How long does it actually take to change?

Longer than any number you have been given, and the visible part arrives near the end. Most people quit during the stretch where it is working and cannot be seen.

· Sals Sky

The short answer: **real change is not slow so much as invisible for most of its length.** The number people quote — twenty-one days, sixty-six days, whatever it is this year — describes habit formation under laboratory conditions, and almost nothing anybody actually wants to change is a habit.

The useful question is not how long. It is what is happening during the part where nothing appears to be happening.

Two different things, one word

Habits are one thing: flossing, a morning walk, a language app. Those do form on roughly the timescales in the studies, and they respond to the standard advice about cues and repetition.

Becoming a different kind of person is not that. Not being the one who snaps at people; not needing everybody to approve; actually doing your work instead of arranging it. These have no cue-and-reward loop, and treating them as habits produces a great deal of effort and no movement — which most people then read as evidence about themselves rather than about the model.

What the invisible stretch consists of

During it, the old behaviour keeps happening and the only thing that changes is *when you notice*.

First you notice a week later. Then the next day. Then an hour after. Then during — still doing it, and watching yourself do it, which is awful and is the single most reliable sign of progress there is. Then, eventually, just before, with a gap in which something else is possible.

That progression is the change. It is not preparation for it. And none of it is visible to anybody, including you, because the observable behaviour is identical until nearly the end.

Which is why people quit at the worst moment

The stretch where noticing has moved from a week to an hour feels like total failure — you are doing the thing *and* you are now conscious of it, so you have added self-disgust and subtracted nothing.

That is the point at which almost everybody concludes it is not working. It is the point at which it is most obviously working, and nobody tells anybody this.

How to tell it is moving

Stop measuring the behaviour and measure the gap. How long between doing it and knowing you did? That number shrinks long before the behaviour changes, and it is the only early evidence that exists.

Where this is in the book

Nothing about the protagonist's change is announced. Read chapter one and then chapter two and the man is different, and no scene contains the moment he became so — which is not a technique. It is how it goes.

More