The Right Numbers: How to Set Your Usage Limit and Break Time So They Actually Stick

Almost everyone who installs a screen-time tool leaves the default settings exactly where they found them. It’s a reasonable instinct — someone presumably thought about the number before shipping it — but a single default can’t actually be right for a remote worker doing focused research on YouTube, a student who checks Reddit for ninety seconds between problem sets, and someone doomscrolling X in bed at midnight. Those are three different problems wearing the same 30-minutes-then-a-pet-video shape, and the version of the settings that works for one of them will actively annoy or completely fail to register for the other two.

The two numbers you’re actually setting — usage limit and break time — do different jobs, and most of the frustration people report (“it never comes on” or “I just Shoo! it every time”) traces back to one of the two being miscalibrated, not to the underlying idea being wrong.

What each number is actually controlling

Usage limit is a budget, not a warning. It’s the number of minutes of focused, visible time on a tracked site before the interruption fires — time only counts while the tab actually has focus and isn’t hidden behind another window or a different tab, so background research tabs and things you forgot were open don’t quietly burn through it. Each tracked domain gets its own independent budget: a heavy YouTube habit and a light Reddit habit don’t share a pool, so fixing one doesn’t require touching the other.

Break time is a full stop, not a suggestion. It’s how long the pet video occupies the tab before you get control back — scrolling, clicking through the page, and touch gestures are blocked for the duration, which is the entire mechanism. A break that’s too short doesn’t give the moment enough weight to actually register as an interruption; one that’s too long starts to feel punitive, which is exactly the kind of friction that gets a tool uninstalled rather than respected.

These two numbers interact. A tight limit paired with a long break reads as a punishment schedule. A generous limit paired with a token break barely registers at all. The combination that tends to work is a limit that’s honestly a little shorter than how long you actually want to be there, and a break that’s long enough to genuinely reset your attention without turning into a chore.

Starting points by situation, not by guesswork

None of the numbers below are the “correct” ones — there isn’t one. They’re reasonable starting points for a few common situations, meant to be adjusted after a week of actually living with them.

Remote workers using YouTube or X for research that drifts. Start closer to 45–60 minutes with a 3-minute break. The goal here usually isn’t to stop the browsing — it’s to catch the moment focused research quietly becomes something else, without interrupting so often that it breaks real work. A short break is enough to ask “was I still doing the thing I opened this for?”
Students studying near a phone or a second monitor. A shorter limit, 15–20 minutes, with a break around 5 minutes, tends to match how quickly a “quick check” during a study session actually happens — the interruption should land before the drift, not after 40 minutes of it. If this feels aggressive at first, that’s often a sign it’s doing its job; it can always be loosened once the habit of checking in has actually changed.
Late-night scrollers. Time-of-day isn’t something the extension can key off directly, but a shorter limit in the evening — 15 minutes rather than 30 — paired with a longer break, 8–10 minutes, works well as a manually-adjusted “evening mode”: long enough to make picking the phone back up feel like a real decision rather than a reflex, at an hour when reflexes are exactly what’s running the show.
Anyone who keeps hitting Shoo! immediately. This is the clearest signal in the whole system: if the break ends the moment it starts, the limit is set to a number you don’t actually agree with. That’s not a discipline failure — it’s a miscalibrated setting, and the fix is to raise the limit to something you’d actually defend, not to feel bad about clicking Shoo!.

The tell that you’ve got it wrong

Two failure modes show up more than any others, and both have the same fix: adjust the number, don’t abandon the tool.

It never comes on. If a week goes by and you never see your pet, the limit is probably too generous for the site in question, or the site itself isn’t the one actually eating your time — check whether the real drift is happening somewhere that isn’t on your tracked list yet.

You Shoo! it every single time. If the break never survives past the first few seconds, the limit was set somewhere you don’t actually believe in, and the tool has quietly become theater instead of friction. Raising the limit by 50% and trying again for a few days is a better move than lowering your expectations of what the setting should do.

Both of these are configuration problems, not evidence that “gentle interruption” as a category doesn’t work for you. We’ve written before about why friction succeeds where hard blocking fails — but friction set at the wrong dial is just as easy to route around as no friction at all.

A five-minute calibration pass

Pick one site you actually lose time to, not every site you’ve ever visited. A budget spread across ten domains you rarely open is noise; a budget on the two or three that actually eat your afternoon is signal.
Set the limit to a number slightly below your honest guess. If you think you spend 40 minutes, try 30. The interruption should catch the drift a little earlier than feels comfortable, not exactly where you’d have stopped anyway.
Give the break enough room to be a real pause — under 3 minutes rarely gets past “annoying flicker” into “actual reset,” per the same logic behind why longer, open-ended breaks work better than a five-second glance away.
Watch what you actually do for a week, not what you predict you’ll do. The Shoo! button is data: if you’re using it constantly, the limit’s wrong. If you never see the pet, the site or the limit is wrong in the other direction.
Revisit the numbers monthly, not once. Habits that respond to a 20-minute limit in week one often tolerate 30 by week four — the point isn’t to lock in forever, it’s to keep the friction sized to the habit as the habit actually changes.

The category of tool this sits in — a gentle interrupter rather than a hard blocker — only works if the numbers behind it reflect something you’d actually stand behind at 2 p.m. on a Tuesday. The default is a starting guess, not a verdict. The version worth keeping is the one you’ve actually adjusted to fit the way you really browse.