Human Attention Span - Statistics, Trends & Research

What is attention span - and is the "humans worse than goldfish" claim actually supported by science?

The much-shared 'human attention span fell to 8 seconds — shorter than a goldfish' line comes from a 2015 Microsoft Canada marketing report and is widely considered weak evidence. The stronger, peer-reviewed body of work is from cognitive scientist Gloria Mark (UC Irvine), who has measured average adult attention span in real work conditions since 2004. Her observational data show the average sustained attention span falling from ~150 seconds (2.5 minutes) in 2004, to ~75 seconds in 2012, to around 47 seconds by the early 2020s. Microsoft's own 2025 Work Trend Index reports knowledge workers are now interrupted every 2 minutes on average. The trend — a shorter human attention span in digital environments — is real; the goldfish comparison is not.

What a degraded attention environment is actually doing to your ability to think deeply and feel satisfied

The much-shared "8.25 seconds, shorter than a goldfish" attention-span figure from a 2015 Microsoft Canada marketing report has been widely debunked — the goldfish comparison was never in a peer-reviewed paper and the underlying study has never been reproduced. The stronger body of work is Gloria Mark's twenty-year research programme at UC Irvine, which measures the average adult attention span in real digital work environments. That figure fell from about 150 seconds in 2004 to around 47 seconds today, and Microsoft's own 2025 Work Trend Index reports workers are now interrupted every ~2 minutes on average. The attention-span decline is real; the goldfish is a myth.

The core issue is not that attention "shrinks" in the way a muscle atrophies, but that it competes against constantly available, algorithmically optimized distraction. Neuroscientist Stanislas Dehaene's work on attention shows that focused attention is an active, metabolically costly state. Smartphones provide constant low-friction alternatives to that effortful state. When checking a phone requires less mental energy than maintaining focus on a difficult task, the brain will consistently choose the check. Over years of repetition, this pattern becomes a default orientation.

The research by Cal Newport and Gloria Mark consistently finds that this isn't just about productivity: the inability to enter and sustain deep focus states correlates with lower reported satisfaction, higher anxiety, and a diminished sense of meaning in work. The timer at the top of this page is a small demonstration: if you're still reading at this point, you've already significantly exceeded the commonly cited average. That's not because the statistic is wrong - it's because context, relevance, and genuine interest are still the most powerful determinants of attention. The algorithms don't compete with genuine curiosity.

Attention span over time

Cognitive scientist Gloria Mark's 20-year programme of observational attention research at UC Irvine is today the most credible empirical measurement of human attention span in digital environments. Across her studies the average sustained attention span fell from ~150 seconds in 2004 to ~75 seconds in 2012, and to about 47 seconds by the early 2020s. The viral 'attention span is now 8 seconds, shorter than a goldfish' figure from a 2015 Microsoft Canada marketing report is not comparable and has been debunked.

Older research on attention

Psychologists were studying attention long before the internet existed. Norman Mackworth’s classic "clock" experiments in the late 1940s showed a reliable vigilance effect. During long, monotonous monitoring tasks, people’s ability to detect rare signals tends to drop within the first stretch of the watch and can continue to decline as time on task increases.

A second line of work, associated with researchers such as Donald Broadbent and Colin Cherry in the 1950s, showed strong limits on divided attention. When two streams of information compete, the mind filters and prioritizes rather than fully tracking everything at once.

Later research, including Michael Posner’s work in the 1980s, measured how quickly attention can be oriented within the visual field. These shifts happen on very short timescales, often in the range of milliseconds, and they are shaped by cues, expectations, and competing stimuli.

What these studies share is that attention is not one thing. It weakens under monotony, it selects under competition, and it shifts rapidly in ways we do not consciously feel. The results depend on which kind of attention is being tested, under what conditions, and over what timeframe. That is why this research does not yield one universal number. The idea of a single, all-purpose human attention span measured in seconds is largely a media simplification, not a standard outcome of the pre-digital scientific literature.

Key attention span statistics

Attention span vs. daily social media time

We sustain focus for 47 seconds on average, yet spend nearly 3 hours a day on social media feeds designed to never end.

The real attention-span numbers: what Gloria Mark's 20-year research actually shows

What the attention and focus research actually says - methodology and findings

From cognitive psychology labs to viral headlines: how the attention span myth spread

1. 2004Gloria Mark (UC Irvine) baseline: average adult attention span in digital work ~150 seconds (2.5 min)
2. 2007iPhone launch: smartphones embed always-on notifications into everyday attention patterns
3. 2012Mark et al. follow-up: average attention span drops to ~75 seconds as push notifications and social feeds mature
4. 2015Microsoft Canada publishes '8.25 seconds, shorter than a goldfish' attention-span figure — viral, but never reproduced in peer-reviewed research
5. 2023Gloria Mark 'Attention Span' (book): average human attention span in digital environments now ~47 seconds — a ~3× drop since 2004
6. 2025Microsoft Work Trend Index 2025: workers interrupted every ~2 minutes; CHIWORK 2025 peer-reviewed study confirms fragmented attention in modern workplaces

In perspective

About this statistic

Unlike most statistics on this site, attention span is a research measurement rather than a cumulative count. The 47-second figure is the average sustained attention span of adults in digital work environments, taken from Gloria Mark's 2023 synthesis ('Attention Span', UC Irvine) of two decades of observational studies in which participants' activity was tracked in real work conditions (not surveys or self-reports). In the same research programme, the average was ~150 seconds in 2004 and ~75 seconds in 2012 — a roughly 3× reduction in twenty years. The viral '8.25 seconds' figure from a 2015 Microsoft Canada marketing report is shown for historical context only; it was based on a consumer survey and short EEG tests, was never replicated in peer-reviewed research, and the goldfish comparison has been debunked.

Sources: Gloria Mark - Attention Span: A Groundbreaking Way to Restore Balance, Happiness and Productivity (UC Irvine / Hanover Square Press). Methodology →

Frequently asked questions

What's the real human attention span in 2025-2026?

The most rigorous figure is around 47 seconds, based on Gloria Mark's (UC Irvine) 20-year observational research summarised in her 2023 book 'Attention Span'. That's down from about 150 seconds (2.5 min) in 2004 and 75 seconds in 2012 — roughly a 3× reduction in 20 years. Microsoft's 2025 Work Trend Index independently finds knowledge workers are interrupted every ~2 minutes on average.

Is the '8 seconds — shorter than a goldfish' claim accurate?

No. That figure came from a 2015 Microsoft Canada consumer marketing report and was never a peer-reviewed cognitive-science measurement. The goldfish comparison has been debunked: goldfish attention is not a well-defined 9-second constant, and the Microsoft report didn't measure attention span the way academic research does. We keep the number on record as a historical reference only.

Does a shorter digital attention span mean people can't focus deeply anymore?

Not exactly. What has measurably shrunk is average sustained attention in digital environments — driven partly by notifications and multitasking, partly by habit. Deep focus on non-digital tasks (reading a physical book, a musician practising) is less clearly affected. Gloria Mark and the 2025 CHIWORK study ('Are Six Minutes of Focus Enough?') both treat attention as a resource that fragments under interruption rather than a single universal shrinking number.