Most statistics pages on this subject are quietly recycling numbers from 2019, or from a phone insurance company's marketing survey, or from a study that no one has been able to locate. This page tries to do better: measured data is separated from self-reported data, every number is dated to when it was collected rather than when it was published, and the well-known figures that turn out to be wrong are corrected rather than repeated.
If you want to cite something here, cite the original source rather than this page — every one is linked. If you find an error, I would genuinely like to know.
Download the figures as CSV or download the machine-readable JSON. Both files include the label, number, explanation, source, source URL and data period shown on this page.
How much we are actually using
About 4 hours 45 minutes a day. This is self-reported, from a survey of more than 240,000 people across 54 economies.
DataReportal / We Are Social / Meltwater — Digital 2026 Mid-Year Global Updatepublished April 2026 · measured April 2026
This one is not a survey. It is measured directly on devices across the App Store and Google Play, which makes it the most reliable screen-time figure available anywhere.
Sensor Tower — State of Mobile 2026published January 2026 · measures all of 2025
Up 3.8% on the year before. That is roughly 14.5 billion hours a day, every day.
Sensor Tower — State of Mobile 2026published January 2026 · full-year 2025
Measured passively rather than self-reported. 18–24s average 6 hours 20 minutes; over-65s 3 hours 20 minutes. 77% of it happens on a phone.
Ofcom — Online Nation 2025 (Ipsos iris panel)published December 2025 · measured May 2025
This counts simultaneous use twice — scrolling while the television is on is two hours, not one. Which is exactly the point: for most of the waking day, a screen is somewhere in the frame.
EMARKETER — H2 2026 forecast updatepublished May 2026 · 2026 forecast
98% own a mobile phone of some kind. 16% have no home broadband and reach the internet only through the phone.
Pew Research Center — Mobile Fact Sheetpublished November 2025 · fielded Feb–Jun 2025
The checking habit
About 11.6 times an hour, every waking hour. Self-reported by around 1,000 US adults.
Reviews.org — Cell Phone Usage Stats 2026published January 2026 · fielded Q4 2025
71% check notifications within five minutes of waking.
Reviews.org — Cell Phone Usage Stats 2026published January 2026 · fielded Q4 2025
Half of adults keep the thing that wakes them up next to the place they sleep.
Reviews.org — Cell Phone Usage Stats 2026published January 2026 · fielded Q4 2025
53.1% say they have never gone 24 hours without it. 76.3% feel uneasy leaving it at home.
Reviews.org — Cell Phone Usage Stats 2026published January 2026 · fielded Q4 2025
Sleep
A further third do it most days or several days a week.
American Academy of Sleep Medicinepublished February 2026 · fielded June 2025, 2,007 US adults
One in four adults knowingly trade sleep for scrolling.
American Academy of Sleep Medicinepublished February 2026 · fielded June 2025
Nearly half of young adults can already name the mechanism damaging their rest.
American Academy of Sleep Medicinepublished February 2026 · fielded June 2025
Attention
In 2004 the same research group measured two and a half minutes. The nearest peer-reviewed anchor is a median of 40 seconds, from software logging every window switch for 40 office workers over 12 working days.
Gloria Mark, UC Irvine — Attention Span (2023); Mark et al., CHI 2016book published January 2023 · underlying study 2016
And in that gap people handled an average of 2.26 other pieces of work. It is not that refocusing takes 25 minutes — it is that the day fragments into pieces that never join back up.
Mark, González & Harris — CHI 2005published 2005 · 700+ hours of workplace observation
How common the problem is
Pooled from 109 studies covering 97,748 people. The rate has grown steadily since 2012 — and is growing fastest among older adults, not teenagers.
Lu, An & Chen — Cyberpsychology, Behavior & Social Networking 27(9)published September 2024 · studies from 2012–2022
Median across 41 studies of 41,871 young people. 17–19-year-olds were the most affected group.
Sohn et al. — BMC Psychiatry 19:356published 2019 · 41 studies
Up from 7% in 2018. Girls 13%, boys 9%. Across 44 countries.
WHO Europe / HBSC — Teens, screens and mental healthpublished September 2024 · ~280,000 adolescents surveyed 2022
What your attention is worth
Three months. From advertising alone.
Meta — Q2 2026 resultsreleased July 2026 · quarter ended June 2026
$16.86 a quarter from each of 3.60 billion daily users — up 42% in two years. Your attention is not the price of the free product. It is the product.
Meta — Q2 2026 results (average revenue per person)released July 2026 · quarter ended June 2026
Of which YouTube alone was $11.1 billion.
Alphabet — Q2 2026 results, filed with the SECreleased July 2026 · quarter ended June 2026
An annual run-rate around $564 billion, spent working out how to hold your eyes a little longer. That is what you are up against when you decide to put the phone down. It was never a fair fight.
Calculated from both companies' Q2 2026 filingsquarter ended June 2026
Children and teenagers
97% use the internet daily. YouTube reaches about 90% of them, TikTok and Instagram about 60% each.
Pew Research Center — Teens, Social Media and AI Chatbots 2025published December 2025 · fielded Sept–Oct 2025, 1,458 US teens
31% find those conversations as satisfying as talking to a real friend, or more so. 33% have chosen the AI over a person for a serious conversation.
Common Sense Media — Talk, Trust and Trade-Offspublished July 2025 · fielded April–May 2025, 1,060 US teens
At age 10 it is 56%. The jump happens in a single year, at the move to secondary school.
Ofcom — Children and Parents: Media Use and Attitudes 2026published May 2026 · 5,008 parents and 3,426 children
Rising to 47% among children with a mental health condition, and 57% among looked-after children. They know.
Ofcom — Children and Parents: Media Use and Attitudes 2026published May 2026
40% have their own tablet by the age of two. Gaming time has risen 65% in four years.
Common Sense Media — 2025 Census: Media Use by Kids Zero to Eightpublished February 2025 · fielded August 2024
The 2026 story: AI arrives
Up 1.4 billion in twelve months — a 141% rise. Roughly 29% of everyone alive.
DataReportal — Digital 2026 Mid-Year Global Updatepublished April 2026
3.6 times more than the year before. A whole new category of screen time appeared in about eighteen months.
Sensor Tower — State of Mobile 2026published January 2026 · full-year 2025
Rising to 66% of 16–17s. 11% of those who use it describe using it 'as a friend'.
Ofcom — Children and Parents: Media Use and Attitudes 2026published May 2026
Down from 40% a year earlier. Only 29% think being online is good for their mental health, down from 33%. The mood is turning.
Ofcom — Online Nation 2025published December 2025 · fielded June 2025
Seven famous statistics that are wrong
These appear on almost every article about screen addiction, including some very reputable ones. None of them survives a check. If you are writing about this subject, please do not pass them on.
“It takes 23 minutes and 15 seconds to refocus after an interruption.”
There is no paper behind it. The number is usually credited to Gloria Mark, but it appears in none of her published work. Her actual 2005 finding is 25 minutes 26 seconds of elapsed time before returning to an interrupted task — and during that gap people handled about 2.26 other pieces of work. Her 2008 study found people worked faster when interrupted, at the cost of higher stress. The oddly precise 15 seconds seems to have entered circulation through an interview.
Traced to: Mark, González & Harris, CHI 2005“Americans check their phones 352 times a day.”
Real survey, wrong decade. It is Asurion — a phone insurance company — with fieldwork from March 2022, never repeated. The equivalent current figure is 186. The older “96 times a day” is from the same company in August 2019.
Traced to: Asurion, fieldwork March 2022“Teenagers spend 8 hours 39 minutes a day on screens.”
Correctly attributed to Common Sense Media — but the fieldwork was October 2021, at the tail of the pandemic, and that census has not been repeated for 8–18s. Every page quoting it as a 2026 figure is quoting five-year-old data. The current Common Sense census covers ages 0–8 only.
Traced to: Common Sense Media, fieldwork October 2021“Human attention span is now 8 seconds — shorter than a goldfish.”
This one is simply not true. It traces to a 2015 Microsoft Canada marketing report citing a source that no researcher has ever been able to locate, and the goldfish comparison has no basis at all. Attention is not a single quantity that can be measured in seconds. What has been measured is how long we stay on one screen before switching, which is a different and more useful thing.
“The average person touches their phone 2,617 times a day.”
From a 2016 study of about 94 people, run by a research firm. Ten years old and a very small sample. Interesting, not authoritative.
Country-by-country screen time league tables
DataReportal stopped publishing country-level time-spent figures. Any 2026 ranking you see is built on 2023 or 2024 data without saying so.
Traced to: DataReportal — notes on dataThe tidy five-year screen time trend line
You cannot draw one. The most-used global series changed its survey method in late 2024 and the publisher states plainly that figures before and after are not comparable. Anyone showing you a smooth 2021–2026 curve has plotted across a break in the data.
Traced to: DataReportal — notes on dataMeasured numbers versus survey numbers
There are two kinds of number on this page and it matters which is which. Measured figures come from software that records what actually happened on the device — Sensor Tower's app data and Ofcom's Ipsos iris panel are both of this kind. Self-reported figures come from asking people, and people are consistently poor at estimating their own screen use, usually underestimating it.
The two disagree, and neither is simply wrong. A measured figure tells you what the device did; a self-reported one tells you what the person believes, which is its own useful fact. What you should be wary of is any page that mixes them without saying so, or that draws a trend line across a year in which the measurement method changed. The most-used global series did exactly that in late 2024, and its own publisher warns that figures before and after are not comparable.
Nothing on this site is sponsored, affiliated or paid for. There are no affiliate links, no advertising, no sponsored placements and no tracking. Nobody can pay to appear here or to be described more kindly. There is also no signup: no email address, no account, no cookie banner. If a company on this page later offers me money, this line is how you will know they were turned down. Why it is all free →
Michael has spent much of his working life online—and much of the rest of it trying to stop that technology taking over. He has thrown a phone into a lake, locked devices in boxes and experimented with long periods offline. He writes about what helped, what failed and how to keep the useful parts of digital life.
Last updated 9 August 2026. This page is checked and corrected as new research comes out. If you spot something out of date or wrong, tell me and I will fix it.