The Habit Loop: Cue, Routine, Reward, Explained

The habit loop is a three part cycle popularised by Charles Duhigg in The Power of Habit (2012): a cue triggers a routine, and a reward tells the brain to store the sequence. James Clear added a craving between cue and response. The reward builds the loop then stops steering it: behaviour continues even when the reward is devalued.

Almost every explanation of the habit loop ends with the same advice: find the reward, change the reward, and the habit changes with it. The experiments psychologists use to prove a behaviour has actually become a habit do the opposite. They strip the value out of the reward, then watch the behaviour carry on anyway. That result is the whole reason habits get studied separately from ordinary goal-directed action, and it is missing from the pages that explain the loop.

What is the habit loop?

The habit loop is a three part cycle: a cue triggers a routine, and a reward reinforces the link between them, so the same cue is more likely to produce the same routine next time. Charles Duhigg named it in 2012 and drew it as a circle. In his words, “over time, this loop, cue, routine, reward; cue, routine, reward, becomes more and more automatic”. The label stuck because it turns a vague feeling (some things you just do without deciding) into three pieces you can point at.

Is the habit loop a real scientific theory?

It is a popular label for real findings, not a theory published under that name. Searched on 11 September 2026 through the NCBI E-utilities endpoint, the exact phrase habit loop returns 0 results in PubMed titles and abstracts. Habit formation returns 972, and outcome devaluation returns 273. The one PubMed abstract containing the string cue routine reward is a 2021 paper in Animals about improving equine welfare by changing human habits.

That is not a debunking. The science under the label is real and old. It is simply filed under different words, which is why searching for the blog vocabulary in a research database finds nothing. Here is the translation.

The brain claim is worth pinning down, because it is repeated loosely everywhere. Jog and colleagues recorded ensembles of striatal neurons in rats learning a T-maze and published the result in Science in 1999. As the task became automatic, neuronal activity reorganised into a pattern emphasising the beginning and the end of the sequence, and those patterns then stayed stable for weeks. That is the finding Duhigg built on. It says a practised sequence gets packaged as a unit. It does not say a reward keeps running it.

Does the reward keep the habit going?

Only while you are learning it. Once a behaviour is well practised, taking the value out of the reward barely changes it, and that insensitivity is how researchers decide a behaviour has become a habit at all. The test is called outcome devaluation, and it has been in the literature for more than forty years.

In Adams (1982), hungry rats were trained to press a lever for sucrose. The sucrose was then devalued by conditioning a food aversion to it, so the animals no longer wanted it. Rats given limited training cut back on pressing, as you would expect from a behaviour aimed at an outcome. Rats given extended training carried on pressing. Same lever, same now-unwanted reward, different amount of practice.

Tricomi, Balleine and O’Doherty reproduced the pattern in people in 2009, with brain imaging. Their conclusion, in their own words: extensive training on a free-operant task “reduces the sensitivity of participants’ behavior to a reduction in outcome value”. Cue-driven activity in a region of the posterior putamen rose as training went on. Smith and Graybiel found the same signature at the cell level in 2013: the striatal action-bracketing pattern they recorded was “insensitive to reward devaluation but sensitive to running automaticity”.

We read the pages currently ranking for habit loop on 11 September 2026 to see how many mention this. The answer is none of them. Two out of seven name a habit researcher at all.

What actually triggers a habit in daily life?

The context does, and you will probably tell yourself it was the goal. Neal, Wood, Labrecque and Lally tested exactly that in 2012 with two studies. In the first, 65 students described where they ran and why they ran, and 53 went on to a lexical decision task where those personal words were flashed below the threshold of awareness. Students with stronger running habits recognised the words running and jogging faster after being primed with their own context words. Priming with their own goal words produced no such effect.

The second study used 102 undergraduates and a different behaviour, speaking loudly in a sports stadium. Students who habitually went to their campus stadium spoke more loudly when primed with pictures of it, and, in the authors’ words, “this effect occurred regardless of whether or not they had a goal to visit stadiums after the study”. Meanwhile the same participants rated their goals as more important the stronger their habit was. Actual trigger and perceived trigger point in different directions.

Strong habits fire on the context. People report that they fire on the goal. Both statements are in the same paper, and both are true.

The cleanest demonstration is a field experiment the same group ran in a cinema in 2011. Habitual popcorn eaters ate the same amount of stale popcorn as fresh popcorn, and their hunger and their liking for the food made little difference. Two things did break the habit. Moving them out of the cinema and into a meeting room broke it, and asking them to eat with their non-dominant hand broke it. In other words, remove the context or block the automatic movement and the behaviour returns to being decided. Make the taste worse and it does not.

Habit loop examples

The useful column in a habit loop example is not the reward, it is what would actually stop the behaviour. Most example tables online stop at three columns. The fourth here is the one you can act on.

How do you change a habit loop?

Work on the cue and on the repetitions, because those are the two parts a practised habit is still sensitive to. Four things follow from the research above, in order of how much they buy you.

1. Change the context, not the payoff. The popcorn eaters stopped in a different room, not when the popcorn got worse. Moving house, changing a commute or moving an object is a bigger intervention on an old habit than any reward you can design.
2. Decide the replacement in advance, and phrase it as an action. A plan written as a refusal can backfire. We covered the experiment on that, and how to phrase the replacement instead, in how to break bad habits.
3. Keep the new behaviour in one fixed context. Repetition in a stable context is the mechanism the whole literature agrees on. It is also what makes automaticity arrive at all, which in the Lally study took a median of 66 days, not 21. The spread was wide, and how long it takes to build a habit goes through the distribution.
4. Count the repetitions rather than trusting the feeling. The Neal study is a warning about introspection: people rate their goals as more influential precisely when goals matter least. Your own read on why you do something is not reliable evidence. A record is.

If you are building rather than breaking, the same logic drives the practical guide in how to build good habits. And if the cue, craving, response, reward version is the one you met first, our Atomic Habits summary walks through the four laws built on top of it. Procrastination runs on the same kind of loop, with relief as the payoff, which we take apart in how to stop procrastinating.

Where does a habit tracker fit in the loop?

A tracker does not touch the cue and it does not manufacture a reward. What it does is make the repetition count visible, which matters because the repetition count is the variable that turns a decision into a habit, and because your own sense of how consistent you have been is exactly the thing the Neal study shows to be unreliable.

StellarHabit is built around that narrow claim. It records whether the behaviour happened, reports it as a consistency percentage rather than a streak, and shows it to friends you chose. The percentage matters because a streak deletes your history the first time you miss, which is the opposite of what a repetition count should do; we argued that out in streaks versus consistency. The visibility matters because a record nobody sees is easy to stop keeping.

The honest limits: it will not move your cues for you, it has no financial stake, and it is not built for tracking alone with nobody watching. If you want a private ledger, a notebook does the same job.