Monday, 9:04am. You open the laptop and forty-one things happen at once. Three emails you meant to answer on Friday. A meeting you forgot you accepted. A task that was due "next week", which has now become this week. A vague sense that there was something important and you can't remember what. By 9:30 you've done nothing except react, and the week's shape has been decided for you.

The weekly review is twenty minutes, once a week, that stops that happening. It's not a productivity system. It's closer to tidying the kitchen before you go to bed so you don't start tomorrow in yesterday's mess.

Why twenty minutes, and why weekly

Daily planning is too close to the work to see it; you're inside the week, reacting. Monthly is too far away; by the time you look back, the details that would have taught you something have gone. A week is the unit at which patterns become visible and are still fixable. And twenty minutes is long enough to do it properly and short enough that you'll actually do it, which beats a beautiful ninety-minute system you abandon by February.

Friday late afternoon or Sunday evening: both work, and they do different jobs. Friday closes the week while it's fresh and lets you leave work at work. Sunday sets up Monday but costs you a slice of weekend. Pick whichever you'll keep, set a recurring twenty-minute block in the calendar, and treat it like a meeting with someone you'd be embarrassed to stand up.

The six questions, in order

The order matters more than people expect. Clear first, then look back, then look forward. If you start with planning, the unfinished stuff from this week quietly poisons it.

1. What's loose? (5 minutes)

Empty the places things accumulate: inbox, notes app, the scraps of paper, the messages you starred, the "I'll deal with that later" pile in your head. You're not doing any of it. You're writing each thing down in one place as a next action — not "Dave's thing" but "reply to Dave about the invoice date". Most Monday dread is just un-captured loose ends, and the act of capturing them removes most of the dread before you've done a single task.

2. What did I actually do? (3 minutes)

Scroll back through the calendar and the sent folder. Write down the three to five things that got done. This isn't self-congratulation; it's calibration. People systematically underestimate what they did and overestimate what they'll do, and seeing the real list corrects both. It also builds a record you'll be grateful for at review time, when "what did you achieve this quarter" otherwise produces a blank stare.

3. What didn't happen, and why? (3 minutes)

The task that rolled over for the third week. The thing you were "going to get to". For each one, a one-line reason, honestly: too big, didn't know how to start; not actually important; blocked on someone; kept getting interrupted; didn't want to. The reason tells you the fix. Too big gets broken down. Not important gets deleted — properly deleted, not rolled over again. Blocked gets a chase. "Didn't want to" is the interesting one, and usually worth thirty seconds of wondering why.

4. What's coming? (4 minutes)

Look at next week's calendar as if it were someone else's. Where are the collisions — two heavy things on the same day? Where is there no gap at all? Which meetings need something prepared beforehand, and is that preparation in the plan or just assumed? This is where you catch the Thursday that was going to flatten you, while there's still time to move something.

5. What are the three things? (3 minutes)

Not a to-do list. Three outcomes that, if they're the only things that happen next week, mean the week was a success. Write them as results, not activities: "proposal sent to Mark" rather than "work on proposal". Then put the first real step of each into a specific slot in the calendar. A priority without a time is a wish, and that's why most to-do lists fail.

6. What's one thing to do differently? (2 minutes)

One. Based on what question 3 told you. "Do the hard task before opening email." "Say no to the Tuesday call." "Block Wednesday morning." Write it on the top of next week's page where you'll see it. Over a year this question alone quietly fixes most of what's wrong with how you work, one small adjustment at a time.

Where to keep it

Anywhere you'll open again. A single notes file with a heading per week works for most people; a paper notebook works for people who think with a pen. What doesn't work is a new app you install for the purpose, because now the review has a login. Keep the six questions as a template at the top of the file so you never start from blank, and never spend the twenty minutes redesigning the template, which is the most seductive way to avoid doing the review.

How to keep it past week three

Week one feels great. Week two feels useful. Week three, something comes up on Friday at four and you skip it, and that's where most people's weekly review ends. A few things that protect it:

Make it smaller before you make it skippable. If you've only got eight minutes, do questions 1 and 5. A short review beats a missed one by a mile, and the habit survives.

Pair it with something pleasant. The good coffee. The chair by the window. Twenty minutes of not being interrupted is a small luxury if you frame it as one, and framed as a chore it will lose to any chore that's louder.

Don't do it at the desk you're trying to escape. A different seat, a different room, even the kitchen table. The change of place tells your brain this is the overview, not more of the work.

Notice the first Monday it works. There'll be one, probably in week two or three, where you sit down at nine and know exactly what the day is for. Pay attention to how that feels. That feeling is what you're protecting when Friday at four tries to eat the slot.

What it isn't

It's not a life audit. Resist the temptation to add questions about your values, your five-year plan and your relationship with your mother. Those are real questions; they belong in a different sitting. The weekly review is small and operational on purpose, because small and operational is what survives.

And it's not a guarantee. Weeks will still go sideways. The difference is that they'll go sideways from a known position, with your three things written down and the loose ends captured, rather than from the 9:04am ambush. You can recover from the first. The second just happens to you.

Related: How to Focus for 90 Minutes When Your Attention Span Is Shot and How to Beat Procrastination When You Already Know Why You Procrastinate.

You said yes to the extra report. You said yes to covering the Thursday call. You said yes to "just having a look" at the new starter's work, and yes to the working group that meets at 5pm. Each one was a small, reasonable yes to a person you like. Together they mean that the one thing you were actually hired to do is now being done at 9pm, badly, and you've started to feel a flicker of dread when a message notification appears.

Here's the part nobody says out loud: the people who say yes to everything don't end up with the best reputation. They end up with the reputation for being late, stretched and slightly frantic, because that's what saying yes to everything does to a person. The colleagues who are trusted with the big things are usually the ones who've learned to say no to the small ones, cleanly.

Why the yes-reflex is so strong

Three things, usually. You want to be seen as helpful, and "no" feels like the opposite. You're imagining a reaction — disappointment, irritation, a mark against you — that in practice almost never arrives. And you're answering in the moment, when the cost of yes is invisible and the awkwardness of no is right in front of you. The cost of yes shows up on Thursday night. The awkwardness of no lasts about four seconds.

So the first move isn't a script. It's a delay.

Move 1: never answer in the moment

Buy yourself an hour, every time, for anything that isn't trivial. "Let me check what I've got on and come back to you by lunch." That sentence is almost impossible to object to, and it moves the decision out of the room where social pressure lives and into a place where you can look at your actual week.

In that hour, ask one question: if I say yes to this, what gets worse? Something always does. Either another piece of work slips, or your evening does, or the quality of the thing you're agreeing to does. Name which one. Then you can say no with a reason, or yes with your eyes open, instead of yes by reflex.

Move 2: the trade-off no

This is the one to learn first, because it's the one that most improves how you're seen. You don't refuse; you make the cost visible and hand the choice back.

"I can do that. If I take it on this week, the quarterly numbers move to next Wednesday — is that the right trade?"

Notice what it does. It shows you understand your own workload, which is a senior skill. It makes the person asking own the consequence rather than discover it later. And it quite often ends the request on the spot, because when the asker has to choose between their new thing and your existing thing, they frequently decide their new thing can wait. You've said no without the word appearing.

With your own manager, this is the only version you need. Managers don't want you to refuse work; they want to know what it displaces, and most of them would rather hear it on Monday than find out on Friday. If the workload is genuinely too much, say so plainly — Acas, the UK's workplace advice body, is direct that too many or conflicting demands are a cause of work stress and that the right first step is to talk to your manager as soon as you can, not to absorb it quietly until something breaks.

Move 3: the scoped yes

Sometimes you want to help but not at the size being asked. Shrink it.

"I can't take on the whole review, but I can give you forty-five minutes on Tuesday to go through the risky section with you."

A scoped yes gives the person something real, keeps you out of a three-week commitment, and — this is the useful bit — is usually what they actually needed. Most requests are inflated at the asking. People say "can you own this?" when they mean "can you help me not get this wrong?" Offer the second thing and both of you win.

Move 4: the clean no

For things that aren't yours, aren't worth a trade, or that you simply can't do: a short, warm, complete no. The mistake people make is padding it until it stops being a no.

"I'm not going to be able to take that on — I'm at capacity through the end of the month. Priya has done that process before, if that helps."

Three parts: the no, one honest reason, and if you have one, a pointer elsewhere. Then stop. Don't apologise three times. Don't list your whole workload as evidence. Don't say "but let me know if you're really stuck", which reopens the door you just closed. A clean no is respectful precisely because it's short; it treats the other person as an adult who can hear it.

And then — this is the step people skip — be reliable about the things you did say yes to. A no is believed when it comes from someone whose yeses are solid. That's the whole reputational logic: you're not the person who refuses things, you're the person whose commitments mean something.

Move 5: the "not me, and here's why" no

Occasionally you're asked to do something that isn't just inconvenient but wrong for you to do: outside your role in a way that matters, something you're not trained for, something that would mean cutting a corner you shouldn't cut. That no is different. It's not about capacity; it's about fit.

"I don't think I should be the one signing that off — I don't have the background to check it properly, and if it's wrong it'll look like I did. Can we get it to someone who does?"

Said calmly, that is not insubordination. It's the thing a careful colleague does, and the people who matter will notice that you noticed.

When the no needs to go above your manager

Most nos stay between you and the asker. A few shouldn't. If you've used the trade-off no and the honest conversation about workload more than once and nothing changes — the work keeps coming, the trade-offs keep being "do both" — that's no longer a time-management problem, and continuing to absorb it will cost you your health or your reputation or both. In the UK, that's the point to put it in writing to your manager (a short email summarising what you've raised and when), and if that doesn't move things, to HR or your manager's manager, with the same email trail attached. You're not complaining. You're documenting a conversation you've already tried to have.

The things that make this easier

Know your three priorities. If you can't say what the three most important things on your plate are this month, every request looks equally urgent and you'll say yes to all of them. Write them down. A request that threatens one of the three gets a trade-off no by default.

Practise on low stakes. Say no to the optional meeting. Decline the lunch you don't want. The muscle is the same, and it's much easier to build it where nothing hangs on it.

Notice who you can't say no to. Most people have one or two. That's information about the relationship, not about the requests, and it's worth sitting with.

What you'll notice

The first few nos feel enormous. Nobody else will notice them at all. What they'll notice, over a couple of months, is that you've become someone whose yes can be relied on, whose workload is visible, and who doesn't look hunted at 4pm. That isn't the reputation of someone who refused to help. It's the reputation of someone who knows what they're for.

Related: How to Ask for a Raise When You've Only Been There a Year and How to Set Boundaries With Family.

You already walk. To the shop, to the station, round the block when the house feels small. And on most of those walks you are not there at all: you're rehearsing a conversation, re-reading an email in your head, listing what's left to do. The legs move; the mind stays at the desk.

An awe walk is the same walk with one change. You point your attention outward, on purpose, at things that are bigger than you. It takes fifteen minutes, it needs no equipment or countryside, and in the one decent trial of it, it did something that "trying to relax" rarely does.

What the study actually found

A team at the University of California, San Francisco, led by Virginia Sturm with the emotion researcher Dacher Keltner, published a small trial in the journal Emotion (online in 2020, in print in 2022). Sixty healthy older adults were asked to take a fifteen-minute walk once a week for eight weeks. Half were given no particular instructions. The other half were told to take an "awe walk": to go somewhere new if they could, to notice things they hadn't noticed before, and to let themselves be struck by what was large, old, intricate or surprising around them.

Everyone took a selfie at the start and end of each walk. Over the eight weeks, the awe-walkers' selfies changed: they made themselves smaller in the frame and the scenery larger, and they smiled more. They also reported more awe, more joy and more of the warm, other-directed feelings the researchers call "prosocial positive emotion", and a bigger drop in day-to-day distress than the control group. Anxiety, depression and overall life satisfaction didn't move in either group, which is worth saying plainly: this is not a treatment for anything. It's a small, cheap shift in how a walk feels, with a measurable effect on mood and on how large you loom in your own picture of the world. (Sturm et al., 2022, Emotion.)

The study was small and the participants were in their sixties and seventies, so hold the numbers loosely. But the idea underneath it is old, and you can test it on yourself for the price of fifteen minutes.

Why "smaller self" is the point, not a side effect

Most of what wears us down day to day is scale. The email, the comment, the thing we said in the meeting — they feel enormous because, inside a head, there's nothing to compare them to. Awe is the experience of being in front of something that doesn't fit: a sky, a 300-year-old tree, the number of people on a station concourse all going somewhere. For a moment your own concerns are measured against something vast, and they shrink to their actual size. The worry doesn't vanish. It stops being the whole picture.

That's why this works where "just relax" doesn't. Relaxing asks you to turn the volume down on yourself, which is hard. Awe turns the volume up on everything else, which is easier, and the self gets quieter by comparison.

How to do one

You don't need mountains. The study was done in ordinary neighbourhoods. You need fifteen minutes and three decisions.

Go somewhere slightly new. Not a different city; a different street. Novelty makes you look. The route you've walked four hundred times has gone invisible, and you'll have to work harder to see it. If a new route isn't possible, go at a different time of day, when the light is different and the same street is not quite the same.

Put the phone away, and the headphones. A podcast is someone else's attention filling yours. The whole exercise is about where your attention goes, so for fifteen minutes it goes where you send it. If you need the phone for a route, check it once and pocket it.

Look up, look close, look at scale. Those are the three places awe tends to live. Up: the sky, the tops of buildings, the undersides of trees, which almost nobody looks at. Close: the pattern in bark, moss on a wall, the structure of a leaf, frost on a railing. Scale: anything that's old, big or many — a church that's been there since before your country had its current name, a river that was here first, a crowd. Pick one and stay with it for longer than feels natural. Twenty seconds on one thing, not two seconds on ten.

Then walk on, and when the mind drifts back to the email, notice it and send it back out. It will drift. That's not failure; that's what the exercise is for.

In a city, specifically

People assume awe needs nature. It helps, but cities are full of it if you break the habit of not looking. Stand at the bottom of a tall building and look straight up its face. Walk through a station at rush hour and notice how many separate lives are passing you. Find the oldest thing on your street; most streets have something — a boundary stone, a church, a tree that predates the houses. Go to the top of a car park at dusk. Walk by a river and think about where it's been. None of this costs anything, and all of it is awe-shaped, which is the point: the feeling doesn't require the Lake District, it requires you to stop and let something be bigger than your afternoon.

A few things that get in the way

Trying to feel awe. You can't force it, and chasing it makes it hide. Your job is to look properly. The feeling comes or it doesn't; the looking is still worth doing.

Photographing instead of seeing. One photo at the end is fine (the study used them). Twenty photos during turns the walk into content, and you're back inside your own head, composing. Look first. Photograph, if at all, last.

Making it a performance. Nobody needs to know you're doing this. It looks exactly like a person going for a walk, which is what it is.

Waiting for a nice day. Fog, rain and wind are often better than sun. Weather is one of the most reliable sources of "bigger than me" on offer, and it comes to your door.

How to make it stick

Once a week is what the study used, and once a week is enough to notice a difference by about week four. Attach it to something that already happens: Sunday morning before the house wakes up, the walk back from the school run, the lunch hour on the one day you don't have meetings. Put it in the calendar as "walk" and don't tell the calendar what kind.

At the end of each one, before the phone comes out, ask yourself one question: what did I see that I wouldn't have? Keep the answers somewhere — a line in a notebook, a note on the phone. After two months you'll have eight things you noticed that you'd have walked past for the rest of your life, and a slightly different relationship with the email.

What it is and what it isn't

It's a walk. It is not therapy, not a cure for low mood, and the study behind it is one small trial. But it's also free, it's fifteen minutes, it fits into a life, and the mechanism underneath it — put yourself in front of something large and let your problems find their size — is one that people across every tradition have been using for as long as there have been people and skies. The research just gave it a name and a stopwatch.

Related: How to Take a Secular Sabbath and How Many Steps Do You Really Need?.

Someone you love is telling you about their terrible day. Thirty seconds in, you've already got the solution. By the time they finish the sentence you're saying "have you tried…", and their face does a small thing you've seen before: it closes. They say "yeah, maybe," and change the subject. You've helped, technically. They feel worse, and so, oddly, do you.

This happens in most relationships many times a week, and the person doing it is almost never being unkind. They're being useful in the wrong currency. This is about how to notice it, and what to do instead.

Why fixing feels like listening (and isn't)

When someone tells you a problem, your brain does what brains do: it starts solving. That's not a flaw. It's just that solving and listening use the same minutes, and the solving takes over. While you're assembling the answer you've stopped hearing the middle of what they're saying, which is usually where the actual thing is.

There's a second reason fixing is so tempting. Someone else's distress is uncomfortable to sit with, and a solution ends the discomfort — for you. The other person didn't ask for the discomfort to be ended. Mostly, they asked to be heard while they had it.

The test is simple. After you've "helped", does the person seem lighter, or do they seem managed? If it's the second, you solved your own problem, not theirs.

Three things to stop doing first

Better listening is mostly subtraction. Before learning any technique, remove the three habits that do the most damage.

Stop finishing their sentences. Even when you're right about where the sentence is going, you've just told them you were ahead of them rather than with them. Let the sentence land. If there's a pause afterwards, let that land too; a lot of the important things people say come after the pause they expected you to fill.

Stop matching their story with yours. "That happened to me, when I…" feels like empathy and functions as a takeover. There's a version that works — one sentence, then straight back to them — but most of us don't do the "straight back" part. If in doubt, keep your story for later. It'll still be true tomorrow.

Stop reaching for the phone. Not just answering it. Having it face-up on the table, glancing at it when it lights. The Samaritans' advice for anyone trying to help someone open up starts with the obvious things for a reason: phone away, face them, give them your whole attention. People can tell when they have seventy percent of you, and they adjust what they're willing to say accordingly.

The one question that does most of the work

Here it is: "Do you want me to help think about it, or do you just want me to listen?"

It sounds clunky the first few times. Ask it anyway. It does three useful things at once. It stops you from launching into solutions. It hands the other person control over what kind of conversation this is, which is itself a form of being heard. And it means that when they do say "actually, yes, what would you do?", your advice arrives as something they asked for rather than something you pushed.

About half the time, in most people's experience, the answer is "just listen". The other half you get to fix things with their permission, which goes far better than fixing without it.

Between partners, agree on a shorthand so neither of you has to ask the full question every time. Some couples use "venting or solving?"; others just hold up one finger for "let me finish". Whatever it is, having it means the fixer doesn't have to guess and the talker doesn't have to feel interrupted.

What to say instead of a solution

"Just listen" doesn't mean sit there in silence like a wall. It means your contributions keep the attention on them. Four kinds of sentence do that.

Say back the thing they actually said. Not a parrot, a summary: "So it's not really the extra work, it's that nobody asked you first." If you've got it right, they'll relax visibly and keep going. If you've got it wrong, they'll correct you, which is also progress; now you both know what the thing is. This is the single most powerful move in listening and it costs nothing.

Ask about the part they skipped. People rush past the bit that matters. "You said the meeting was 'fine' quite quickly — what happened in it?" You're not interrogating; you're showing you noticed.

Name the feeling, tentatively. "That sounds humiliating" or "You sound more sad than angry about it, actually." Tentative matters. You're offering a word, not assigning one. If it's wrong they'll give you the right one, and you've just helped them find it.

Ask what they want to happen. Not "what are you going to do", which pushes them toward a plan. "What would you want, if it were up to you?" People often don't know until they're asked, and finding out is worth more than any plan you could hand them.

The silence you have to learn to stand

This is the hard part and nobody warns you. When you stop fixing and stop filling, there are gaps. Three, four, five seconds where nothing is said. Every instinct says to rescue the moment.

Don't. Count to five in your head if it helps. The pause is where the person decides whether to say the next, truer thing, and if you fill it they'll never have to. Most of the conversations people remember as "they really listened to me" contained a pause that the listener survived.

When advice is actually wanted

Sometimes they say "yes, help me think." Then help, but notice how differently it goes if you do these things first: ask what they've already considered (so you don't suggest the obvious thing they tried last week), offer options rather than an instruction ("two things come to mind…"), and leave the decision clearly with them. Advice given this way gets used. Advice fired across the table in second thirty gets politely ignored, which is why your previous excellent advice never seemed to land.

A note on the people you listen to least

We listen worst to the people we've known longest. We think we already know what they're going to say, so we stop hearing them say it, and some of the loneliest people in the world are married. If one relationship in your life would benefit from this, it's probably the one where you'd swear you already listen. Try the "venting or solving" question there first. The reaction will tell you how overdue it was.

This week

Pick one conversation a day, with anyone. Phone away. No matching stories. Say back one thing they said. Let one silence run. That's it. You won't become a different person by Friday, but the people you talked to will have noticed something they can't quite name, and they'll come back to tell you more. Which is the whole point.

Related: Fighting Fair: How to Argue With Someone You Love and How to Talk to Your Teen So They Actually Talk Back.

You've been deciding for three weeks. You have fourteen tabs open. You've asked four people, who said four different things. You've made a spreadsheet, which was honestly more about feeling productive than deciding. And every evening the choice is still sitting there, exactly where you left it, a little heavier.

The problem is rarely the decision. It's that "think about it more" feels like progress and isn't. What follows is a thirty-minute method for getting unstuck, in the order that works.

Minute 0–5: work out which kind of stuck you are

There are four, and they need different fixes, so this is worth the five minutes.

Too many options. Six flats, nine laptops, every possible holiday. You're not stuck between two things; you're drowning in choice. Fix: criteria, below.

Fear of regret. Two options, both fine, and you can already picture the version of you that chose wrong. Fix: reversibility and the 10-10-10, below.

Missing information. You genuinely don't know a fact that matters: the salary, whether the landlord allows pets, what the course involves. Fix: get that one fact, today, and nothing else. This is the only kind of stuck that more research helps, and it needs exactly one search, not fourteen tabs.

You already know, and you don't like the answer. Be honest. If you've been "deciding" about leaving for months, you've decided; what you're doing is working up to it. Fix: skip to the coin test.

Minute 5–10: sort it by how reversible it is

The single most useful question about any decision is: if this turns out wrong, what does it cost to undo?

Some choices are two-way doors. Pick a laptop, hate it, return it. Try a class, don't like it, stop. Move teams, move back. These deserve minutes, not weeks, because the cost of a wrong answer is small and the cost of three weeks' dithering is three weeks.

Others are one-way, or close to it: a house purchase, a move abroad, ending a marriage, a loan. Those earn real deliberation. But be honest about which category you're in. Most of what we agonise over is a two-way door we've dressed up as a one-way one, because treating it as momentous feels more respectful of the fear.

If it's two-way: give yourself the rest of the thirty minutes and then decide. If it's one-way: give yourself a week, with the deadline written down, and use the rest of this method inside that week.

Minute 10–15: three must-haves, set before you look again

If your problem is too many options, this is the fix. Before you open a single tab, write three things the answer must have. Three, not ten. "Under £900 a month, within 40 minutes of work, allows a cat." "Full-time, remote at least twice a week, doesn't involve selling." The first option that meets all three wins. You stop looking.

This feels like settling. It's the opposite. There's a well-known line of research, starting with a 2002 paper in the Journal of Personality and Social Psychology, that compared "maximisers" — people who try to find the best possible option — with "satisficers", who set a bar and take the first thing that clears it. Maximisers often ended up with objectively better outcomes and were less happy with them, with more regret and more second-guessing. Later studies have found the effect is smaller and messier than the first paper suggested, so hold it lightly, but the practical lesson survives: the search for the best option has a cost, and the cost is paid in your evenings.

Minute 15–20: cap the information

Three rules, all of them about limits.

A timer. Thirty minutes of research, once, with a clock running. When it rings, you're done researching, whatever you've found.

A source limit. Three sources. One expert or official page, one person who's done it, one review or comparison. The fourth source never tells you anything the first three didn't; it just gives you one more opinion to reconcile.

A deadline, with a date, written down. "I will decide by Friday 6pm." Then tell someone, because a deadline only you know about is a suggestion.

Minute 20–25: two questions for the regret-afraid

If you're stuck between two fine options because you're scared of picking wrong, run each option through these.

10-10-10. How will I feel about this choice in ten minutes? Ten months? Ten years? The ten-minute answer is usually fear, which is noise. The ten-month answer is the real one. The ten-year answer tells you whether this is even a one-way door — most things you can't remember deciding ten years ago.

The pre-mortem. Imagine it's a year from now and the choice went badly. Write one sentence on why. Do it for both options. Often one failure story is vivid and specific ("the commute broke me") and the other is vague ("I suppose I'd have been bored?"). The vague one is the safer choice. Specific fears are information; vague ones are just fear.

Minute 25–28: ask one person one question

Not four people. One, chosen because they know you and don't have a stake. And not "what should I do?", which invites a speech. Ask: "What would you do?" Then notice your reaction to their answer, because that's the data, not the answer itself.

If they say "take the job" and you feel relief, you wanted to take the job. If you feel a flare of "but you don't understand", you didn't. You've just learned which option you're defending, and you only defend the one you want.

Minute 28–30: the coin

For the people who already know. Assign the options to heads and tails. Flip it. Don't look at the result; look at what you hoped for while it was in the air. That half-second of hope is the decision, and it was there the whole time. The coin didn't decide anything. It just made you admit it.

After you decide: write it down and commit for a set period

Three lines, in a note you'll keep: what you decided, the main reason, and a review date. "Took the job. Reason: more interesting work, 10-month answer was clearly yes. Review 1 March."

Then commit until the review date and don't reopen it before. This is the part people skip, and it's why decisions keep coming back. A choice reviewed every Tuesday isn't a choice, it's a hobby. Pick a sensible period — thirty days for small things, three months for a job — and give the decision a fair run before you grade it.

The honest trade-off

Some of your decisions will be wrong. That's not a flaw in the method; it's the price of deciding at all, and it's a price you were already paying. The three weeks you spend not choosing have a cost too — the flat goes, the offer expires, the course starts without you — and because that cost is invisible, nobody puts it on the spreadsheet. Put it on the spreadsheet.

A decent decision made on Friday beats a perfect one made never. Set the timer.

Sources: Schwartz B, et al. "Maximizing versus satisficing: happiness is a matter of choice." Journal of Personality and Social Psychology, 2002;83(5):1178–1197. Related: How to Quiet an Overthinking Mind at 2am and How to Know When to Quit Your Job.

Somewhere in your bank statement there is a payment you don't recognise, a subscription you stopped using in February, and a broadband bill that quietly went up when the contract ended. You know this. Everyone knows this. The reason you haven't done anything about it is that "sort out my bills" is a vague, unpleasant task with no edges, so it sits on the list for a year.

This gives it edges. One hour, four timed passes, done once. Set a timer, because the timer is what makes it finishable.

Why it's worth an hour

Surveys of UK households regularly find that around half are paying for at least one subscription they don't use, and the typical waste comes out somewhere in the region of ten to fifteen pounds a month per household. Treat the exact figure loosely — it depends who's asking and how — but the shape is consistent and it matches what people find when they actually look. Add the out-of-contract broadband, the insurance that auto-renewed at a worse price, and the energy direct debit that's been over-collecting, and an hour's audit finding a few hundred pounds a year is ordinary, not lucky.

Notice what this isn't. It isn't an investing article or a side-hustle article. It's the boring end of money, the end where the gains are certain and nobody writes about them because "cancel the thing you don't use" doesn't sell a course.

Prep: five minutes

Download or open the last three months of statements for every account and card you pay things from. Three months, not one, because quarterly and annual payments hide in the gaps. Open a blank note or spreadsheet, or MoneyHelper's free budget planner if you'd rather have a structure handed to you. Make a cup of tea. Start the timer.

Pass 1: find every recurring payment (15 minutes)

Go down the statements and list every payment that repeats. Don't judge yet, just list. Name, amount, how often. You'll find the obvious ones in the first five minutes and the sneaky ones — the app you subscribed to on a free trial, the "annual" that was last year's impulse, the thing in your partner's name that you both forgot — in the last ten.

Then tag each line with one of three words:

Need. Rent, council tax, energy, insurance you're required to hold, the phone you actually use.

Use weekly. The streaming service you watched last night, the gym you went to on Tuesday. Keep, for now.

Haven't used in 30 days. Be honest. If you can't remember the last time, it's thirty days.

Pass 2: cancel the third column, today (10 minutes)

Everything tagged "haven't used in 30 days" gets cancelled before the pass ends, not added to another list. The test, if you waver, is: would I sign up for this again today, at this price? Not "might I use it one day". Today, at this price. Most of the column fails.

One genuine trade-off: a few things are good value even at once a month — a membership you use for one big thing in winter, say. For those, the move isn't "keep just in case", it's a calendar reminder: "Did I use this? If not, cancel." Dated, three months out. "Just in case" is how the column got long in the first place.

Use a two-line cancellation message for anything that makes you phone or email: "Please cancel my subscription effective today and confirm by email. I don't need a retention offer, thank you." The last sentence saves you from the script they're required to read you. If you signed up online or by phone within the last fortnight, you usually have a 14-day cooling-off right under UK consumer rules, which is worth knowing for the trial you forgot about.

Pass 3: the big three (20 minutes)

Subscriptions are the easy wins. The money is in the three bills most people never renegotiate.

Broadband and mobile. Find out whether you're in or out of contract. In the UK, Ofcom requires providers to tell you when your contract is ending and what their best deal is, and to remind you every year if you've stayed out of contract — so the information is in an email or a letter you ignored. Out of contract almost always means paying more than a new customer. Either switch, or phone and use this: "I've been a customer for six years and I'm out of contract. I can see [competitor] offering [price] for the same speed. What can you do to keep me?" Have the competitor's price actually open when you say it. Expect to be transferred to a retentions team; that's where the real offers live.

Insurance. Car, home, contents, anything that auto-renews. Get your renewal quote, get two comparison quotes, and if they're lower, phone your current insurer with the number. Loyalty gets you nothing in insurance; a quote gets you a match, or you switch.

Energy. Check two things. First, whether the tariff you're on is the best your supplier offers — prices move with the regulated price cap and a fixed deal may or may not beat it, so compare today rather than on memory. Second, whether your direct debit is building up a large credit balance; if you're hundreds of pounds in credit going into winter, you can ask for it to be reviewed or refunded.

Do what you can in twenty minutes. If broadband turns into a forty-minute phone call, that's fine; the audit's job was to get you to pick up the phone.

Pass 4: set the traps for future you (10 minutes)

You will end up here again in a year unless you lay a few tripwires.

Put every contract end date and annual renewal in your calendar, a month before, with the amount in the title so you feel it. Every free trial you start from now on gets a reminder two days before it converts. And add the three-month "did I use this?" reminders from pass two.

Then the line that makes the whole hour mean something: tally the annual saving and decide now where it goes. Not "I'll save it". A named destination: the emergency fund, the credit card, the thing you were putting off. Set up a standing order for the monthly amount to a separate pot on the same day. If you don't, the money dissolves back into the account and in six months you won't be able to prove the hour happened.

If you're not in the UK

The method is identical; only the names change. Every country with a regulated telecoms or energy market has a regulator with rules on contract notices and switching, and a government-backed money-advice service with a budget tool. Find yours, and run the same four passes.

What you'll have at the end

A single list of everything you pay for, with each line judged. A shorter third column. At least one phone call made or scheduled. A set of calendar reminders doing the remembering for you. And a monthly number, going somewhere on purpose.

That's an hour. It's less than the time most of us spend each week looking at things we might buy. Prices change and this article will date, so check current rules and tariffs when you do it — but the shape of the job won't change, and neither will the fact that nobody is going to do it for you.

Related: Emergency Fund 101 and How to Build Your First Budget.

Think about your last Sunday. Was it rest? Or was it the big shop, the washing, the admin you'd put off all week, forty minutes of the phone on the sofa, a vague sense of dread about Monday, and then bed, more tired than Friday?

Most people haven't had a day of genuine rest in years. Not a holiday — holidays are logistics with a sea view — but a plain, ordinary day where nothing was produced, fixed, bought or scrolled. Several of the world's oldest traditions built one into every week and guarded it fiercely. The rest of us let it get eaten by errands, and we feel it.

What a secular sabbath is, and what it's borrowed from

A sabbath, in the plain sense used here, is a fixed window every week — somewhere between a few hours and a full day — in which you deliberately do not do productive work. The idea comes from Jewish Shabbat and the Christian Sunday, and it's worth being straight about that rather than pretending it was invented by a productivity blog. Those traditions have rules, meaning and community that a secular version doesn't try to copy. What it does take is the structural insight: rest that happens "when there's time" never happens, and rest that is fixed, weekly and defended does.

You don't need to believe anything to use the structure. You just need to treat the window as if it were an appointment with someone you'd be embarrassed to cancel on.

Choose the window, and start smaller than you want to

A full day is the goal; four hours is the start. People who try for the whole of Sunday on week one usually break it by 11am, decide they're not the type, and that's the end of it. Four hours, kept, beats twenty-four hours, abandoned.

Pick the edges and write them down. The traditional shape, Saturday 6pm to Sunday 6pm, works surprisingly well for secular people because it protects Saturday evening and leaves Sunday evening for the gentle re-entry everyone needs. If that's too much, Sunday 10am to 2pm. Fixed edges are what make it a sabbath rather than a lie-in.

Prepare the day before

The day of rest only works if the day before does its homework. Shabbat has a whole rhythm of Friday preparation for exactly this reason: if the shopping isn't done, you'll do it on Sunday, and then it's Sunday as usual.

So on Saturday morning (or whenever your "day before" is): do the food shop, put a wash on, and write the to-do list for next week on paper and put it in a drawer. That last one matters more than it sounds. The list in the drawer is the thing that stops your brain running it on a loop at 3pm on Sunday, because it knows the list is safe.

Then tell the people who need telling. A partner, a housemate, the family group chat. Something like: "From 10 till 6 on Sunday I'm off duty — no admin, no errands, phone in a drawer. Here's what I'm cancelling in advance so nobody's surprised." You'll find people are far less bothered than you fear, and a few of them will ask if they can join in.

The no list: three things, not twenty

Resist the urge to write a monastic rulebook. Three prohibitions you'll actually keep are worth more than a long list you'll resent.

No work. No email, no "just checking", no laptop open "in case". If your job genuinely requires you to be reachable, define exactly how (one named person can phone you) and nothing else.

No chores or admin. No washing, no bills, no sorting the cupboard, no returning the parcel. If it would be on the list in the drawer, it waits.

No scrolling. This is the one that makes the day feel different. Phone in a drawer, or in another room, with calls allowed from named people. If you want to know how to do this without a fight, the doomscrolling plan has the mechanics; here, it's just one day, which is far easier.

Everything else is allowed. That's the point. The no list is short so that the yes list can be long.

The yes list

The strange thing you'll notice around hour two is that you don't know what to do with yourself. The day has always been full of tasks; take them away and there's a blank. Plan for the blank with a short list of things that count as rest for you specifically.

A long meal, cooked slowly or not cooked at all, eaten at a table. A walk without headphones. A paper book. A nap, which is not laziness, it's the thing the day is for. Seeing one person you actually like. Something with your hands: bread, a drawing, a jigsaw, a plant. And one small ritual that marks the window — a candle lit at the start, the same pot of tea at the same time, the same walk to the same bench. Rituals sound precious until you have one, and then they're the thing that makes the day feel like a different kind of day.

What the NHS calls the five steps to mental wellbeing — connect, be active, take notice, keep learning, give — fit inside a rest day almost without trying, which is a decent sign the structure is sound.

The guilt, and the exceptions

The first few weeks you'll feel guilty around mid-afternoon. A voice will say that useful people don't sit in a garden reading on a Sunday while the recycling piles up. Let it talk; it has nothing to say that the list in the drawer can't handle tomorrow. The guilt fades by about week four, and what replaces it is the thing you've been missing: a week with a floor in it.

Some lives can't take a day. If you're caring for someone, working shifts, or parenting small children, a full day is a fantasy and a four-hour window may be one too. Make it two hours. Make it a Tuesday morning. Share it in rotation with a partner — you take Saturday 2 to 5, they take Sunday. The principle survives scaling; the shape doesn't matter as much as the fact that it's fixed and defended.

Why not just "have a lazy Sunday"?

Because you've tried that, and it became the admin day with a nap in it. The difference between a lazy Sunday and a sabbath is the edges. One is what's left over when the chores are done; the other is a decision made in advance that the chores will wait. The first always gets smaller. The second holds.

There's a modest research literature on stepping away from screens for a week and on structured rest, and it points gently in this direction, but you don't need a study to test the idea. You need four weeks. Keep the window four Sundays in a row, then write one line about what changed. Most people find the week before the window gets more bearable, because there's something at the end of it that's actually theirs.

If after a month it's done nothing, you've lost sixteen hours to reading and walking, which is not the worst outcome. If it's done what it does for most people, you'll wonder how you managed without it.

Related: What to Do When You Feel Spiritually Empty and The Practical Guide to Gratitude.

Your watch wants 10,000 steps a day. You're doing about 3,500. So most evenings you get a little buzz of failure on your wrist, and after a few months you stop looking at it, which is the opposite of what the number was supposed to do.

Here's the thing nobody told you when you bought the watch: 10,000 was never a medical target. And the actual research says something far more encouraging for people who sit down for a living.

Where 10,000 came from

The number is usually traced to a Japanese pedometer sold in the 1960s whose name translates roughly as "10,000-step meter". It was a round, memorable figure chosen for a product, and it stuck because it's a round, memorable figure. It was not the output of a trial, and for sixty years nobody checked whether it was the right number because it sounded so plausible.

What the research actually says

In 2025 a team published a systematic review and meta-analysis in The Lancet Public Health pooling dozens of studies that tracked people's daily step counts and followed what happened to their health over years. The headline finding: compared with people doing around 2,000 steps a day, those doing around 7,000 had roughly 47% lower all-cause mortality over the follow-up periods. Benefits for several other outcomes rose steeply from low step counts and then flattened out somewhere around 5,000 to 7,000. Beyond that, more was still a little better for some things, but the curve goes flat fast.

Two honest caveats, because this is exactly the kind of statistic that gets mangled. First, it's observational. People who walk more differ from people who walk less in a hundred ways, and the studies adjust for what they can, but nobody randomised anyone to 7,000 steps. So "associated with" is the right phrase, not "will add years to your life". Second, the paper has a published erratum, which is normal for a study this size and doesn't change the shape of the finding, but it's worth knowing the figures have been corrected once.

What you can safely take from it is this: the biggest gains in the whole curve happen between "almost nothing" and "a moderate amount". If you're at 3,500, the move to 7,000 is where nearly all of the benefit lives. The move from 7,000 to 10,000 is a bonus. You've been grading yourself against the bonus.

Step 1: find your real baseline

Your phone has been counting already. Open the health app and look at three ordinary days — not the Saturday you went to the market, not the day you were ill. Average them. That's your baseline, and it's probably lower than you'd like, which is fine. The plan starts from where you are, not from where the watch thinks you should be.

Step 2: add a thousand a week

A thousand steps is about ten minutes of ordinary walking. That's the whole increment. Add a thousand to your baseline, hold it for a week, then add another thousand.

From 3,500 that looks like: week one 4,500, week two 5,500, week three 6,500, week four 7,000 and hold. A month, and you're in the part of the curve where the research gets interesting. From 5,000 it's a fortnight. The reason to go slowly isn't that your body can't handle more; it's that habits added at ten minutes survive and habits added at an hour don't.

Step 3: attach the walking to things that already happen

You will not "find time" to walk. Nobody finds time. You attach walking to anchors that are already in the day, so that the decision gets made once rather than daily.

The ones that reliably work for desk-bound people:

The after-lunch ten. Eat, then walk for ten minutes before you sit back down. A thousand steps, and most people report the afternoon slump is smaller, though that's your experience to test rather than a promise.

Call-walks. Any call where you don't need a screen, you take walking. One-to-ones, catch-ups with your mum, the call with the bank where you're on hold for twenty minutes anyway.

One stop early. Bus or train: get off a stop before yours. Driving: park at the far end. It's usually five to eight hundred steps and it needs no planning after the first time.

The long way to the loo. In an office, use the toilets on another floor. At home, make the tea in the kitchen and drink it standing in the garden or by the window. This sounds trivial and adds up to about a thousand a day for most people.

Step 4: the bad-weather, bad-day plan

Every walking plan written in good weather dies in November. Decide now what the rain version is, so you're not deciding at 6pm in the dark.

Indoor laps count. Three lots of five minutes — round the flat, up and down the stairs, round the office — is about 1,500 steps and nobody has to see you do it. A bad day doesn't need to hit the target; it needs to not hit zero. The streak you're protecting isn't "7,000 every day", it's "didn't skip".

Step 5: pace matters, a bit

Not every step needs to be brisk. But some of them should be, and the test is simple: at a good pace you can talk but you couldn't sing. That's the "moderate" intensity in the NHS adult activity guidelines, which ask for 150 minutes a week of it. The after-lunch ten and the call-walks, done at talk-not-sing pace, get you most of the way there without a gym in sight.

What steps don't do

Steps are a proxy. They measure one thing, moving about, and they measure it imperfectly. They don't replace the other half of the same NHS guidance, which is strength work on at least two days a week, and they're not a weight-loss plan; walking is one of the least efficient ways to burn energy and one of the best ways to be a person who moves, and those are different goals. If you've got a health condition, or you haven't been active for a long time, check with your GP before changing much, and start lower than this plan suggests.

What steps are good for is being honest. The watch doesn't care about your intentions. If it says 3,500, you sat down a lot, and a week of adding one anchor will change that number more than any amount of planning to join a gym.

The target, then

Find your baseline. Add a thousand a week until you're around 7,000. Attach it to lunch, calls and the commute so it stops being a decision. Have a rain plan. Walk briskly some of the time. That's it.

And change the goal on your watch. There's nothing noble about a target you ignore. Set it to 7,000, hit it, and let the little buzz on your wrist mean something again.

Sources: Ding D, et al. "Daily steps and health outcomes in adults: a systematic review and dose-response meta-analysis." The Lancet Public Health, 2025;10(8):e668–e681 (erratum 2025;10(9):e731). NHS, "Physical activity guidelines for adults aged 19 to 64." Related: How to Start Exercising When You Hate Exercise.

There's a name that comes to mind every few weeks. You see a shop, a song comes on, someone mentions a city, and there they are. You think, I should message them. Then you think, it's been three years, it'd be weird, and you don't.

This article is about the gap between those two thoughts, which is almost entirely imaginary.

The awkwardness is mostly in your head

In 2022 a group of researchers ran thirteen experiments on exactly this question: how do people feel when someone they've lost touch with reaches out, compared with how the sender expects them to feel? The result, published in the Journal of Personality and Social Psychology, was the same every time. Senders consistently underestimated how much the message would be appreciated. And the effect was strongest when the contact was a surprise — the very thing that makes you hesitate is the thing that makes it land.

That matches ordinary experience, if you flip it. Think about the last time an old friend messaged you out of the blue. Were you annoyed? Did you think "how odd"? Almost certainly not. You were pleased, probably more than you expected, and you probably told someone about it. The person on the other end of your message is you.

Two caveats so this doesn't become a rule. The research didn't look at reaching out after a proper falling-out, and it certainly didn't look at contacting someone who has asked you not to. If either of those is your situation, the advice below changes, and the last section covers it.

Pick one person, by the right criterion

Don't make a list of everyone you've ever lost touch with. That turns into a project and projects get postponed. Pick one person this week.

The criterion is "who would I be pleased to hear from?" Not "who owes me a message", not "who's useful to know", not "who would make me look good to be friends with again". Those are motives the other person can smell through a screen. The friend whose name you were glad to see on the shop sign is the one.

Three messages that work

The message has one job: to be easy to reply to. That means short, specific, and free of anything the other person has to apologise for.

1. The memory prompt. Best for a gap with no bad feeling in it, which is most gaps. Lead with the thing that made you think of them, because it tells them they exist in your head for a real reason.

"Just walked past a bakery with a terrible pun for a name and thought of you and that place in Leeds. How are you? Properly, not the LinkedIn version."

2. The no-guilt acknowledgement. Best when you both know the silence has been long and you'd rather name it than tiptoe. Own it in one line, then move on.

"I've been rubbish at keeping in touch and I've missed you. No excuses. Fancy a call on Sunday? Twenty minutes, I'll do the dialling."

3. The open door. Best after a long gap or a wobble you'd both rather not relitigate. It asks for nothing, which is what makes it safe to answer.

"You've been on my mind. No pressure at all to reply — I just wanted you to know I think of you fondly."

Notice what none of them do. None contain a six-hundred-word explanation of why you went quiet. The explanation feels necessary to you because you've been carrying the guilt; to them it's a wall of text that makes them responsible for reassuring you. One line, maximum. "I've been rubbish" covers it.

Turn the text into a plan within one exchange

Here's where reconnections die: a warm reply arrives, you reply warmly back, there's a flurry of "we should catch up!", and then nothing. "We should" is where friendships go to be remembered fondly.

So in your second message, propose something small with a date in it. Not dinner, not a weekend; those are too big to schedule and too easy to postpone. A twenty-minute call. A walk if you're in the same city. A specific day.

"Brilliant. Are you about Sunday afternoon for a call? Say 4? If not, throw me a time."

Small and dated is the entire trick. A twenty-minute call that happens beats a dinner that gets rescheduled three times, and once you've actually spoken, the next plan is easy because the awkwardness has been spent.

If they don't reply

Sometimes they won't. Life is loud, messages get buried, and the research only says people are glad to hear from you, not that they're organised about answering.

Send one gentle follow-up after two or three weeks. Something with zero edge in it: "No worries if you're flat out — just resurfacing this in case it got buried. Hope things are good." Then let it go. Not with a dramatic internal speech about how they've clearly moved on; just let it go. You did the kind thing, it cost you a minute, and the door is open on your side. That's a complete action.

Make it stick without making it a chore

The reason you drifted the first time is that friendship after your twenties doesn't maintain itself. Nobody's timetable throws you together any more. So give it a rhythm that's small enough to survive a busy month.

Options that actually work for people: a monthly call that lives in both calendars as a recurring event; a shared notes file or photo album you both drop things into; a voice-note thread, which is lower effort than typing and sounds like a real person. Pick one. Big gestures are lovely and unsustainable; the friendships that last are the ones with a boring, reliable channel.

The honest trade-off

Some friendships have ended, and that's allowed. You were friends because you sat next to each other for three years, and the thing you had in common was the seat. If you reach out and the call is pleasant and flat and neither of you suggests another, that's not a failure. You've answered a question you'd been carrying, and now you can stop carrying it.

The test is simple: does either of you miss it? If yes, the rhythm above is worth the effort. If not, you can wish them well and mean it.

And if the gap exists because something went properly wrong between you, the open-door message is the only one of the three to use, sent once, with no expectation. If they've asked for distance, respect it; being thought of fondly is not the same as wanting to be contacted, and that's their call to make.

For the ordinary gaps, though — the drift, the move, the years that just went — stop waiting for a less awkward moment. There isn't one. There's the message, which takes forty seconds, and the person on the other end, who will be gladder than you think.

Related: How to Make Real Friends After College and Solitude vs Loneliness.

You picked up the phone to check the time. Forty minutes later you know about a flood on another continent, a politician's bad week, three people's holidays and a dog that can say "mama", and you feel worse than when you started. You didn't decide to do that. Nobody does.

The usual fix is to delete the apps, which works for about nine days and then you're reinstalling them "just for messages". This is a different approach. You keep the apps. You change the conditions that make the loop so easy to fall into, one small change a day, for a week.

Why this isn't a willpower problem

Three things are stacked against you. Feeds are infinite, so there's no natural end — no last page, no "you're all caught up" that you actually believe. Bad news holds attention better than good news, so the algorithm serves more of it. And the phone is already in your hand at the exact moments your attention is weakest: in bed, on the loo, on the sofa after work.

Put those together and you have a system designed by very clever people to do precisely what it does. Blaming your discipline for losing to it is like blaming yourself for being out-run by a car.

Worth knowing, too, that the research here is more modest than the headlines. A scoping review of doomscrolling studies found it's consistently associated with higher anxiety, low mood and stress, but most of the studies are snapshots, so nobody can honestly say which direction the arrow points. Maybe scrolling makes you feel worse; maybe feeling worse makes you scroll. Probably both. Either way, the loop is worth breaking, and you don't need a diagnosis to break it.

Day 1: measure, change nothing

Open the screen-time report that's already on your phone (Screen Time on iPhone, Digital Wellbeing on Android). Write down three things: the app that eats the most time, the total, and the two times of day it happens. Don't change anything yet. You're taking the "before" photo, and most people find the number alone is motivating enough to carry the rest of the week.

Be specific about the two time slots. "Evenings" is useless. "10:40pm in bed" and "the first ten minutes after I sit down at 6pm" are things you can design around.

Day 2: add friction

Log out of the worst app. Not delete — log out. Move it off your home screen into a folder on the last page. Set the built-in app limit to twenty minutes.

Now the honest part: you will override the limit. Everyone does. The limit isn't a wall, it's a speed bump, and the point of a speed bump is to make you notice you're driving. The log-out does the same job: three seconds of typing a password is enough to ask "do I actually want this?" Surprisingly often, the answer is no.

Day 3: move the phone out of the bedroom

If one of your two slots is in bed, this is the single highest-value change in the week, and it costs about eight pounds.

Buy an alarm clock. Charge the phone in the kitchen or the hall. The NHS's sleep advice is to keep phones and screens away for an hour before bed; you don't have to hit an hour on day three. Just make the phone not be the last thing you touch and the first thing you reach for. The 2am scroll can't happen if the phone is in another room, and that's the whole trick. You're not being stronger. You're removing the option.

Day 4: replace, don't just remove

This is where "just stop" plans die. You removed the thing you did at 6pm on the sofa and left a hole, and holes get filled by the nearest habit, which is the phone.

Pick a replacement that is finishable. That's the key property. A paper book has a chapter end. A podcast episode ends. A newspaper app with an actual edition — a front page that's the same at 6pm as at 7pm — ends. A crossword ends. The feed doesn't, and that's why you can't put it down. Leave the replacement physically where the phone used to be: the book on the sofa arm, the puzzle on the kitchen table.

Day 5: two news windows and one question

If part of the pull is "I need to know what's happening", honour that properly instead of pretending you'll go cold turkey on the news. Give it two windows a day, say 12:30 and 6pm, ten minutes each, and read a source with an edition rather than a feed. You'll be better informed than the scroller, not worse, because you'll read whole articles instead of forty headlines.

For the anxious moments in between, when the thumb is already moving, use one question: "Is there something I can do about this today?" If yes, do that thing. If no, close it. Most of what the feed shows you fails the test. That isn't callousness. Being upset about something you can't act on, at 11pm, helps nobody, including the people the story is about.

The trade-off you're accepting is real and small: you'll hear about breaking news an hour or two late. Think about the last time that cost you anything.

Day 6: curate the feed you're keeping

You kept the apps, so make them less bad. Spend fifteen minutes doing the thing you've been meaning to do for two years. Mute the keywords that reliably spike you. Unfollow or mute the five accounts that leave you feeling worst — not the ones you disagree with, the ones that make the evening worse. Turn off every notification except messages from actual people.

You'll be struck by how reluctant you are to unfollow things that make you miserable. That reluctance is the design working. Do it anyway.

Day 7: look at the number

Open the screen-time report again. Compare it to day one. Then keep what worked and drop what didn't, without guilt about either. For most people the alarm clock and the finishable replacement are the keepers; the app limit gets ignored and the two news windows stick about half the time. That's fine. A plan that survives at sixty percent beats a detox that survives at zero.

What this isn't

It isn't a cure for anything. If you're scrolling because you can't face what happens when you stop — if the quiet is the problem, not the phone — a week of friction will show you that, which is useful information, but it's not a fix. The fix for that is a conversation with a GP or someone you trust, and the phone was only ever the symptom.

For everyone else, which is most of us, the loop is mostly environmental, and environments can be changed in a week. The phone goes in the kitchen. The book goes on the sofa. The news gets two windows. And the thing you were going to check "for a second" waits until a time you chose.

Related: How to Quiet an Overthinking Mind at 2am.