Notifications, phones and the attention economy

People blame themselves for checking their phone. The honest framing is that they are losing a contest with teams of engineers whose objective is precisely that they check it.

What the design actually does

Several mechanisms, all well documented and none accidental.

Variable reward. Content that is sometimes interesting produces more checking than content that is reliably interesting.

Intermittent reinforcement schedules produce the most persistent behaviour, which has been known in learning research for decades and is applied deliberately.

The pull-to-refresh gesture is the clearest example: a physical action, followed by an uncertain outcome, repeatable indefinitely.

Infinite scroll. Removing a natural stopping point. A finite page ends and prompts a decision; an infinite one does not.

Social obligation. Read receipts, typing indicators and streak counters convert use into a duty.

Notification design. Badges, sounds and vibration exploit orienting responses that are largely involuntary.

Autoplay, which removes the moment at which a person would otherwise choose to stop.

Self-interruption is most of it

The finding that changes what the response should be.

Observational studies of workplace interruption consistently find that a substantial share are self-generated. Nobody prompted the check.

Which means silencing notifications addresses only part of the problem. The habit is cue-driven, and the cue is frequently internal — a moment of difficulty, boredom or uncertainty in the task.

That is the same mood-regulation mechanism that drives procrastination, and it explains why the checking increases precisely when the work becomes hard.

What works

Distance. The most effective intervention available.

A phone in another room is checked far less than a phone in a pocket, and a phone face-down on the desk is checked nearly as much as one face-up.

Studies have found that the mere presence of a phone is associated with reduced performance on attention-demanding tasks, even when it is not used — the effect sometimes described as brain drain.

The effect size is debated. The practical advice is unaffected: put it somewhere else.

Turning off notifications properly. Not silencing them — disabling them, per application, leaving only those that genuinely require immediate response.

Most people find that number is very small, and that the anxiety about missing something does not survive a week.

Removing applications rather than resolving to use them less. Using a service through a browser rather than an app removes the notification channel and adds friction.

Greyscale. Reducing colour saturation makes the interface less rewarding. The evidence is thin and the reports from people who try it are consistent enough to be worth an experiment.

Charging outside the bedroom. Which addresses the phone problem and the sleep problem simultaneously, and sleep is the larger effect.

What does not work

Willpower. Discussed elsewhere on this site, and it applies with particular force here because the opposing effort is professional.

Screen time reports. Awareness is not an intervention. Most people who check their weekly figure are briefly alarmed and change nothing.

App limits that can be dismissed. A limit with a "remind me later" button is a suggestion.

The proportionality point

Worth stating, because this topic attracts moralising.

Phones are genuinely useful. Messaging is how a great many people maintain relationships. The evidence on screen time and wellbeing is considerably more mixed and more modest than the coverage suggests, and much of it cannot distinguish cause from effect.

The defensible claim is narrower: that continuous availability fragments attention, that fragmented attention makes certain kinds of work impossible, and that this is worth managing deliberately.

Not that phones are ruining everything, which is neither supported nor useful.