A habit is a cue and a routine, and the cue is the part you can move

What automatic means here

A habit, in the research sense, is a behaviour that has become associated with a context to the point where the context alone is enough to start it. The defining feature is not frequency or duration but the loss of a deliberate decision step: the behaviour begins before any weighing of options occurs, and often before you notice it has begun.

That is why habitual behaviour is so resistant to intentions. An intention operates on the deliberate system, and the habit has stopped consulting it. People routinely report doing something they had firmly decided not to do, and then being unable to describe the moment of choosing, because there was no such moment to describe.

The cue is usually a situation, not a feeling

Popular accounts describe habits as triggered by emotional states, and sometimes they are. But the cues that do most of the work in everyday behaviour are duller: a location, a time of day, the preceding action, the presence of a particular object, or a particular person. Arriving home. Sitting down at a desk. Finishing a meal. Picking up keys.

This matters because situational cues can be changed and emotional states largely cannot, at least not directly. If a behaviour begins when you sit on a particular sofa with a particular device within reach, moving the device is an intervention available today. Resolving to feel differently while sitting on the sofa is not really an intervention at all.

It also explains a phenomenon that surprises people: habits often break by themselves during a move, a change of job or a disrupted routine. The behaviour did not weaken. Its trigger stopped occurring.

Identifying the cue is harder than it sounds, precisely because the behaviour bypasses the part of you that would have noticed. The usual method is to record the circumstances of the next several occurrences — where you were, what had just finished, what was within reach, who else was present — rather than to introspect after the fact. Reconstruction from memory tends to produce a plausible story about mood, because mood is what introspection has access to, and the actual trigger is often something dull enough to be invisible.

Why substitution beats suppression

A cue that has reliably preceded a behaviour will keep arriving, and it will produce something. Attempting to respond with nothing leaves an unresolved prompt, which is why suppression is effortful in a way that redirection is not. Attaching a different routine to the same cue gives the prompt somewhere to go.

The substitute has to fit the slot. It needs to be available in the same situation, take roughly the same amount of time, and require no more setup than the original, because any friction at the moment of the cue will send the behaviour back to the path with less resistance. A replacement that requires going to another room will lose to one that does not, regardless of which is better in principle.

Where the popular model overreaches

The cue-routine-reward description is a useful frame and it is not a complete theory. The role of reward in maintaining an established human habit is less clear than the tidy version implies; much of the underlying work comes from animal learning, where reward can be controlled precisely, and the translation to complex human behaviour in ordinary environments involves assumptions that are hard to test.

Some habits persist with no identifiable reward at all, and continue after the original reward has disappeared, which is one of the observations that made habit a separate concept from goal-directed action in the first place. Treat the model as a way of organising the question rather than as an established mechanism, and be suspicious of any account that explains every behaviour in its terms without difficulty.

What this implies about effort

The practical consequence is that effort spent early, on arrangement, is worth much more than effort spent later, on resistance. Deciding where an object lives, what happens immediately after a particular daily event, or which room contains which activity is a one-off cost that then operates automatically. Resisting in the moment is a recurring cost paid at exactly the times you have least capacity for it.

This is also why advice framed as trying harder tends to fail and advice framed as changing the situation tends to work better. It is not about willpower being a myth — the research on that is contested and the strong resource model did not survive replication — but about the simpler observation that a cue which never fires cannot start anything.