The Digital Whale

Attention and habit

How often people actually use their phones

Metered panels put UK adults at four and a half hours online a day, while the most-quoted pickup counts come from a 94-person vendor study and self-report marketing polls.

The best-measured answer is that a typical adult's phone is in use for between three and four and a half hours a day, in dozens of short bursts, and that nobody has a trustworthy population-level count of how many times a day people pick it up. The figures split into two families. Logged measurements come from small samples and are reasonably accurate. Survey figures come from large samples and are not, because people estimate their own use badly. The gap runs one way: people underestimate. That disqualifies most of the numbers in circulation, including several in official-looking places. For a workplace-oriented comparison point on how digital activity is translated into metrics, see employee PC activity tracking.

Self-report does not match the logs

The largest synthesis compared survey answers with device logs wherever both existed.

2021

Across 106 effect sizes, self-reported media use correlated only moderately with logged use, and measures of problematic use tracked the logs even more weakly.

Direction: Increase. Strength of evidence: Strong.

Parry, Davidson, Sewall, Fisher, Mieczkowski & Quintana, Nature Human Behaviour, 2021106 effect sizes, self-report against device logs

Caveat Included studies varied in logging method and duration, and the finding undermines most survey-based screen-time research, including figures elsewhere on this page.

A smaller study sized the error, comparing donated iOS Screen Time records with estimates.

2021

Measured use averaged 257 minutes and 69.6 pickups a day against self-reports of 146 minutes and 55 pickups, with 70% underestimating screen time by an average of 125 minutes and 84% underestimating pickups by 37.

Direction: Increase. Strength of evidence: Mixed.

Ohme, Araujo, de Vreese & Piotrowski, Mobile Media & Communication, 202147 Dutch adults who donated iOS Screen Time data

Caveat Only 47 people donated data, self-report and log correlations were just 0.3–0.4, and donation biases the sample towards more engaged users.

The metered numbers are lower than the survey numbers

The UK has one of the few large metered panels reporting publicly, run for Ofcom by passive measurement.

2025

UK adults averaged 4 hours 30 minutes online per day in May 2025, up 10 minutes year-on-year, with 77% of that time on smartphones; 18–24s averaged 6h20m and over-65s 3h20m.

Direction: Increase. Strength of evidence: Strong.

Ofcom, Online Nation 2025, 2025UKOM-accredited Ipsos iris passive panel, UK adults 18+

Caveat It measures personal-use internet time, not total screen time — offline app use, calls and camera are excluded — and covers adults only, so it is not comparable with other sources.

The figure most often quoted globally is larger. DataReportal's Digital 2025 report (January 2025) puts average daily internet time at 6 hours 38 minutes for the third quarter of 2024, from GWI's commercial survey panel — self-reported throughout, and broken as a time series by a methodology change in early 2024 that added over-65s and lowered the averages.

Use is fragmented, not continuous

Long unbroken sessions are rare; logged German data found the opposite of the "hours lost to the phone" image.

2025

Most smartphone sessions last under seven minutes and individual app use rarely exceeds two minutes.

Direction: Increase. Strength of evidence: Mixed.

Toth, Parry & Emmer, Journal of Quantitative Description: Digital Media, 20251,797 German adults, Android logs, iOS donations and experience sampling

Caveat German adults only, and the iOS portion relied on voluntary donation, which skews towards more privacy-tolerant and engaged users.

The frequency of switching matters more than the daily total, and it is what shows up in the work on interruptions and how long attention lasts.

Heavy use tracks weakly with mood

The largest passive study of use and well-being found very little.

2025

Associations between logged screen time, app categories and unlock frequency and mental well-being were weak or statistically insignificant, and age and gender predicted mood better than phone use did.

Direction: No detectable effect. Strength of evidence: Strong.

Sultan, Hernandez et al., PNAS, 2025more than 10,000 US adults, more than 250,000 days of passively recorded use

Caveat Google Research collaborated on recruitment and analysis and a Google employee is a co-author, and participants self-selected into a Google health app, so the sample is not representative.

Teenagers are measured worse than adults

Teen figures are almost entirely self-reported, and the headline one is a category, not a quantity.

2024

46% of US teens say they are online almost constantly, roughly double the 24% of a decade earlier, and 95% have access to a smartphone.

Direction: Increase. Strength of evidence: Strong.

Pew Research Center, Teens, Social Media and Technology 2024, 20241,391 US teens aged 13–17, recruited via parents, Ipsos KnowledgePanel

Caveat "Almost constantly" is a self-categorisation rather than a measured quantity, and the figure has been flat for several years despite being reported as rising.

The other commonly cited teen number — 8 hours 39 minutes of screen media a day for 13- to 18-year-olds, 5 hours 33 minutes for 8- to 12-year-olds — comes from Common Sense Media's 2021 census wave, compiled by the American Academy of Pediatrics. It is self-reported time-diary data from an advocacy non-profit, excludes school and homework use, and should not be combined with that organisation's later figures for younger children. Household rules built on it deserve reading against the evidence on screen-time limits.

The pickup counts are the weakest numbers in circulation

The most repeated statistic here has a small footing.

2016

The "2,617 touches a day" figure — and 5,427 for the heaviest 10% — comes from a five-day tracking exercise of 94 Android users.

Direction: Increase. Strength of evidence: Weak.

dscout, Mobile Touches study, 201694 Android users tracked for five days

Caveat Never peer-reviewed, Android only, published by a company selling mobile research tooling, and roughly a third of participants did not complete all five days.

The other family of pickup figures is worse. The claim that Americans check their phones 186 times a day comes from a Pollfish online survey of about 1,000 US adults, run in the fourth quarter of 2025 and published by Reviews.org in January 2026 — a site funded by affiliate revenue from mobile-plan comparisons, using a non-probability panel with a margin of error of about four points.

The short version

  • Logged use in a Dutch donation study averaged 257 minutes a day against self-reports of 146 minutes, with 70% underestimating their own screen time (Ohme et al., Mobile Media & Communication, 2021).
  • Ofcom's metered panel put UK adults at 4 hours 30 minutes of personal online time a day in May 2025, below self-reported global averages.
  • Phone use is fragmented: most sessions run under seven minutes and most app visits under two minutes (Toth, Parry & Emmer, 2025).
  • The best-powered passive study of use and well-being, covering over 250,000 days, found weak or statistically insignificant associations (Sultan, Hernandez et al., PNAS, 2025), though Google collaborated on it.
  • No population-representative logged count of daily pickups exists, so any "times a day" figure is a survey artefact rather than a measurement.