bundled under the word focus: sustained attention (maintaining vigilance over time), selective attention (filtering distractors), and executive control (managing goals). They have different limits and different failure modes.
The Interruption Cost, Precisely
This is the best-supported finding in the area and the one most worth acting on.
Gloria Mark's observational research on office workers is the standard citation, reporting that people switched tasks every few minutes and took an average of around 23 minutes to return to an interrupted task. The figure deserves a caveat that is usually dropped: it reflects a specific workplace and includes intervening tasks, so it is not a clean measure of cognitive recovery time. The finding beneath it, that interruptions cost far more than the seconds they occupy, is robust across methods.
Laboratory work adds a mechanism. Task switching produces measurable costs even when the switch is voluntary and brief, because the previous task's goal set does not clear immediately. Some of it lingers and competes. There is also evidence that the mere presence of a smartphone, face down and silent, is associated with reduced available cognitive capacity, presumably because not attending to it requires something.
The practical implication is unglamorous: the largest available gain in focused work is usually not learning to concentrate harder. It is reducing the number of times concentration has to restart.
How Long Focus Actually Lasts
| Common claim | What research supports |
|---|---|
| Eight-second attention span | No traceable source; not a meaningful measure |
| Attention collapses after 20 minutes | Vigilance declines on monotonous tasks; engaging work sustains far longer |
| The Pomodoro 25-minute interval is optimal | No evidence for that specific number; structured breaks generally help |
| Four hours is the daily limit for deep work | Popular claim from expertise research on deliberate practice, not a measured universal |
| Interruptions cost more than their duration | Well supported across multiple methods |
The pattern in that table is worth noting: the specific numbers people repeat are mostly not measurements, while the general principle that fragmenting attention is expensive holds up well. Anyone selling a system built on a precise number is selling the packaging.
Flow, Described Accurately
Flow is the state of absorbed engagement where time distorts and the work seems to carry itself. The research describes conditions associated with it fairly consistently: a match between challenge and skill, clear goals, immediate feedback, and freedom from interruption.
What the research does not support is a method for producing it on demand. The conditions are necessary rather than sufficient, and much of the work in this area relies on self-report about a state that is defined partly by not noticing yourself being in it, which is a genuine methodological difficulty.
The realistic version is subtractive. You cannot summon flow, and you can reliably prevent it: by working on something with no clear goal, at the wrong difficulty, with notifications on. Removing those obstacles is what an individual can actually control.
What the Evidence Supports
Reduce interruption sources, structurally. Notifications off rather than dismissed, phone in another room rather than face down, and a block of time protected rather than defended. This follows directly from the best-supported finding in the field.
Single-task deliberately. The evidence that multitasking degrades performance is strong. The people who believe themselves best at it perform worst in several studies, which is an unusually clean result.
Match difficulty to skill. Work that is too easy invites drift; work that is far too hard produces avoidance. This is the one flow condition that is actually adjustable in most jobs.
Sleep, before technique. Sleep restriction degrades sustained attention measurably, and lapses in vigilance are one of the most reliable findings in sleep deprivation research. No focus method compensates for chronic short sleep. The Sleep Foundation's summary of sleep deprivation effects covers the cognitive consequences.
Physical activity. Aerobic exercise has reasonable evidence for short-term improvements in attention and executive function, which is a modest effect from a free intervention.
Breaks that differ from the work. A break spent on a phone is a switch to another demanding visual task. Attention restoration research suggests time in natural settings has more benefit than equivalent indoor time, with modest effect sizes.
Meditation, Fairly Assessed
Mindfulness training is the intervention with the most study behind it for attention, and its evidence is better than brain training and weaker than its reputation.
Meta-analyses report small to moderate improvements in attentional measures. The literature has known problems: many trials are small, active control groups are the exception rather than the rule, and expectancy effects are difficult to eliminate when participants know they are meditating. Reviews have repeatedly flagged these limitations while still finding a signal.
The honest summary is that the evidence for meditation and attention is real, modest, and imperfect. That is a better position than most things in this category occupy.
Where an Audio Program Fits
Brainwave entrainment programs sit at the intersection of a real laboratory phenomenon and an oversold consumer claim.
The phenomenon: the brain does show measurable responses to rhythmic stimulation, and EEG changes in response to auditory beats have been documented. Gamma oscillations are genuinely associated with attention and perceptual binding in neuroscience research.
The gap: the step from a measurable EEG response to improved concentration during actual work is where the evidence becomes thin. Studies of binaural beats and cognitive performance are small, short, methodologically varied, and inconsistent in their results. They are also difficult to blind, since participants can hear whether they are receiving the intervention, and expectation is a powerful confound in exactly this kind of outcome.
The Brain Song is one commercial example, a gamma-focused audio program sold with a 90-day refund window. Research on gamma oscillations in the brain is not research on this product, and we found no independent trials of it, so the available evidence is insufficient to say what it does. A more defensible description of what an audio track offers is behavioral: it masks environmental noise and can function as a start signal for a focused block, both of which are real effects that require no claims about oscillations. It belongs behind sleep, interruption control, and single-tasking, which are where the evidence points. Our full Brain Song review covers the claims and the refund process.
When It Is Worth Asking a Professional
Lifelong difficulty with attention that predates smartphones, causes problems across multiple settings, and has been present since childhood is a different matter from a fragmented workday. Adult ADHD is real, underdiagnosed in some groups, and evaluated by a clinician rather than an online quiz.
New concentration difficulty that arrives with low mood, sleep disturbance, or loss of interest may reflect depression, which affects concentration directly. Persistent cognitive change is also worth evaluating for the common medical causes, including thyroid dysfunction, anemia, and sleep apnea.
Frequently Asked Questions
How long does it take to refocus after an interruption?
A widely cited study of office workers reported an average of around 23 minutes to return to the original task, though that figure reflects a specific workplace context and includes intervening tasks rather than measuring pure cognitive recovery. The broader finding, that interruptions carry a resumption cost well beyond their own duration, is well supported.
Is the eight-second attention span claim true?
No. The claim circulated widely after a 2015 marketing report and traces back to no identifiable research. Attention span is not a single fixed quantity: it varies enormously with task, motivation, interest, and context. The same person who cannot read a page can watch a three-hour film.
What is flow state and can you produce it on demand?
Flow describes a state of absorbed engagement characterized by a match between challenge and skill, clear goals, and immediate feedback. The conditions associated with it are reasonably well described, and reliably producing it on demand is not something research supports. Removing obstacles to it is more realistic than summoning it.
Does background music help concentration?
The evidence is mixed and depends heavily on the task and the person. Music with lyrics generally interferes with reading and language tasks. For repetitive work it may help by maintaining arousal. Individual differences are large enough that population-level averages are of limited use for any specific person.
Does having your phone nearby affect focus even if you do not use it?
Some research has reported reduced available cognitive capacity when a smartphone is merely present, even face down and silenced, with the proposed explanation that inhibiting attention to it consumes resources. The finding has not replicated uniformly, so it is suggestive rather than settled, and the intervention it implies costs nothing to try.
The Bottom Line
Attention is not a fixed personal quantity that technology has drained, and the statistics claiming otherwise mostly trace to marketing rather than measurement. What research does establish is that interruption is expensive well beyond its duration, that multitasking degrades performance while feeling productive, and that sleep loss reliably damages sustained attention. The useful moves follow from that: fewer interruptions, one task at a time, difficulty matched to skill, and enough sleep to make any of it possible. Meditation has modest, imperfect evidence. Audio programs marketed on brainwave claims do not, and what they plausibly offer is noise masking and a ritual, which is a smaller claim that happens to be defensible.