Implementation Intentions: The If-Then Plan That Works

Most intentions are stated as goals: I'll exercise more. I'll review my notes weekly. I'll reply to that email. Almost all of them fail, and the reason is not motivation. It is that a goal contains no information about when the behaviour actually happens, so it competes with everything else for a slot that never arrives.

An implementation intention adds that information in a specific format:

If [situation], then I will [behaviour].

"If it is Tuesday at 9am, then I will review last week's notes." "If I finish a call, then I will write three lines before opening anything else."

Peter Gollwitzer's meta-analysis with Paschal Sheeran, covering a large number of studies across many behaviours, found a medium-to-large effect on goal attainment. For a technique that costs one sentence, that is an unusual return.

Why it works

Two mechanisms, and both matter for how you write the plan.

The cue becomes highly accessible. By specifying a concrete situation, you make that situation more likely to be noticed when it occurs. Much of the failure of good intentions is not refusal — it is that the moment passes unnoticed.

The response becomes more automatic. Linking a specific cue to a specific action delegates the initiation to the environment rather than to a decision. You are not deciding whether to review your notes on Tuesday; the arrival of Tuesday 9am triggers it.

That second part is the important one. The technique works by removing a decision, and decisions at the moment of action are where intentions reliably die — which is also why a capture step with five choices in it gets abandoned and a one-gesture one survives.

How to write one properly

The format is simple and most people get it wrong in the same three ways.

Be concrete about the cue. "When I have time" is not a cue. "After I put the kettle on" is. Cues that work are things that reliably happen and that you cannot miss: a time and place, the completion of an existing habit, or a specific event.

Be concrete about the action. "Then I will work on the report" is too big. "Then I will open the document and write the section headings" is startable. Vague actions reintroduce the decision you were trying to remove.

Anchor to something that already happens. Existing routines are the most reliable cues available, because they are already automatic. This is why "after coffee" outperforms "at 9am" for many people.

Bad: I'll review my notes more often. Better: Every Sunday I'll review my notes. Good: If it is Sunday evening and I have made tea, then I will open my inbox note and process it for ten minutes.

The obstacle variant

A second form handles interference, and it is the one people skip:

If [obstacle], then I will [response].

"If I open my laptop and see 40 emails, then I will close the tab and start the writing block first." "If I feel too tired to review, then I will do only the first three notes."

This works because the obstacle is usually predictable. You already know what derails you. Deciding the response in advance, once, is far easier than deciding it in the moment when the obstacle is present and the easy option is right there.

Pairing a goal intention with obstacle plans is essentially what "mental contrasting with implementation intentions" formalises, and it outperforms the if-then plan alone.

Where it fails

Three honest limits.

It does not create motivation. If you do not actually want the outcome, an if-then plan will not manufacture the desire. The technique converts an existing intention into action; it does not supply the intention.

It degrades with too many plans. A handful of active if-then plans works. Twenty does not, because the cues start competing and none becomes automatic.

Effect sizes in the wild are smaller than in the lab. As with most behavioural findings, real-world implementation is messier and the benefit is real but smaller than the headline numbers suggest.

It is also worth noting that some of the larger claims in this area rest on studies that have not all been replicated at full strength. The core finding — specifying when and where helps — is well supported; treat the more dramatic numbers with the usual caution.

Where this applies to notes

Almost every note-taking habit that fails, fails for exactly the reason this technique addresses: there is no specified moment.

  • Capture. "I'll write things down" fails. "If I have a thought while walking, then I will speak it into my phone before I reach the door" works. See the capture habit.
  • Processing. "I'll clear my inbox" fails. "If it is Friday at 4pm, then I will empty my note inbox" works. See the note inbox and the weekly review.
  • Review. "I'll revisit my notes" fails. Attach it to something that already happens.
  • After meetings. "If a call ends, then I will write three lines before opening anything else." See how to take notes on a phone call.

And the design consequence for tools: the smaller the action after the cue, the more likely the plan holds. A system where the then is "unlock, open app, choose folder, title, type" has too much in it. That is the reasoning behind one-gesture capture in Clair Mind — not convenience for its own sake, but because a short then is what makes an if-then plan survive.

The minimum viable version

Write one sentence. If [something that reliably happens], then I will [one small specific action]. Put it somewhere you will see it once.

One plan at a time until it is automatic.

More in how to stop procrastinating, time blocking, and memory and focus.

More in Memory & Focus

75 more notes on this branch.