What Are Implementation Intentions? Why If-Then Plans Beat Willpower

It's Sunday night and the plan sounds great. This week you're getting up early, doing the hard thing first, and not touching email until 10.

Then it's Wednesday, 7:12 a.m. The phone's in your hand before your feet hit the floor, there's a Slack from your boss, and the hard thing is already sliding to Thursday.

Nothing was wrong with the goal. The plan just never said what you'd do at 7:12 on a Wednesday when the phone lights up. I'm guilty of this too, more than I'd like to admit. Psychologists call the fix an implementation intention, and it's way less fancy than it sounds.

An implementation intention is an if-then plan that ties a specific moment to a specific action. "If I sit down at my desk, then I open the draft before I open email." That's it.

The idea comes from psychologist Peter Gollwitzer, who laid it out in a 1999 paper with a pretty honest title: "Implementation intentions: Strong effects of simple plans." He separated two things most of us mash together. A goal intention says what you want: "I want to write more." "I'm going to stop scrolling in bed." An implementation intention says when, where and what you'll do about it, so the decision is already made before the moment shows up.

Most of us stop at the goal. The goal is the fun part to say out loud on a Sunday.

Yes. But the honest answer is "less dramatically than you've probably read." The effect is real, it's consistent, and it's small.

The number that gets passed around comes from Gollwitzer and Paschal Sheeran's 2006 meta-analysis of 94 tests, which found a medium-to-large effect on reaching goals (d = .65).

In 2025, Sheeran, Listrom and Gollwitzer published a much bigger update in the European Review of Social Psychology. They pulled together 642 independent tests, most of them with adults, run in the UK, Germany, the US, the Netherlands, Canada and China. Across all of it the average effect was d = .36. For behavior change it was d = .27. For cognitive stuff like concentration and memory it was d = .47.

Then they did something I respect a lot. They checked whether the research leaned optimistic, and it did. Small studies with positive results showed up more than they should have. When they ran a stricter correction, the estimate dropped to d = .15, which they call small. The same analysis found "extreme evidence" that the effect is real.

Two more findings stuck with me. The effect didn't fade. Follow-ups six months or more later were as strong as follow-ups under a week (d = .51 vs .48). And plans worked on things measured objectively, not just on what people said about themselves.

So no, an if-then plan isn't going to change your life by Friday. But a small edge that holds up for six months and costs you one sentence? I'll take that every morning.

Because the plan does the deciding ahead of time, so you're not deciding in the moment when you're tired, rushed or staring at a notification.

Gollwitzer's explanation is that an if-then plan hands control to the situation. You pick the cue in advance. When the cue shows up, the action gets going without the usual back-and-forth in your head. The format matters too. In the 2025 review, plans written as a true "if this, then that" beat plans written as a schedule of when, where and what (d = .43 vs .29).

There's a second payoff. Psychologists E.J. Masicampo and Roy Baumeister found that unfinished goals keep popping into your head and pulling at your attention, and that making a specific plan quieted those thoughts, even before the task was done. I went deeper on that in the piece on why unfinished tasks keep you up at night.

Look, willpower is just what you're left with when you didn't decide anything ahead of time. At 7:12 with a phone in my hand, I don't love my odds.

Name a time and a place, write it as an if-then, say it to yourself once, and pick something you actually want. That's the short version of what 642 tests point to.

That last row surprised me. When people kept a copy of their plan and could go back over it, the plans worked a bit less well. The authors think going back over a plan turns it into something you deliberate about all over again, which is the exact thing the plan was supposed to save you from.

These are patterns across hundreds of different studies, not one experiment testing each piece. The authors flag that the number-of-plans finding could be tangled up with other differences between studies. Treat the table as good bets, not laws.

Pick one thing, tie it to a moment you'll definitely run into, and write it as "if, then." It takes about as long as reading this sentence.

If you're stuck on step one because everything feels like the one thing, start with how to decide what matters most when everything feels urgent.

The question picks the one thing. The if-then plan gets it done. Put them together and you've got an implementation intention before breakfast.

That's the question I ask every morning, and I wrote about why this one question works if you want the long version. Here's what it looks like when you add a when and a where to the answer:

If you want the question in front of you before the inbox gets a vote, that's why we built the app. The Question is a free iPhone app that asks one question every morning: What will make today matter? Answering takes under a minute. Midday, it reminds you what you said. In the evening, it checks in. No list to manage. You can get it on the App Store here.

"If it's 6:30 and I've poured my coffee, then I write for 20 minutes before opening email." The format is always a specific situation followed by a specific action. Goals like "write more" or "be healthier" aren't implementation intentions because they don't name a when or a where.

A goal says what you want, like "I want to get the proposal done." An implementation intention says what you'll do when a particular moment arrives, like "If I open my laptop, then the proposal is the first thing I touch." Gollwitzer's research treats them as two separate steps, and the plan is the one that does the work in the moment.

Fewer than you think. In the 2025 review, a single plan had one of the stronger effects (d = .41 across 478 tests), while forming three to five plans had weaker ones. The authors say the number of plans may be tangled up with other differences between studies, so one clear plan a day is a good place to start.

The effect held up over time in the 2025 meta-analysis of 642 tests. Follow-ups six months or more later showed an effect as large as follow-ups under a week (d = .51 vs. .48). Making the planning itself a daily habit takes longer; in Lally and colleagues' 2010 study, new habits took a median of 66 days, and you can read more on that in how long it really takes to build a habit.