Is it true that human attention span is now shorter than a goldfish's?

No. That claim — eight seconds, down from twelve — is usually attributed to Microsoft, but Microsoft's report cited a consumer statistics site whose own source cannot be traced to any study. There is also no meaningful sense in which a goldfish's attention span was ever measured for comparison. It is one of the most repeated fake statistics on the internet.

Has attention actually got worse?

Capacity, probably not. Conditions, certainly. There is no good evidence that ordinary phone use permanently reduces the ability to concentrate, but there is plenty that interruption is constant and that unbroken stretches have become rare. People who remove the interruptions generally report reading at length within weeks.

How long should I be able to concentrate for?

There is no correct figure, and any specific number you see quoted is invented. Sustained attention varies enormously by person, task, interest, sleep and time of day — the same person who cannot manage ten minutes on a form will manage three hours on something absorbing.

Do brain-training apps improve attention?

They reliably make you better at the app. Transfer to unrelated tasks is weak in the evidence, and one of the largest brain-training companies was fined for overstating exactly that. Practising the thing you actually want to be able to do is a better use of the time.

Does meditation help?

It is one of the few practices with reasonable evidence behind it for attention, and it has the same shape as everything else here: a short daily session of deliberately sustaining attention. If sitting still is unappealing, a daily reading session trains something similar and produces books.

How to improve your attention span

Start with the statistic, because it is almost certainly the reason you are reading this.

The eight-second attention span is not real

The claim goes: human attention has fallen from twelve seconds to eight, which is shorter than a goldfish. It appears in news articles, conference talks and roughly every productivity book of the last decade.

Follow it back and it collapses. The figure is usually credited to a Microsoft report, which did not measure it — the report cited a consumer statistics website, whose own citation cannot be traced to any study that measured anything of the kind. There is also no established measurement of a goldfish’s attention span to compare against, and the notion of a single number for “human attention span” is not how attention researchers describe it in the first place.

It is worth knowing this for two reasons. First, because it is false. Second, because believing it is actively harmful: if you think your capacity has been permanently damaged to below that of a fish, the sensible response is to stop trying, and that belief does more to keep people out of books than any amount of TikTok.

What is actually true

Attention is not one thing. Sustained attention, selective attention and working memory are separate systems that vary independently and by task. There is no single dial.

It varies enormously by conditions. Sleep, interest, stress, time of day and hunger move it more than most people credit. The person who cannot read two pages at eleven at night is often perfectly capable at ten in the morning.

Interruption is the real change. The evidence that phones permanently reduce capacity is thin. The evidence that modern life leaves almost no uninterrupted stretches is overwhelming — and a long stretch of attention cannot form in a fifteen-minute gap. That is a conditions problem, and conditions can be changed today.

It responds to practice. Sustained attention improves by being practised, and it declines by not being. Both directions are ordinary. Nothing is broken.

The practice

There is no shortcut here, and the method is boring enough that most articles avoid saying it plainly: do one sustained, undistracted thing every day.

1. Pick one task with a natural pull

Reading, an instrument, drawing, writing, chess, cooking something slow. It must be something you can want to do, because the wanting supplies attention that discipline otherwise has to.

Reading is the most common choice for a good reason: it demands continuous attention, it is available at any hour, and it produces something other than the training itself.

2. Twenty minutes, daily

Not an hour twice a week. Daily and short beats occasional and long, and twenty minutes is long enough to get past the settling period and short enough to survive a bad day.

3. Remove the interruption rather than resisting it

Phone in another room — not face down. Notifications off. One thing on the desk.

Resisting an available distraction spends attention from precisely the pool you are trying to build, which is why “leave it there but don’t look” is a worse plan than it sounds.

4. Expect the first ten minutes to feel bad

Every session starts slow, and a session started straight after a feed starts slower. That is the entry cost, not the verdict. It shortens as the weeks go on, and the single most common way people conclude their attention is broken is by quitting inside it.

5. Add five minutes a week

Slowly, and only when twenty is comfortable. The goal is a stretch you can reach every day, not a record you set once and then quietly avoid repeating.

What the evidence does not support

Brain-training apps. They make you better at the app. Transfer to real tasks is weak, and one of the largest such companies was fined over claims of exactly that kind.
Supplements and nootropics for ordinary attention in healthy adults.
Any specific number of days to fix it. The round figures are invented.
The idea that this is permanent. It is the one claim in the whole area with the least behind it and the most cost.

Where reading fits

If you want the training and something to show for it, a daily reading session is the most efficient version: it is the practice and the reward at once. How to focus while reading covers the settling period in detail, and how to read more covers making the session happen at all.

If the interruptions are the whole problem, how to stop scrolling is about breaking the specific reflex that prevents long stretches from forming.

BookGate is one narrow implementation of everything above: twenty minutes at a fixed hour, the apps you nominate shielded for the duration, and nothing counted except whether the night happened.