If-then plans are the least glamorous thing that keeps working

The gap the technique addresses

People routinely intend things they do not do. The gap between intention and action is one of the most reliably documented facts in behavioural research, and it is not primarily a failure of sincerity: the intention is real at the time it is formed and remains real afterwards. What is missing is any connection between the intention and the moment at which action would have to occur.

A general intention has no trigger. It sits in memory as a goal, and goals have to be retrieved and acted upon deliberately, which requires noticing that this is an appropriate moment — precisely the thing that fails when you are busy, tired or absorbed in something else.

What the technique is

An implementation intention is a plan in a specific form: if situation X arises, then I will do Y. Not an intention to exercise more, but a specification that when the working day ends on a particular set of days, the next action is to change and leave the building. The situation must be concrete enough to be recognised without judgement, and the action specific enough to begin without further decisions.

The proposed mechanism is that the plan builds an association between a situation and a response, so that the situation itself brings the response to mind, without the deliberate retrieval a general goal requires. In that respect it borrows the structure of a habit deliberately, rather than waiting for one to form.

How good is the evidence

This is one of the better-supported techniques in applied psychology, with a large body of studies across many behaviours and populations, and summaries of that literature generally find a positive effect on whether people act on their intentions. It has also fared better than several neighbouring ideas that collapsed under replication.

That said, the caveats are real and worth stating. Much of the evidence comes from short-term studies of relatively simple behaviours, often with volunteers and self-reported outcomes. The published literature in this area, like most, is likely to overrepresent positive results, so the true average effect is probably smaller than a naive reading of the summaries suggests. And the technique appears to work best where the obstacle is forgetting or failing to notice the opportunity, which is not every obstacle.

When it does little

If the reason you are not doing something is that you do not want to, an if-then plan will not supply the motivation. The technique addresses the translation of an existing intention into action, and it assumes the intention exists. This is a genuine limit rather than a caveat, and it explains a number of failed attempts: people apply it to goals they have adopted because they think they should, and the plan fires and is ignored.

A related limit shows up with behaviours you are trying to stop rather than start. Here the competing response is often already automatic and already attached to the cue, so the new plan is not filling an empty slot but contesting an occupied one. The technique still helps in that situation, but expecting it to perform as it does on a fresh behaviour sets it up to look like a failure when it is simply working against more resistance.

It also does little where the barrier is structural. No specification of when you will study helps if the room is occupied, the material is unavailable, or the time named does not actually exist in your week. A plan that collides with reality is not a planning failure but a scheduling one, and treating it as the former produces a lot of unnecessary self-blame.

Writing one that works

The cue should be something that will unambiguously happen, that you will notice, and that recurs. Anchoring to an existing routine works well because the routine already occurs reliably: after the last meeting of the day, immediately on returning home, once the table has been cleared. Anchoring to a feeling or an intention — when I feel ready, when I have a moment — reproduces the original problem, since those states never announce themselves.

The response should be the first physical action rather than the whole activity. Putting on shoes, opening the specific document, taking the item out of the bag. The first action is what the plan can trigger; the rest follows from being started, and specifying the whole activity tends to make the plan feel large enough to postpone.

One more thing, which the evidence supports more weakly but which costs nothing: write the plan down in the if-then form rather than merely thinking it. The formulation is doing work, and rehearsing it as a sentence is how it gets stored in the form the mechanism needs.