Why Friction Beats Blocking: Screen-Time Limits That Actually Last
Why friction beats blocking for screen time: the University of Michigan study, the habit loop it exploits, and how to add friction on Android and iOS that you keep.
Quick answer
Friction beats blocking because a small, repeatable cost at the moment of reaching changes behaviour without triggering the rebellion that a hard wall does. In a five-week University of Michigan field study (CHI 2024), making the phone slightly more annoying to use cut screen time about 16% more than timed lockouts, and 62% of participants kept the friction on versus 36% for lockouts. Timers get ignored because the 'Ignore limit' button is always there; friction is paid every time, so it keeps working.
Key takeaways
- University of Michigan's InteractOut study: friction cut screen time ~16% more than lockouts and was kept on by 62% vs 36% (CHI 2024).
- Blocking fails at the moment of highest craving, which is exactly when the 'Ignore limit' button gets pressed.
- Friction works on the habit loop: it sits between cue and routine, and the cost is paid on every repetition, not once.
- Good friction is small (10–30 seconds), unavoidable and ideally useful — a question to answer beats a countdown to wait out.
- On Android and iOS you can add friction today with grayscale, home-screen distance, notification pruning and roll-to-scroll style guards.
You set a 30-minute limit on Instagram. On day one it fired and you closed the app, slightly proud. On day three you tapped “Ignore limit for 15 minutes”. By day ten you tapped “Ignore limit for today” without reading it. The limit still exists in your settings. It just does not exist in your life.
This is not a failure of discipline. It is a predictable property of walls. This post is about the alternative, friction, why it works where blocking does not, and how to set it up on the phone you already own.
What did the University of Michigan study actually test?
The clearest head-to-head comes from a field study presented at CHI 2024, the main academic conference for human-computer interaction. University of Michigan researchers built a tool called InteractOut that did not block anything. It just made the phone slightly more annoying to use: taps registered with a short delay, swipe gestures were shifted, scrolling took a little more effort.
They compared it with the standard approach, a timed lockout of the kind Screen Time and Digital Wellbeing provide. Forty-two people used the tools for five weeks. The results (University of Michigan, 2024):
| Measure | Friction (InteractOut) | Timed lockout |
|---|---|---|
| Reduction in screen time | ~16% greater than lockout | Baseline |
| Participants who kept it switched on | 62% | 36% |
| Study length | 5 weeks | 5 weeks |
| Participants | 42 | 42 |
Two things stand out. The friction cut more screen time, which is the headline. But the second row is the one that matters for anyone who has lived through the “Ignore limit” cycle: people kept the friction on at nearly twice the rate they kept the lockout.
A tool you keep using at 62% beats a tool you abandon at 64%, even if the abandoned one worked better on paper for the week you tolerated it.
It is a small study, and forty-two people over five weeks is not the last word. But it matches what every person who has set and then ignored an app timer already knows.
Why does blocking fail?
It fires at the worst moment
A limit triggers when you have used the app for 30 minutes. That is, by definition, a moment when you are engaged and want more. The wall appears at peak craving and asks you to make a sober decision. It is the equivalent of putting the biscuit tin lid on after you have eaten half the packet.
The door is always there
Both Apple and Google ship an override, and they must. A limit with no override is a parental control, and adults uninstall parental controls. So every lockout is a wall with a door, and once you have walked through the door a few times, the wall is decoration.
It is all or nothing
Blocking offers two states: allowed and forbidden. Real life has a third, which is “I could, but is it worth it right now?” That third state is where behaviour change happens, and blocking skips it entirely.
Deleting the app moves the scroll
The harder version of blocking, deleting the app, fails for a different reason. The habit is not “open Instagram”. It is reach, unlock, swipe. YouTube Shorts, Reddit, the news app and the browser all satisfy it. How to Stop Doomscrolling covers why replacement matters more than removal.
How friction works on the habit loop
A habit has three parts: a cue (boredom, a notification, a dead minute), a routine (reach, unlock, swipe), and a reward (novelty, a hit of something). The loop runs automatically, which is the whole problem: by the time you notice, you are already scrolling.
Friction inserts itself between the cue and the routine. Grayscale makes the reward duller. A delay makes the routine slower. A question makes the routine conscious. None of them forbids anything. Each one converts an automatic sequence into a tiny decision, and most of the time, when you are forced to decide, you decide not to bother.
That is why the cost must be paid every time. A lockout is paid once, when you hit the quota, and you can pre-decide to override it. Friction is paid on every single reach, so there is no pre-decision to make. Each reach is its own small vote.
What makes friction good rather than merely irritating
- Small. Ten to thirty seconds. Enough to break automaticity, not enough to resent.
- Unavoidable. No “skip” button, or the skip becomes the habit.
- Immediate. At the moment of reaching, not at a quota you hit later.
- Ideally useful. A plain delay is a cost you pay for nothing. A question you answer is a cost that pays you back, and people tolerate it for much longer.
That last point is the design idea behind “roll to scroll”. More on it below.
How to add friction on Android and iOS today
None of this needs a new app. Start with the built-ins.
Both platforms
Grayscale. On iOS: Settings > Accessibility > Display & Text Size > Colour Filters > Grayscale, then add it to the Accessibility Shortcut so a triple-click toggles it. On Android: Settings > Accessibility > Colour correction (or Digital Wellbeing’s Bedtime mode, which includes grayscale on many phones). Feeds are designed in colour. In grey they look like a spreadsheet of strangers.
Distance. Move feed apps off the first home screen, into a folder, on the last page. Make them a search away rather than a tap away. Twenty seconds of extra effort is often enough to tip the decision.
Notification pruning. Turn off every notification that is not from a human. Each app notification is a cue, and Common Sense Media found over half of teens receive 237 or more a day (Common Sense Media, 2023). Fewer cues, fewer loops.
iOS specifics
Use Focus modes to hide feed apps from the home screen entirely during work or sleep hours, rather than blocking them. The app still exists; it is just not in front of you. Set a one-minute Screen Time limit on your worst app and treat it as a pause, not a wall: when it fires, the point is to notice you reached, not to stop you.
Android specifics
Digital Wellbeing’s Focus mode greys out chosen apps and pauses them with a tap to resume, which is closer to friction than to blocking. Bedtime mode turns the screen grey on a schedule. And because Android allows apps to sit in front of other apps, it supports a kind of friction iOS does not yet: a guard that asks something of you before another app opens.
Roll to scroll: friction that pays you back
Wiseroll is the worked example here because it was built around the friction model rather than the blocking model. On Android, you choose which apps to guard: Instagram, TikTok, YouTube, anything. When you open one, Wiseroll asks you to solve one “roll” first, a 30-second multiple-choice, numeric or photo question drawn from channels you follow, or a short clip you have to lock in an answer to. Get it right and the app opens. That is about 20 seconds.
Nothing is locked. There is no quota and no “ignore” button because there is nothing to ignore; you just answer the question. The cost is paid every time, which is what the Michigan study suggests you want, and it is paid in something useful, which is why people keep it on. If you want to compare it with the built-in timers directly, Wiseroll vs Screen Time Limits does that.
The honest limits: roll to scroll is Android-first, because iOS does not let one app intercept another. And Wiseroll is coming to Android and iOS via a waitlist, so for now the iOS version of this idea is the grayscale-plus-distance-plus-pause stack above.
A friction stack that survives a month
| Layer | What | Cost per reach | Keeps working because |
|---|---|---|---|
| Visual | Grayscale on by default | Nothing | The reward is duller every time |
| Spatial | Feed apps on last page, no dock | ~10 s | The reach is conscious |
| Cue | App notifications off | Nothing | Fewer loops start |
| Gate | Roll to scroll or a one-minute pause | ~20 s | Each reach is a vote, paid in something useful |
Four layers, all small, none of them a wall. The point is not to make the phone unusable. It is to make the autopilot slice, the 36% of UK phone use that happens with no clear purpose (Virgin Media O2, 2026), require a decision. Decisions are what autopilot cannot survive.
Where to start
Turn on grayscale tonight and move your two worst apps to the last home screen. That is ten minutes of setup and two layers of friction with zero ongoing cost. Live with it for a week, then add a gate: roll to scroll if you are on Android, or a one-minute Screen Time pause you honour as a moment to notice rather than a wall to climb. Keep whatever you still have on after a month. The Michigan numbers say that is the one that counts.
Frequently asked questions
Why don't screen time limits work?
Because they are walls with a door in them. Apple's Screen Time and Android's Digital Wellbeing both have to offer an override, otherwise users would uninstall the feature entirely. The limit fires at the moment you most want the app, you press 'Ignore limit', and within a week the limit is a formality. A University of Michigan study found only 36% of participants kept a timed lockout on after five weeks.
What is the difference between friction and blocking?
Blocking stops you using an app, either for a period or once a quota is hit. Friction lets you use it but adds a small cost each time: a delay, a question, a greyed-out screen, an extra few taps. Blocking produces an all-or-nothing choice that people override. Friction produces a small, repeated moment of decision that tilts behaviour without provoking the urge to switch it off.
What did the University of Michigan screen time study find?
Researchers compared a timed lockout with InteractOut, a tool that made the phone slightly harder to use by delaying or shifting touch gestures. Over five weeks with 42 participants, InteractOut reduced screen time about 16% more than the lockout, and 62% of people kept it on compared with 36% for the lockout. The paper was presented at CHI 2024.
How much friction is enough to change a habit?
Less than most people think. The useful range is about 10 to 30 seconds of unavoidable effort at the moment of reaching: enough to break the automatic reach-unlock-swipe sequence and force a decision, not so much that you resent it. If the friction is also useful, such as answering a question, people tolerate it far longer than a plain delay.
Is roll-to-scroll a blocker?
No. Roll to scroll, the Wiseroll feature for Android, guards apps you choose and asks for one solved roll, a roughly 20-second question or puzzle, before the app opens. It never locks you out and has no quota to hit. It adds a small, useful cost to every reach, which is the friction model rather than the blocking model.
Swap the scroll, keep the swipe.
Wiseroll is a free feed of 30-second quiz rolls, puzzles and knowledge clips — with a roll-to-scroll guard for Instagram and TikTok. Coming to Android and iOS.