Two and a half minutes. That is how long, on average, a person used to stay on a single screen before flicking to something else, back when researchers first started measuring it. Two decades later, that figure has fallen to about 47 seconds. Same task, same kind of worker, but a fraction of the staying power.

The temptation, the moment you read that, is to reach for the obvious villain. I did. But the woman who actually collected the numbers thinks the obvious story is only half right, and the half it misses is the more interesting one.

A quick note before going further: I am a writer, not a psychologist or a neuroscientist, and this is one researcher reading her own data rather than a settled scientific consensus. The figures below come largely from a single line of research. Take them as a clue about how a lot of us work now, not a diagnosis of you.

The number that keeps shrinking

The person doing the measuring is Gloria Mark, a professor of informatics at the University of California, Irvine, who trained as a psychologist and wrote the 2023 book Attention Span. She has spent roughly twenty years watching how people actually behave at their computers, and the trend line she found bends only one way.

As Mark describes it, “When we first started measuring this, back in 2004, we found that people averaged two and a half minutes on a screen before switching.” By 2012, her logs put the average at around 75 seconds. Then it kept falling. In her words, “In the last, I would say, five or six years, since around 2016, we found that attention spans average 47 seconds on a screen before switching.”

One thing worth being honest about: the 2004 figure came from researchers shadowing people, while the 47-second number comes from later automated computer logging. Different methods, so it is not a perfectly clean comparison. But the shape of the drop holds up. The median in the recent data sits at about 40 seconds, and other researchers have landed in the same neighbourhood, one at 50 and another at 44. The number is real, and it is small.

The easy story we tell about it

So who took our minutes? The reflex answer writes itself. Phones, social media, the algorithm engineered in some open-plan office to keep your thumb moving. It is a satisfying story because it hands us a clean villain and lets us off the hook entirely — our attention was stolen, we are victims of design.

Not entirely wrong. Mark herself puts tech companies near the top of the list. But she refuses to stop there. She says, “I think there are a lot of things going on. First, tech companies bear some responsibility. I don’t think that is the sole reason why our attention spans have shortened.”

If the apps were the whole story, you would expect to be pulled away by them. Buzz, notification, ping. But a lot of the switching, in Mark’s account, comes from inside the house.

What the data actually points to

We assume distraction arrives from outside, a colleague at the desk or a notification on the lock screen. Mark’s work, however, suggests that we are as likely to interrupt ourselves as to be pulled away by something external. Nobody pinged us; we just reached for the other tab.

I know this one from the inside. Many mornings I sit down to write, spend hours researching, never quite find the angle, and somewhere in the fog I drift. A message, then YouTube, then a different tab, then back, jumping from one thing to the next until I look up and the day is gone. No app reached into my skull and pulled. I left.

Mark’s explanation is less about theft and more about training. As she puts it, “We have developed very deep habits of switching attention. A lot of these habits, as you know, are automatic.” Years of being rewarded for the flick, of checking email something like 77 times a day, and the move stops being a choice and starts being a reflex you carry into every quiet moment.

That reframe matters because it shifts the problem from “they did this to us” to “we built this together, and we keep maintaining it.” It is less flattering and more useful.

What 47 seconds might mean for how we work

The figure I bumped into years ago, while researching a different post, was the cost of recovery. Mark reports that it takes around 25 minutes to pick up an interrupted task again. That stopped me, because at the time I was breaking my own focus constantly and barely noticing. A Slack message mid-sentence, a quick reply, and then that draggy stretch where I am sort of back but not really back, the old thread still half-loaded somewhere behind my eyes.

What I have landed on is not a system, just a few small defences. The phone goes in another room. Every social tab is closed before I start, not minimised, closed. I put rain sounds on instead of music, because anything with lyrics pulls me sideways. And I try to give the task a hard stop. None of this makes me a disciplined person, and it does not always work. Plenty of mornings still dissolve. But removing the easy exits at least makes the next flick a decision again rather than a reflex.

The thing I keep coming back to is Mark’s own framing, which is gentler than the panic the number usually inspires. She argues that “Our ability to focus isn’t lost, the way we focus is just changing.” Not a brain that has been broken, just a brain that has been trained, which means it can in principle be trained back. Her one piece of plain advice, stated simply: “We have limited and very precious attentional resources — use them wisely.”

What the 47 seconds is really measuring is the shape of a habit, recorded one screen-switch at a time, mostly by people who did not notice they were switching. That is a habit we can work with, if we choose to.

If your own focus is fraying in a way that worries you beyond the everyday, a doctor or psychologist is a better guide than any blog post.