What Does a 5-Second Pause Actually Do? The Research Behind Pausing

Most advice about phone use comes down to taking something away. Block the app. Set a limit. Delete it and start again on Monday.

This article is about a much smaller intervention: putting a few seconds between wanting to open an app and actually opening it.

Not a lecture.

Not a block.

Not a 30-minute limit.

Just a pause.

So much app opening is not really a decision in the first place — that is how automatic app checking becomes a habit, and it has its own article. What I want here is narrower: what does the research say happens when you interrupt the sequence?

The answer is more interesting than “five seconds gives you willpower.” A short delay can change the structure of the decision. But the evidence also says something the marketing tends to skip: there is nothing magical about five.

The useful idea is the pause.

It started after the worst thing that had happened to me. My fiancée ended our engagement.

I went to Italy for the first time. To forget, to recharge, to notice small things again. That was the entire point of the trip.

And I spent it on my phone.

Not deciding to. My hand would go to my pocket, and Instagram would be open, and I had no memory of choosing either.

I tried the obvious things. Screen time limits, which I dismissed without reading. Then apps built for this, which either shamed me or did nothing at all.

I’m a software developer, so I went home and started building the one I pictured, the one that could actually help me and potentially a lot more people like me.

A pause changes the decision, not just the clock

When an app opens instantly, the decision is almost frictionless. The cue is there. The action is available. The reward is one tap away.

Adding a delay changes the sequence. Instead of:

cue → tap → app → content

you get:

cue → tap → pause → decision → app or exit

That extra step matters because it puts a small separation between the impulse and the behaviour.

This is related to self-control, but it helps to stop thinking of self-control as “trying harder.” Duckworth, Gendler and Gross (2016) argue that some of the most effective self-control is situational rather than effortful: choosing or changing the situation so the impulse is weaker, instead of fighting the impulse once it has arrived. Their examples are ordinary — the partygoer who sits away from where the drinks are poured, the student who goes to the library without a phone.

A pause is a very small form of situation modification. It does not remove the choice. It changes the conditions under which the choice is made.

The PNAS study that matters

The most directly relevant study is Grüning, Riedel and Lorenz-Spreen (2023), published in the Proceedings of the National Academy of Sciences. They studied one sec, a self-nudge app that interrupts an app opening with a deliberation message, a brief wait, and an option to abandon the attempt.

The field study followed 280 users over six weeks, and the numbers are striking. On average, participants abandoned the target app on 36% of attempts after the intervention appeared. Over six weeks, attempts to open target apps fell by 37% compared with the first week. Combining the two, actual openings of target apps fell by 57%.

Two things belong next to that 57%, and most write-ups leave them out.

First, the six-week study had no control group. It followed people who chose to install the app, so some of that change is the tool and some is the kind of person who installs one. The authors do not hide this; it is a limit on how much weight the number carries.

Second, one of the authors is one sec’s developer. The paper discloses this openly, to their credit — and it is still a fact you deserve when a headline number comes from a study of the author’s own product.

The stronger part of the paper is the preregistered online experiment, with 500 participants, which tested the three components separately. The option to dismiss the attempt had the strongest effect. The waiting-time friction also reduced consumption. The deliberation message, on its own, did not work.

That finding cuts against the simple version of this idea. It is not that any interruption helps. Being told to reconsider did nothing. Being given a moment and a way out did.

So, the lesson is not “wait five seconds and your screen time falls by 57%.” It is narrower and more defensible: a small interruption can turn an automatic opening into a moment where another choice is available.

So what does five seconds actually do?

Five seconds is long enough to notice and short enough not to feel like a punishment. That makes it a reasonable design interval, not a discovered constant.

It does not need to produce a transformation. It only needs to interrupt the sequence long enough to register what is happening. The difference between:

“I opened Instagram.”

and:

“I was about to open Instagram. Do I actually want to?”

is awareness, and awareness is the whole product.

Stillman and colleagues (2025), also in PNAS, give a broader view of how that moment unfolds. Using mouse-tracking to watch choices evolve in real time, they found evidence for both impulse inhibition and a more dynamic competition between immediate temptation and longer-term goals — rather than for inhibition alone.

In other words, the decision may still be forming. A pause gives it room to finish before the behaviour does.

The pause is not the same thing as willpower

This distinction matters, because “use more willpower” is a poor design strategy. Willpower asks you to notice the temptation, resist it, and keep resisting it.

A well-designed pause asks less. It says:

Wait.

That is different from:

Resist.

Raio and Glimcher (2021) put a price on the difference. Across five studies, they measured what people would pay in real money to avoid situations that required self-control — and the more tempting the item, the more they paid. Resisting is not psychologically free, and people know it well enough to buy their way out of it.

Milyavskaya and colleagues (2015) point the same direction from another angle: people who succeed at their goals tend to experience fewer temptations in the first place, rather than winning more fights against them.

Both suggest the same design conclusion. If a tool can make an unwanted action slightly less automatic, it may reduce how much effort is needed at all. The goal is not to make you stronger every time. It is to make the action less automatic.

Five seconds is not a magic number

This is where digital-wellness writing usually becomes too confident. There is good evidence that delays and interruptions change behaviour. There is no evidence that five seconds is a threshold at which impulse disappears.

The component results above are the clearest argument against a magic number. If friction alone were the mechanism, the deliberation message — which also costs time and attention — would have worked. It did not. What worked was the exit.

Friction fails in two directions. Too little and it is ignored. Too much and people disable the tool, at which point its effect is zero.

So, the useful design question is not “why exactly five seconds?” It is:

How much friction is enough to interrupt an automatic behaviour without making the tool itself annoying?

For anyone building one of these, that is a far better question.

The strongest evidence is about interruption

The best case for a pause does not rest on one experiment about phones. It rests on several lines of work pointing at the same principle: change the timing of a response and you change the response.

McCurry, May and Donaldson (2024), in Communications Psychology, had couples compete in a reaction-time game where the winner could blast noise into the loser’s headphones. Short enforced breaks — including breaks of five seconds — reduced negative emotion and aggressive behaviour. The primary sample was 104 participants across 3,095 trials, with a replication of 58 participants across 3,167 trials.

That study was about conflict, not smartphones. It is not evidence that a five-second pause reduces social-media use, and I am not going to present it as if it were.

What it shows is the same shape: interrupt an immediate response and the sequence can change. The mechanism should not be stretched further than it goes.

Why this matters for smartphone habits

Phone checking is a good candidate for this kind of intervention because so much of it is habitual. Oulasvirta and colleagues (2012) described the “checking habit” — brief, repetitive inspections of dynamic content — as a learned association between a cue and a response, built by repetition rather than chosen each time.

It is worth keeping habitual and compulsive apart. Opening an app automatically is not a disorder — it is what repetition does to any behaviour.

The objective is not to convince you your phone is bad. It is to create one moment where you can tell the difference between:

“I want to open this.”

and:

“I automatically opened this.”

That distinction is small. Repeated a few hundred times, small distinctions add up.

Add a moment, not a wall

The tempting fix is always a stronger restriction.

Those work for some people, and they also turn the tool into an opponent. A pause takes a different line. It does not say:

No.

It says:

Not yet.

A block removes the behaviour from the set of things you can choose. A pause keeps the choice and makes it more deliberate. The one sec components suggest the exit does more work than the timer — an argument for keeping a way through.

What the research does and does not say

The evidence supports a few modest conclusions. A short interruption can add friction to an automatic behaviour. Self-control can be supported by changing the situation rather than by effort, and a pause creates room to reconsider before the action completes.

What it does not establish is that five seconds is optimal, that everyone responds the same way, or that a timer alone fixes anything.

That is exactly why the pause is worth building. It does not need to be a miracle. It only needs to make a little space between the impulse and the action.

Sometimes that space is enough to notice you never really wanted to open the app.