The number 21 has lived in motivational books, courses and app descriptions for half a century. It has a source, and the source is not about habits.
Where the number came from
Maxwell Maltz was a plastic surgeon. In “Psycho-Cybernetics”, published in 1960, he described an observation from his practice: patients needed about twenty-one days to stop noticing a new face in the mirror, and someone who had lost a limb needed roughly the same to stop feeling it there.
He wrote “a minimum of about 21 days”. He was describing adjustment to a changed body, not the formation of new behaviour. The book sold in the millions, the qualifier “minimum” vanished, “adjustment” became “habit”, and out came a rule people still repeat.
What the study found
In 2010 Phillippa Lally and colleagues at University College London published a study in which 96 people spent twelve weeks performing one self-chosen action daily and rating how automatic it felt.
The median was 66 days. The range ran from 18 to 254. So 21 days sits at the bottom edge of the distribution: enough for something very simple, like a glass of water with lunch, and nowhere near enough for squats or a run.
Which gives the first practical conclusion: the difficulty of the action matters more than the calendar. “Drink a glass of water” automates faster than “do fifty squats”, and the difference is measured in months rather than days.
The most useful part of that study rarely gets quoted
A missed day did not break the process. A single lapse had no meaningful effect on the eventual level of automaticity. The formation curve turned out to be robust to individual failures, which directly contradicts how motivation works in most trackers.
A tracker draws a streak. The streak breaks on the first miss and the number resets to zero. Someone sees a zero after forty days and quits, not because the habit disappeared but because the counter did. The punishment the interface hands out for one missed day is wildly out of proportion to what that day actually cost.
That is why habits in KIPRI have a freeze: a day that does not break the streak. It is not indulgence, it is an attempt to make the interface agree with what is known about habit formation. Illness, a flight and a bad week at work do not undo the forty days that already happened.
What genuinely speeds things up
Anchoring to something that already happens. Peter Gollwitzer showed in 1999 that phrasing an intention as “when X happens, I will do Y” sharply raises the chance of following through, compared with a vague “I should”. Not “study Spanish” but “after I put the kettle on in the morning, ten minutes of Spanish”. The trigger is something that already occurs daily without your involvement.
The same time and the same place. Automaticity is built on a repeated context, not just a repeated action. The same action at a different time each day automates noticeably worse.
Shrinking it to something laughable. There is a whole argument about that, but briefly: streaks are held together by bad days, not good ones, and the goal has to be achievable on the worst of them.
How to count honestly
Count the proportion rather than consecutive days: how many times out of thirty. Eighty per cent is an excellent result, and it resembles real life far more closely than thirty out of thirty.
And stop waiting for the date after which it becomes easy. There is no such date. There is a gradual decline in resistance stretched across two or three months for anything of moderate difficulty. The 66-day figure is useful for exactly one thing: it saves you from disappointment on day twenty-two, when the book said it should have gone automatic and it has not.
The study’s numbers
Where you are on Lally’s scale
The marks on the scale are figures from the 2010 study: the fastest time, the median and the longest. Move the slider to how many days you have kept your habit.
Did this help?
Thanks, that is the best answer. If it helped, it may help someone else too.
Thanks for saying so. Tell us what was missing and we will add it: [email protected]
