If a fixed nighttime boundary fits your routine, you can download 9to9 from the App Store and choose the apps and websites you want quiet from 9 PM to 9 AM.
There is a version of staying up too late that has nothing to do with insomnia and nothing to do with losing track of time. You know exactly what time it is. You know the alarm is set for 6:30. You know, with the quiet certainty of experience, that tomorrow-you will pay for tonight-you's decisions. And you stay up anyway, because these dark, undemanding hours are the first ones all day that belong entirely to you.
If that description stings a little, you are describing what researchers call bedtime procrastination — and what the internet, with its usual flair, renamed revenge bedtime procrastination. The “revenge” is aimed at your own schedule: after a day spent meeting other people's demands, stealing hours back from sleep can feel like the only available form of protest. Understanding that motive matters, because it explains why the obvious fixes — earlier bedtimes, sterner rules, more guilt — so often make things worse.
What the term actually means
Researchers introduced bedtime procrastination as a specific pattern: going to bed later than intended, when no external circumstance prevents it, while fully expecting a cost the next day. Each element matters. It is not staying up late for a reason — a crying baby, a deadline, a red-eye flight. It is delay without a justification, chosen anyway, with the consequences visible in advance.
That definition came out of a 2014 paper by Floor Kroese and colleagues, who noticed that procrastination research had never really looked at bedtime, even though “I'll go in ten minutes” is one of the most commonly broken promises people make to themselves. Later work gave the pattern a more vivid name and a sharper motive. A 2023 paper in Sleep Health called the phone “the thief of (bed)time,” documenting how social media in particular turns a vague intention to sleep into a prolonged negotiation. The revenge framing adds what the clinical definition leaves out: the delay is often deliberate, and it often feels earned.
One clarification worth keeping: bedtime procrastination is not insomnia. Someone procrastinating bedtime is postponing the attempt to sleep; someone with insomnia may be lying there trying and failing. They can look identical from the outside — a person awake at 1 AM — but they are different problems with different responses. If your nights involve genuine inability to sleep, persistent distress, or daytime impairment that will not lift, that deserves a conversation with a professional rather than a productivity fix.
There is also a tell that separates this pattern from ordinary night-owl preference: the morning-after mismatch. A true night owl stays up late and feels fine about it. Bedtime procrastination comes with a signature regret — you did not even enjoy the extra hours that much, and you would not have chosen them at the price they cost. That gap between the choice and the wanting is the whole diagnosis in miniature.
Sources: Bedtime procrastination: introducing a new area of procrastination (Frontiers in Psychology), The thief of (bed)time (Sleep Health, via PubMed)
Why shame is such a poor strategy
The default response to revenge bedtime procrastination, in most advice and in most people's own heads, is some flavor of discipline: be more responsible, respect your bedtime, act like an adult. It fails for an interesting reason. The scolding frame treats the late-night hours as pure waste — a bad habit to be suppressed. But the person staying up does not experience those hours as waste. They experience them as the only leisure the day produced.
Think about who revenge bedtime procrastination hits hardest. It clusters around people whose days are most claimed by others: demanding jobs, caregiving, long commutes, schedules with no slack in them. When your waking hours are spoken for from 7 AM to 10 PM, the stretch between “finally done” and “must sleep” becomes precious out of proportion to its size. Squandering it on a feed is a poor trade — but choosing to sleep instead can feel like surrendering the whole day. The scroll is not the point. The autonomy is. The feed is just the cheapest, lowest-effort form autonomy comes in.
This is why “just go to bed” backfires. It asks you to give up the only free hour in exchange for a less miserable tomorrow — a trade that loses, reliably, at the exact moment of night when tomorrow feels abstract and freedom feels concrete. A plan that works has to keep some of what the night is for.
What the late hours are actually buying
It helps to be precise about what the night is purchasing, because the purchase is rarely “content.” Look closely and the late-night scroll is usually buying one of three things. Sometimes it is decompression — a way to metabolize a demanding day before sleep, to let the static of other people's drama rinse out your own. Sometimes it is solitude — the only stretch where nobody needs anything from you, where the social surface of the day finally goes still. And sometimes it is simply postponement — tomorrow exists, and as long as you stay awake, it has not started yet.
Each of those is a legitimate need wearing an inefficient solution. Decompression does not require a feed; a shower, a walk around the block, or twenty minutes of a show you actually chose does the same job without the two-hour tail. Solitude does not require the internet; a quiet room and a book are more solitary than a feed full of strangers. Postponement is the trickiest, because no rearrangement of the evening makes tomorrow arrive later — but naming it honestly at least lets you ask whether the day itself needs to change, rather than its ending.
The reason this inventory matters is that it turns “I can't stop staying up” into a more workable question: which need is going unmet, and what would meet it in a form that can actually conclude?
Protect a real leisure window instead of stealing one
The counterintuitive move is to make the leisure official. Not less of it — less sneaky, less infinite, less accidental.
Choose a window before your digital closing time — twenty to forty-five minutes is a reasonable size — and designate it as yours in advance. This is the part where you watch the episode, play the game, read the thing, call the person. The specifics matter less than three properties: it is chosen beforehand rather than delegated to a feed's judgment; it has an ending built in — one episode, one chapter, one call — so it can actually conclude; and it happens while you still have some capacity to enjoy it, rather than in the dregs of the night when you are too tired to do anything but scroll.
There is a real difference between forty minutes of something you picked and two hours of whatever the algorithm served. The first leaves a memory; the second leaves a fog. People who build a protected window often find the strange result that less leisure time feels like more, because they were actually present for it.
Give the end of the night a shape
The other half of the problem is the cliff edge. Most failed evenings do not fail at 9 PM — they fail at midnight, when “one more” has been said eleven times and stopping feels arbitrary. What helps is making the ending small, predictable, and physically easy.
Keep the shutdown routine minimal. A long, optimized bedtime routine is itself a form of resistance — one more demanding thing between you and sleep, easy to postpone. Try a minimum viable version instead: the phone plugs in across the room (or in another room entirely), the basic hygiene happens, and one quiet, finite activity carries you to the bed. Ten pages. A short stretch. A conversation that ends when it ends. You can always expand later; the first job is making “stop” repeatable on a night when nothing went right.
Environment does quiet work here. A charger that lives outside arm's reach removes the midnight scroll without requiring a decision at midnight. A book on the pillow makes the replacement the path of least resistance. These sound trivial because they are — that is the point. You want the right action to be the lazy one, because at 11 PM lazy is all you have.
One boundary, seven nights
Rather than redesigning your whole relationship with your phone, run a single experiment: pick one stopping rule and keep it identical for a week. “Entertainment feeds end at 9 PM” is a good default — memorable, early enough to matter, late enough to allow a real evening.
For seven nights, let the rule be boring. Each morning, note two things and nothing more: roughly when you stopped scrolling, and whether you would repeat the rule. That tiny record does two jobs — it shows you whether anything actually changed, and it keeps the experiment honest on the nights your memory would rather forget.
What people usually notice, somewhere around night four, is that the resistance concentrates at the boundary itself — the half hour around closing time — and then dissipates. Once the feed is genuinely unavailable, the evening does not feel deprived so much as open-ended, and the protected window you built earlier starts doing the work the scroll was pretending to do. That discovery is the real payload of the week: not that you can survive without the feed, but that you barely miss it once it is truly off the table.
When the rule fails — and it will, at least once — resist the urge to escalate. Do not add more blocked apps, an earlier cutoff, and a punishment clause in one panicked revision. Make the rule smaller before you make it harsher: block only the two apps that actually trap you, move the boundary to a time you can credibly keep, or fix the replacement activity that turned out to be no fun. A boundary survives by being livable, not by being strict.
When the problem is bigger than the phone
A caveat that deserves its own space, because kindness cuts both ways. Sometimes the late-night pattern is a symptom of something the evening cannot fix: a workload that genuinely never ends, a relationship or living situation with no room to breathe, anxiety that treats 2 AM as a strategy meeting. If you protect a leisure window and it does not help — if the need behind the scrolling is bigger than entertainment — that is not a failed experiment. It is information about where the real work is.
The same goes for sleep itself. If you regularly cannot fall asleep or stay asleep even when you put the phone down, or if daytime exhaustion is costing you work, driving, or relationships, talk with a qualified professional. No routine and no app is a substitute for that conversation, and treating them as one only delays it.
The part a tool can do, and the part it cannot
This is where a scheduled blocker earns its place — not as a substitute for the leisure window, but as its bodyguard. When 9 PM arrives and the feeds simply do not open, the negotiation you would otherwise lose is never held. Something like 9to9, which shields your chosen apps from 9 PM to 9 AM every day, exists precisely for the person who understands all of this and still finds their hand drifting toward the phone at midnight. The blocker holds the line; the protected hour before it is where the “revenge” actually gets satisfied — on purpose, in a form you will remember.
What no tool can do is decide what your evenings are for. If the day leaves no room for you at all, the fix is bigger than an app — it is a conversation about where the hours go. But for the common version of this problem, the fix is smaller and kinder than it looks from the outside: a little real freedom, earlier, with a door that closes when it is over.