How to write an if-then plan for stopping something

Updated 2 September 2026

The short answer

Write it as a situation followed by an action: when I sit down after dinner, I make tea before I sit down. The situation half has to be something you will physically notice happening. The action half has to be something you could begin within ten seconds of noticing it, where you will actually be standing.

The reason for the shape is that the plan has to be usable at a moment when you are not deliberating. A meta-analysis of 94 studies found that plans written in this if-then form produced a medium-to-large effect on whether people did what they intended, and the proposed mechanism is that the situation itself starts cueing the action rather than a decision having to.

Why if-then and not a resolution

A resolution — I am going to drink less this month — is a statement about a period. Nothing in it tells you what to do at half past six on Thursday, which is when the question actually gets asked. An implementation intention names the moment and answers it in advance, so that at half past six on Thursday there is nothing left to work out.

Peter Gollwitzer and Paschal Sheeran gathered 94 studies of this and found a medium-to-large effect on whether people carried out what they had intended. The size is not the interesting part; the mechanism is. The claim is that the specified situation comes to cue the specified action directly, so acting stops requiring a decision at the time — which is exactly the resource you do not have in the situations that matter here.

Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69–119.

The half people get wrong

Almost everybody writes a good if and a bad then.

The bad then is an absence. “When I sit down after dinner, I do not pour one.” That is a rule, not a plan, and it has the defect the whole approach exists to fix: at the moment of use it hands you back a decision. There is no such thing as beginning to not do something, so there is nothing for the situation to cue.

The fix is to name a thing you would be doing instead, and to screen it with one question: could I start this within ten seconds, where I will actually be? Making a pot of tea passes. Going for a run does not, at half past nine, in the rain, from a sofa. The point is not that the replacement is worthwhile. The point is that it occupies the same ten seconds.

Put it before the seam, not after

Most of these situations have a seam — a moment where one thing ends and the next has not started. Sitting down. Getting in the car. Finishing what you were watching. The seam is the last point at which you are still deciding anything, and a plan placed after it has already missed.

So “when I sit down after dinner, I make tea” is weaker than “when dinner finishes, I put the kettle on before I sit down”. Same intention, ten seconds earlier, and those ten seconds are most of the difference.

Mini Breaks is a place to keep one of these and have it in front of you when the situation arrives, rather than in a note you will not open. See Mini Breaks on the App Store.

Three worked examples

Weak: I will not scroll in bed. Better: when I go upstairs, the phone goes on the charger in the hall. The action is physical, it happens before the situation completes, and it does not require you to be sensible at midnight.

Weak: I will smoke less at work. Better: when I step outside for a break, I walk to the end of the road and back. Same break, same door, the twenty minutes accounted for.

Weak: I will stop buying things when I am fed up. Better: when I open the app in the evening, I put it in the basket and close the tab. That one is unusual in that the replacement is a version of the thing, which is fine — the plan only has to survive the next ten seconds.

Where to keep them

Written down somewhere you will see in the situation, and nowhere else. A plan in a note you never open is a resolution with extra steps. The reason a plan for one situation beats a list covering five is partly this: one plan can be in front of you at the moment it is for, and five cannot.

Working out which situations to write plans for is its own job, and usually a surprising one. There is a method on how to work out the situations it happens in.

Common questions

How many if-then plans should I have?

One, until it stops needing attention. The evidence is about plans for a specified situation, and a plan covering several situations is not that. It is also a practical matter: four plans means four chances a day to discover that this situation was not one of the four.

What if the situation does not come up for a week?

Then you have learned something useful about which situation to pick, and it is worth swapping to one that occurs most days. A plan for a rare situation is untestable, and untestable plans stay theoretical.

Can the action half be “leave the room”?

Yes, and it is one of the better ones. It is physical, it takes under ten seconds, it is available almost everywhere, and it changes what is within reach rather than relying on you to want it less.

Does writing it down matter, or is thinking it enough?

The studies generally had people write or state the plan, so written is the version with evidence behind it. There is also the practical argument: you cannot look at a thought at half past six on Thursday.

Is this the same as a habit?

It is the mechanism people usually mean by one. The difference worth keeping is that a habit is a description of something that has become automatic, and an if-then plan is a thing you write on purpose before that has happened.