I wrote an if-then, but in the moment I ignore it.

That's expected. The meta-analysis reports a medium-to-large effect, but on its own it's weak against automatic action during a strong urge. Pair it with environment design that adds friction before you open the app, and the if-then gets room to work.

How many should I make at once?

Start with one. Confirm the link between cue and action in a single scene before adding more, so you can tell which ones work. When you add, keep the conditions — a concrete cue, an action under five seconds.

Can you stop the "suddenly it's open" reflex? How to build an if-then plan

About 2 min read Last verified: 2026-07-22 Published: 2026-07-22 By the Detroppo editorial team

In short

"Suddenly it's open" isn't weak will; it's an automated action that never passes through a decision.
If-then planning — "if X happens, I do Y" — was reported at an effect size of d=0.65 (medium-to-large) in a meta-analysis of 94 tests.
But it's weak on its own against strong urges, so it assumes you pair it with environment design. Three ready examples for night phone use.

1 primary source Last verified 2026-07-22 See sources

This is general wellness information, not medical advice. It describes what research has reported; it is not a diagnosis or treatment, and individual results vary. If you're struggling, please reach out to a professional or a crisis line (see the footer).

You don’t remember deciding to open it, and your thumb is already at the top of the timeline. Sound familiar?

The trickiest part of “suddenly it’s open” is that there’s no decision at the moment you open it.

Is “suddenly it’s open” weak will?

No. It’s an automated action that skips the decision step.

From unlocking to launching the app, your finger’s path is automated by countless repetitions. An automated action never passes the “should I open this” moment. So a resolution — “I won’t open it tonight” — never gets its turn, and the night ends. The resolve isn’t weak; it just isn’t placed where it can act.

What is if-then planning? Does it work?

It’s deciding in advance “if X happens, I do Y,” and a meta-analysis backs it.

A meta-analysis pooling 94 tests (Gollwitzer & Sheeran 2006) reported an effect size of d=0.65 (medium-to-large) for if-then planning toward goal attainment. The mechanism is simple: by fixing a cue and an action together ahead of time, you fire the wanted behavior without needing an in-the-moment decision. You answer automation with automation.

But to be honest: on its own, if-then is weak against the automatic action of a strong urge in full swing. Don’t count on one sentence to win head-on in the thick of a midnight craving. It assumes you pair it with environment design — rearranging apps, changing where you charge. Think of if-then not as the last line, but as one layer laid on top of environment design.

How do you write a night-phone if-then?

Write it on two conditions: a concrete cue, an action under five seconds.

Here are three examples. Use them as is if you like.

If my hand reaches for the phone in bed, I turn the screen face-down and slide it under the pillow.
If the 11 p.m. alarm sounds, I plug the phone into the charger outside the room.
If a video ends and the next is about to start, I go back to the home screen.

A vague line like “go easy on the phone at night” doesn’t work as an if-then. Trim the cue down to a visible event, and the action down to something the body can do without thinking.

Tonight’s step: write just one line

Pick one example above, or write one line for your own “moment it falls apart” (30 seconds).
Put that line at the top of a phone note, or on paper by the bed.
Tomorrow night, try just that one scene. Leave the other scenes unchanged for now.

Common questions

I wrote an if-then, but in the moment I ignore it.: That's expected. The meta-analysis reports a medium-to-large effect, but on its own it's weak against automatic action during a strong urge. Pair it with environment design that adds friction before you open the app, and the if-then gets room to work.
How many should I make at once?: Start with one. Confirm the link between cue and action in a single scene before adding more, so you can tell which ones work. When you add, keep the conditions — a concrete cue, an action under five seconds.

Sources

Primary sources only (peer-reviewed research and public bodies). Last verified: 2026-07-22.

Gollwitzer & Sheeran (2006) Implementation intentions meta-analysis (94 tests). Advances in Experimental Social Psychology If-then planning (implementation intentions) reached an effect size of d=0.65 (medium-to-large) for goal attainment; weak on its own against strong in-the-moment urges.