A habit loop is the three-part structure behind a repeated behaviour: a cue triggers it, a routine runs, and a reward tells your brain the sequence was worth keeping. The habit loop examples further down name all three parts for five ordinary behaviours, which is the step most people skip. Charles Duhigg popularised the model in The Power of Habit, and the reason it's still useful two decades on is that it tells you where to intervene. You almost never change a habit by attacking the routine in the middle. You change the cue or the reward at either end.
Below are five real habit loops with all three parts named, which is harder than it sounds — most people can describe their routine and have never identified their cue. That's usually the whole problem.
If you're looking for which habits are worth building in the first place, keystone habits is the companion piece from the same book; this post is the mechanics of a single loop.
Five habit loops, pulled apart
| Habit | Cue | Routine | Reward | The change that works |
|---|---|---|---|---|
| Evening phone scrolling | Sitting on the sofa, TV on | Pick up phone, open app, scroll | Relief from boredom, low-effort novelty | Move the phone to another room — the cue is proximity, not the app |
| Afternoon biscuit at 15:30 | Energy dip plus walking past the kitchen | Kettle, biscuit tin | Break from the screen, sugar | Keep the break, change what's in the tin — the reward was the pause |
| Skipping the gym after work | Arriving home and sitting down | Coat off, sofa, "tomorrow" | Immediate comfort | Don't go home first; the cue is the front door |
| Morning run that actually happens | Alarm, then shoes already by the bed | Same route, same distance | Feeling ahead of the day by 07:15 | Keep the visible cue; the shoes do the work |
| Checking email first thing | Phone is the alarm clock | Unlock, inbox, 20 minutes gone | Feeling on top of things | Buy a clock — this loop is a hardware problem |
Read the last column. Four of the five changes are environmental, not motivational. The one that isn't — swapping what's in the biscuit tin — works because it keeps the reward intact and only changes the routine's ingredients.
Why the cue is the part people miss
Ask someone why they scroll in the evening and they'll say they're tired, or bored, or that it's a bad habit. None of those are cues. The cue is the sofa, the time, and the phone being within arm's reach — a context, not a feeling.
This is the practical core of Wood and Rünger's review of habit psychology: habits are context-cued responses, which means the context is doing more of the work than intention is. It also means a habit you can't perform when the context is absent isn't broken — you've just left its trigger behind. People who move house famously lose habits they'd had for years, in both directions.
So when you write your own loop down, be concrete about the cue. Not "stress" but "the moment I close the laptop". Not "evenings" but "after the kids are in bed and I sit down".
Drawing the loop for one of your own habits
The diagram everyone draws is a circle with three labels, and you can do it in four lines of text instead. Take one habit you want to keep or lose and finish these sentences:
- Cue: the last thing that happens before it, every time. If it varies, list the two most common.
- Routine: what you actually do, in the order you do it, including the small first step.
- Reward: what you get in the ten seconds after. Be honest — usually it's relief, novelty, or a break, not the thing you'd say out loud.
- The craving: what you notice wanting when the cue fires and the routine is unavailable.
That fourth line is the one that identifies the loop correctly. If you take the biscuit away and what you miss is the walk to the kitchen, the reward was the break. If what you miss is the sugar, it wasn't.
Which part to change first
Change the cue to stop a habit. Change the reward to keep one.
Stopping works best by removing or relocating the trigger, because a cue that never fires needs no willpower. That's why the phone-in-another-room intervention outperforms every screen-time app: it deletes the loop instead of interrupting it.
Building works best by making the reward arrive sooner. Most habits worth having pay off in months — fitness, reading, saving — and months is far too slow to reinforce anything. A visible record is the standard fix, because it produces a small immediate reward for the behaviour on the day you do it, rather than in April. How to be consistent with habits goes further into that gap between the effort and the payoff.
Where the model runs out
The loop is a good map and not a complete one. It's better at explaining simple, repeated, context-bound behaviours — the biscuit, the scroll, the shoes — than at explaining anything that requires planning. Nobody files a tax return on a habit loop.
It also under-sells how long the process takes. The often-quoted UCL work on habit formation found automaticity taking anywhere from 18 to 254 days depending on the person and the behaviour, so a loop you've correctly identified still needs months of repetition before it runs itself. Identifying the cue is the fast part. Everything after that is patience.
For the specific case of dismantling a loop you'd rather not have, habit tracker for breaking bad habits covers the count-the-clean-day approach, and examples of bad habits has the wider list with the swap for each one.
FAQ
What are the three parts of a habit loop?
Cue, routine, reward. The cue is the context that triggers the behaviour, the routine is the behaviour itself, and the reward is what you get immediately afterwards — usually relief, novelty or a break rather than the long-term benefit you'd name if asked.
What is an example of a habit loop?
Sitting on the sofa (cue), picking up the phone and scrolling (routine), relief from boredom (reward). The useful part of writing it out this way is that it shows the intervention: move the phone, and the cue stops firing.
Should I change the cue or the reward?
Change the cue to break a habit and the reward to build one. Removing a trigger costs no willpower, whereas a habit you're trying to establish usually fails because its real payoff is months away and needs something immediate standing in for it.
How do I make the reward arrive sooner?
Make the behaviour visible on the day you do it. A filled square in a tracker like init.Habits is a small immediate reward for a habit whose actual benefit is months out, which is the gap that stops most new habits from sticking.