Implementation Intentions: If-Then Plans Tested in 94 Studies
An implementation intention is a plan in the form "when X happens, I will do Y". Across 94 independent tests, making one raised goal attainment by a medium-to-large margin (d = 0.65), according to the largest review of the evidence. It is one of the best-tested tools in the psychology of habits, and one of the cheapest: it takes a sentence.
This page covers what an if-then plan is, what the studies found, how to write one that works, and where the evidence stops. Every figure links to its source so you can check it.
What is an implementation intention?
An implementation intention is a plan that ties a specific action to a specific moment: "If it is 7 a.m. and I have finished my coffee, then I will walk for ten minutes." A goal says what you want. An implementation intention says when, where and how you will act on it, so the moment itself prompts the action instead of your memory.
The term comes from the psychologist Peter Gollwitzer, who set out the idea in the 1990s. His starting point was a problem most people know: wanting something, even wanting it badly, does not reliably lead to doing it. In a review of that gap, Gollwitzer and Paschal Sheeran note that goal intentions account for only a modest share of what people actually do.
An if-then plan works on a different step. It does not try to make you want the goal more. It decides in advance what you will do when a particular situation shows up, so that when it does, you do not have to stop, weigh it up and decide again.
Two parts carry the plan:
- The "if" part is a cue you will meet: a time, a place, or something you already do. "When I sit down at my desk after lunch."
- The "then" part is one concrete action. "I will write the first line of the report."
A plan that says "exercise more, starting tomorrow" has neither part. It names no moment and no action, so the day arrives and nothing in it tells you to begin.
What 94 studies found about if-then plans
The biggest review pooled 94 independent tests and found that people who made if-then plans reached their goals more often than people who only set the goal. The average effect was medium-to-large (d = 0.65). Plans helped people get started (d = 0.61) and helped them stay on track once started (d = 0.77).
The review is Gollwitzer and Sheeran, Advances in Experimental Social Psychology, 2006, with an open copy at the University of Konstanz. The two sub-effects come from the same authors' summary of that review.
What does d = 0.65 mean in plain terms? A standard way to read it: the average person who made an if-then plan did better than about three in four people who only set the goal. It is an average across very different studies, so some plans helped a lot and some barely at all.
The review found plans helped with four separate problems:
- Getting started. Remembering to act, spotting the chance, and pushing past the first reluctance.
- Staying on track. Keeping a goal going when something distracting or tempting turns up.
- Letting go of a plan that is not working, instead of pouring more effort into it.
- Saving effort for later, because the decision was already made.
For habits, the first two matter most. A new habit often fails before it starts: the moment passes before you notice it.
Classic experiments behind if-then planning
The early studies used ordinary goals. Students who made a when-and-where plan for a hard holiday project finished it about two times in three, against one in four without a plan. In another study, three in four students with a plan sent a required report on time, against one in three without. Plans mattered most when the task was hard to start.
These come from Gollwitzer's 1999 paper in American Psychologist, "Implementation intentions: strong effects of simple plans", which reviews the first wave of experiments. An open copy is free to read.
The holiday projects. Before a Christmas break, students named two projects they meant to finish, one easy and one hard. For hard projects, about two thirds of those who had decided when and where to start finished them; only about one in four of those without such a plan did. For easy projects the plan made no difference: about 80% finished either way.
The Christmas Eve report. In a follow-up, students were asked to write a report on how they spent Christmas Eve within 48 hours. Half were asked to write down exactly when and where they would write it. Three quarters of that group sent it on time, against one third of the rest.
A health check. In a study the 1999 paper describes, women who intended to do a breast self-examination in the next month all did it when they had also made a when-and-where plan, against 53% of those with the intention alone.
Exercise. A later trial with 248 people split them into three groups: nothing, a motivation session about the health risks of inactivity, or the same session plus an if-then plan for exercise. The motivation session made people intend to exercise more but did not change what they did. Adding the plan, the authors wrote, had a "dramatic effect" on exercise over the following week (Milne, Orbell and Sheeran, 2002).
The pattern across these studies is the same one the review found later: the plan did not change how much people wanted the goal. It changed whether they acted when the moment came.
Why a simple plan works
Making an if-then plan seems to do two things. It makes the cue stand out, so you notice it when it comes. And it links the cue to the action ahead of time, so the action starts with less thought. In the review, both processes showed up in the data. In effect, the plan hands the start of a habit to the situation.
The 2006 review tested both ideas. People who had made a plan picked out the cue more readily than people who had not, and their response to it became more automatic: it started quickly once the cue appeared, with less need for a conscious decision.
This is close to how habits form in the first place. A habit is an action that a context triggers without much thought. An if-then plan borrows that structure from the start: you pick the context yourself, instead of letting it form by accident.
It also explains why plans help most with hard-to-start tasks. When a task is easy and pleasant, you start it anyway. When it is dull or daunting, the moment to begin is exactly when your mind looks for a reason to wait. A plan removes that choice point.
Implementation intentions vs habit stacking
Habit stacking is one kind of implementation intention. It uses something you already do as the "if": "After I pour my morning coffee, I will write three lines in my journal." A plan can also use a time or a place. A randomised trial found that linking a new habit to a daily routine and linking it to a set time worked about equally well.
The trial is Keller and colleagues, 2021. Adults picked a nutrition habit and linked it either to a daily routine or to a specific time of day, then reported every day for 84 days. Both groups gained in automaticity and followed their plans more often, with no difference between them. Among people who formed the habit, it took a median of 59 days to reach peak automaticity. What predicted the habit was doing the planned action again and again in response to the planned cue.
So the choice between "after my coffee" and "at 7:30" matters less than two things:
- Pick a cue that happens every day you want the habit, without fail.
- Act on it every time it happens, especially in the first weeks.
Habit stacking goes deeper on choosing an anchor habit. For how long the early weeks last, see how long it really takes to form a habit.
Implementation intention examples
A good if-then plan names one cue you cannot miss and one small action you can do on the spot. "If it is Monday" is too broad; "when I close my laptop on Monday evening" is clear. Write the plan down. The examples below cover starting habits, getting through obstacles and replacing an unwanted habit.
Starting a habit
- If I get out of the shower, then I will do ten push-ups before I dress.
- When I sit on the train home, I will read one chapter before I open my phone.
- When I finish dinner, I will put my running shoes by the door.
- If it is 9:30 p.m., then I will put my phone on the charger in the kitchen.
Getting past an obstacle
- If it is raining when I planned to walk, then I will do the 15-minute video in the living room instead.
- If I get home too late for the gym, then I will do one set of stretches before bed.
- If I skip a day, then I will do the habit at the first cue the next day, at any size.
Replacing an unwanted habit
- When I feel the urge to check my phone at my desk, I will stand up and fill my water glass.
- If someone offers me a second drink, then I will ask for sparkling water.
- When I reach for a snack out of boredom, I will step outside for two minutes first.
The obstacle plans matter as much as the starting plans. The review found if-then plans helped people stay on track when something got in the way, and the obstacle is often predictable: rain, a late meeting, a tired evening. How to break bad habits covers the replacing side in more detail.
How to write an if-then plan in four steps
Name the goal, find the moment that usually stops you, pick one small action for that moment, and write it as "if [cue], then I will [action]". Then check it: could a stranger tell, from the sentence alone, exactly when you will act and what you will do? If not, make the cue and the action more specific.
- State the goal. "Walk most days." The plan serves the goal; it does not replace it. Gollwitzer and Sheeran's work suggests plans help most when you genuinely want the goal in the first place.
- Find the critical moment. When does the habit usually fail? Leaving work, the first ten minutes after waking, the evening sofa.
- Choose one action you can start at once. Small enough that you can begin on the spot. "Put on my shoes" is easier to start than "run five kilometres". Small does not mean trivial: some of the small habits that change your life in health studies take a minute a day.
- Write it down in if-then form. One sentence per plan. Put it where you will see it at the cue, or in the note field of your tracker.
Two practical points. First, one or two plans per goal is plenty; no study sets a number, so treat that as our advice. Second, check your plan against real days for a week. If the cue does not happen every day, pick another one.
Tracking if-then plans
Recording whether you acted on the plan closes the loop. A 2016 review of 138 experiments found that tracking progress helped people reach their goals, most of all when the record was written down. Tracking the plan, not only the result, also shows you which cues work and which you keep missing.
That review is Harkin and colleagues, 2016, and does habit tracking work? goes through what it found. For if-then plans, the useful record is simple: did the cue happen, and did you act on it? Two weeks of that tells you whether the plan needs a new cue or a smaller action.
A paper grid works. So does an app with a one-tap check-off; in Aura, you can name the habit after the plan itself ("After coffee: 3 journal lines") so the cue sits in front of you each time you check it.
What the research does not say
It does not say an if-then plan makes you want something you do not want, or that it works equally well for every goal. The effects are averages across many short studies, often with students and often over days or weeks rather than months. Plans help most when the goal is strong and the start is hard.
A few limits worth stating plainly:
- Plans need a real goal behind them. Gollwitzer and Sheeran report that strong effects showed up mainly when the goal intention was strong. A plan for a goal you do not care about is words on paper.
- Most studies are short. Many measured behaviour over a week or a month. Plans help habits start; whether they hold a habit for years is less studied.
- Easy tasks gain little. In the holiday-project study, plans made no difference for easy projects, which most people finished anyway.
- The average hides wide spread. An effect of d = 0.65 is an average across very different goals and groups; your own result may be larger or smaller.
- The popular numbers are often wrong. Claims that if-then plans make you "2 to 3 times" more likely to succeed, or that they are backed by "over 200 studies", do not come from the 2006 review. The review's figure is 94 tests and d = 0.65.
Frequently Asked Questions
What is an implementation intention?
It is a plan that links a specific situation to a specific action, written as "if X happens, then I will do Y". For example: "When I finish breakfast, I will take my vitamin." The situation acts as the cue, so you do not have to remember or decide in the moment.
Do implementation intentions work?
Yes, on average. A 2006 review of 94 independent tests found that people who made if-then plans reached their goals more often than people who only set the goal, with a medium-to-large effect (d = 0.65). Plans helped most with goals that were hard to start.
What is the difference between a goal and an implementation intention?
A goal says what you want: "I want to read more." An implementation intention says when, where and how you will act: "When I get into bed, I will read ten pages before I pick up my phone." The goal gives you a reason; the plan gives you a moment.
Is habit stacking the same as an implementation intention?
Habit stacking is one form of it. A habit stack uses an existing habit as the cue ("after I brush my teeth, I will floss one tooth"). An implementation intention can also use a time or a place. A 2021 trial found routine cues and time cues worked about equally well.
How many if-then plans should I make?
No study sets a number. We suggest one plan to start a habit and one for the most likely obstacle, per goal. More plans means more cues to watch for, and a plan you cannot remember cannot prompt you.
Who came up with implementation intentions?
The psychologist Peter Gollwitzer introduced the idea in the 1990s, and he and Paschal Sheeran later reviewed the evidence in 2006. Authors such as James Clear popularised the "when-where" plan, but the research behind it is Gollwitzer's and his colleagues'.