Why “use it less” keeps failing

The phone stopped being a thing you own and became the place your life is kept. It is the bank, the camera, the maps, the school group, and for many people the office.

That is not a personal failing, and here it is close to universal. Statistics South Africa’s General Household Survey 2024, released in May 2025, found that 96,1% of households owned at least one mobile phone and 82,1% had internet access by some means, while the share of households still using a landline telephone had fallen to 3,4%, from 25,5% in 2002. For most South African households the phone is not a distraction from the real tools. It is the tool.

So a detox asks you to give up the thing your salary, your groceries and your school notices run through. You manage a weekend, and by Tuesday you are back where you started with a layer of guilt on top. The research is no kinder. A 2022 systematic review in Mobile Media & Communication, led by Theda Radtke, gathered 21 studies covering 3 625 people asked to take timeouts from their phones or social apps. Effects “varied across studies”: some found improvement, some found nothing, and some found the timeout made wellbeing worse.

The useful question was never how many hours. It is which twenty minutes you actually want back, and what it would cost you to get them.

Be wary of certainty in either direction. A 2019 paper in Nature Human Behaviour by Amy Orben and Andrew Przybylski put the question to three datasets covering 355 358 young people, and found the link between technology use and adolescent wellbeing “negative but small, explaining at most 0.4% of the variation in well-being”. So the frightening headlines carry more than the data can hold, and four settings changes will not remake your mood. What they return is countable time: the twenty minutes at the end of an evening you cannot remember spending.

The interactions that actually cost you

Spend four minutes looking at what is happening. Not to feel bad about it: nobody’s screen time number is a verdict on their character.

iPhone. Settings, then Screen Time. Look at the last seven days, not today.
Android. Settings, then Digital Wellbeing and parental controls.

Read the number the phone gives you, not the one in your head. A 2021 meta-analysis in Nature Human Behaviour, led by Douglas Parry, compared self-reported use against logged use across 106 effect sizes. What people report “correlates only moderately” with what their devices record, and was “rarely an accurate reflection” of it.

Then ask three questions of that screen, and write the answers on paper.

1. Which app has the most pickups, not the most minutes? Minutes tell you where time went. Pickups tell you which app has trained a reflex, and reflexes are what settings changes interrupt.
2. Which app do you open without deciding to? Usually it opens between two other things — the kettle boiling, the lift arriving, the ad break.
3. Which one leaves the evening worse than it found it? Not which is a waste of time. Which you close feeling flatter than when you opened it.

Three app names. Not ten. A change made to three apps is still in place next month; a change made to thirty gets reversed by Wednesday.

Change one: triage, not silence

The instinct is to switch everything off. There is good reason to think that is the wrong move.

In 2019 Nicholas Fitz, Kostadin Kushlev and colleagues published a randomised field experiment in Computers in Human Behavior. For two weeks, 237 people received notifications as usual, batched three times a day, batched hourly, or not at all. The three-times-a-day group “felt more attentive, productive, in a better mood, and in greater control of their phones”. Hourly batching did close to nothing. The group receiving none at all got few of the benefits while reporting “higher levels of phone-related fear of missing out” and “significantly more anxiety than controls”, and kept checking about as often as everyone else. One study, young sample — but the shape matches what people report after blanket silence: the interruptions stop and the checking does not, because now you have to check.

So triage instead. Go through the notification list once, app by app, sorting each into three buckets.

KEEP IT INSTANT
  A human being is trying to reach you.
  Calls, WhatsApp, messages, your calendar.
  Money moving, or a code you need to log in.

PUT IT IN A SUMMARY
  Things you want today, not now.
  News, email that is not from a client,
  delivery updates, group chats, sport.

TURN IT OFF ENTIRELY
  A company wanting your attention back.
  Games, shopping, streaming, anything that
  says "we noticed you haven't been on lately".

iPhone. Settings, then Notifications, then Scheduled Summary. Switch it on, choose two or three delivery times, tick the apps that belong there. For the off pile: Settings, Notifications, tap the app, turn off Allow Notifications.
Android. Settings, then Notifications, then App settings, and work down the list. For quiet hours, Settings, then Modes, then Notification filters, where you allow specific People and Apps. Google’s documentation flags the setting that matters most for safety: “To let a call through if the same person calls twice in 15 minutes, turn on Allow repeat callers.”

Who this will not work for

If your income depends on answering within minutes — a plumber, a tow operator, anyone whose enquiries arrive on WhatsApp and move to the next name on the list if you are slow — do not summarise the channel your work comes through. Leave that one instant and triage everything else. A rule that silences your work is not a reset.

Change two: take the colour out

Turning the screen black and white is the cheapest change here, so here is the evidence rather than the claim.

Cynthia Dekker and Susanne Baumgartner ran a preregistered study, published in Mobile Media & Communication in 2024, in which 84 Android users, average age just under 22, had their phone use logged for six days and then spent six days in greyscale. Daily screen time fell by about 20 minutes. People reported more control and less stress. Sleep quality and productivity did not change at all, and unlocks did not move — roughly 80 a day in both periods. The checking reflex survived; what shrank was how long each visit lasted. One more detail: participants rated greyscale annoying rather than helpful, and just over half said they would never or rarely use it again. Twenty minutes a day is a real return on a two-minute change, but it is not a transformation and you may dislike it.

iPhone. Settings, then Accessibility, then Display & Text Size, then Colour Filters. Switch it on and choose Greyscale. Then the important part: Settings, Accessibility, Accessibility Shortcut, tick Colour Filters. A triple-click of the side button now flips colour back for the two minutes you need to look at a photograph.
Android. Settings, then Digital Wellbeing and parental controls, then Bedtime mode, then Bedtime routine. Google lists Grayscale among the options there: “Changes your screen to black and white.” That gives you greyscale on a schedule, which for most people is the version worth having. Some phones also have a full-time switch under Accessibility, filed under vision or colour, though the wording differs by manufacturer.

Who this will not work for

Anyone who works in colour: photographs, paint, checking that a client’s brand blue is the right blue, reading charts where the colour is the information. The scheduled version solves most of that — colour in working hours, grey after eight. One group more: Colour Filters is an accessibility feature, and for some people with low vision, colour is what makes a screen readable. If greyscale makes text harder for you to pick out, leave it off and use the other three.

Change three: three apps off the home screen

The reflex here lasts about a second. Phone out of the pocket, thumb to a memorised position, app open — and only then does the conscious part of you arrive to find out where it is. No decision happened.

Friction interrupts that. The best evidence comes from a 2023 study in PNAS by David Grüning, Frederik Riedel and Philipp Lorenz-Spreen, who tracked 280 people using an app called one sec, which inserts a short pause before a chosen app opens. Across six weeks the pause stopped people going on into the app in about a third of attempts — 43% in the first week, settling at 32% to 34% — and by week six they opened those apps 57% less often than at the start. Weight that carefully: these were people who went looking for a self-control app, which is not a random sample, and the app is a paid product. What it supports is the mechanism — a small gap between impulse and app changes what happens next. The free version of that gap is moving the app two seconds further away.

iPhone. Press and hold the icon, choose Remove App, then Remove from Home Screen. Read that screen properly: Remove from Home Screen keeps the app and everything in it. Delete App is the other button, and not the one you want, particularly on a banking app. The app now lives in the App Library, one swipe left of your last page, reached by swiping down and typing its name.
Android. Press and hold the icon and drag it to Remove. Again: Remove takes it off the home screen, Uninstall takes it off the phone. The app stays in your app drawer and in search.

Who this will not work for

People who already open everything by searching — if your habit is to swipe down and type two letters, the app was never really on the home screen. And the harder case: when the app you want to interrupt is one you need. If it is WhatsApp, you cannot bury the way your family, your work and your building’s security group all reach you; the lever is inside the app instead, muting the groups that make most of the noise and leaving real messages alone. If the answer is the browser, greyscale and the evening habit will do more than rearranging icons.

Change four: the phone charges somewhere else

This change has the biggest effect on most people’s evenings and the least tidy research behind it. The story everyone has heard is blue light, and the evidence is thinner than the confidence around it. A 2020 systematic review and meta-analysis in SLEEP Advances, by Ari Shechter and colleagues, gathered 12 studies of interventions that cut short-wavelength light in the evening. Samples ran from 6 to 21 people each, and the authors found “some, albeit mixed, evidence” that it helps, with effects showing up more in what people said about their sleep than in what was measured. Night mode is not the mechanism.

The mechanism that needs no study is more ordinary. A phone within arm’s reach at eleven at night is a decision you keep making at the hour your judgement is worst. Move it to the kitchen and the decision gets made once, at nine, by a version of you who is thinking clearly.

So put the charger somewhere that is not your bedroom, and buy the cheapest alarm clock you can find, because “but it is my alarm” ends this change more often than every other reason combined. It is also the rule doing the heaviest lifting in our guide to keeping kids safer online, and it lands better in a house where the adults do it too.

Who this will not work for

A real list, not a token one. People on call — nurses, security, IT support, anyone on a duty roster. Parents whose teenagers are out. People caring for an elderly relative who might phone at two in the morning. Anyone whose street or complex WhatsApp group is a genuine part of how the neighbourhood stays safe. Anyone living in one room, where “another room” does not exist. And households still on load reduction: Eskom said on 4 August 2026 that the country had passed 441 days without load shedding, but that it “continues to implement load reduction in some areas affected by illegal connections, meter tampering, infrastructure damage and overloading” — so the kitchen plug may be dead exactly when you need it.

If you are on that list, run the compromise. The phone stays in the room but across it, face down on the far wall — far enough that answering means standing up. Set a Focus on an iPhone, or a Mode on Android, that lets through calls from favourites and repeat callers and silences the rest. You keep the line open for the thing you are worried about and close it to the other four hundred.

One caveat, since it matters most here: we are not doctors and this is not sleep advice. If you have been sleeping badly for weeks, that is a conversation with a GP, not a settings change.

An evening that survives a bad day

Settings hold the line on an average day. What decides whether any of this is standing in a month is the bad ones — the evening the client shouted, the invoice bounced, the child would not sleep. A routine that only works when you are in a good mood is not a routine. It is a mood.

Three parts do the work. An anchor, which is an event rather than a time, because 21:00 is easy to argue with and “when the kettle goes on” is not. A landing place, one spot the phone goes to, so putting it down is a motion rather than a decision. And something to do instead, because an evening with a hole in it gets filled by whatever is nearest, and the nearest thing is always the phone.

THE ORDINARY EVENING

  Kettle on for the last cup  -> phone goes on
                                 the shelf by
                                 the door
  Alarm set on the clock, not the phone
  One thing that is not a screen: book, guitar,
    dishes, a walk to the gate, a conversation
  Phone charges in the kitchen overnight

THE BAD EVENING

  Phone face down on the far side of the room
  Thirty minutes before bed. That is all.
  Tomorrow, the ordinary version again.

  No making up for it. No stricter rules as
  punishment. Tomorrow is only tomorrow.

The bad-day version is not the good version with more willpower applied. It is a smaller thing, chosen in advance, so that on the night you have nothing left you are not negotiating with yourself.

Give it two weeks, and judge it on the right question: not whether the screen time number fell, but whether you reached the end of an evening and could remember how you spent it.

The week you fall off it

You will fall off it. A deadline, flu, a fortnight of visitors — and one evening you look up at midnight with two weeks of the routine gone.

What happens next is the whole thing. Most resets do not end because the settings stopped working. They end the day after a broken streak, when the story turns from “I am doing this” into “I tried that once”, and picking it up again feels like admitting something.

So take the pressure out of it. In 2010 Phillippa Lally and colleagues at University College London published a study in the European Journal of Social Psychology following 96 volunteers building a new daily habit. It took a median of 66 days for the behaviour to feel automatic, with a huge spread — 18 to 254 days, depending on the person and the habit. And the finding that matters here: “missing one opportunity did not materially affect the habit formation process”. A missed evening is one evening, and the habit underneath does not notice.

Do not restart from zero. There is no counter and nothing reset. Tomorrow evening the kettle goes on and the phone goes on the shelf, exactly as it did a fortnight ago.

Do not run a stricter version as penance. The instinct after a bad week is to add rules: nothing after seven, apps deleted, a full detox to make up for it. That version is harder than the one you could not sustain, and it takes the original down with it.

Change the setting, not yourself. If the same app got past you three nights running, that is information about the setting, not about your character. Move it further away, put it in the summary, add it to the Focus. The point is to keep the environment doing the work, so your willpower is not the load-bearing part.