The 66-days number is everywhere. It's on productivity podcasts, in the back-of-book pitch of about half the habit books published since 2014, and on a thousand Instagram tiles in Helvetica Light.
It comes from one real study, and the study's authors would probably wince at how the number is being used.
The actual study
Lally et al., 2010, published in the European Journal of Social Psychology. They followed 96 people who picked a single behaviour they wanted to make a habit — eating fruit at lunch, doing 50 sit-ups before breakfast, that kind of thing — and tracked how long it took until the behaviour felt automatic.
The average was 66 days. The range was 18 to 254 days. That range is the part that gets cut from the Instagram tile.
Some habits became automatic in three weeks. Some took most of a year. The variance depended on how complex the behaviour was, how reliable the daily cue was, and how the person felt about the habit. Drinking a glass of water before lunch is easier to automate than doing 50 sit-ups before breakfast. This will not surprise anyone who has tried both.
What 66 days isn't
It isn't a deadline. It isn't a guarantee. It isn't a clock that starts on day one and ends with you cured.
Most habit-formation literature treats it like an oven timer. Set it for 66 days, walk away, come back to a new life. That's not how the study describes it and not how the brain works.
What's actually happening
The neuroscience version: when you repeat a behaviour, the neural pathway involved gets myelinated. Myelin is the fatty insulation that wraps nerve fibres and makes the signal travel faster. The more you use a pathway, the more myelin gets laid down, and the more "automatic" the behaviour feels.
The catch is that myelination isn't all-or-nothing. It's a gradient. Day 30 you've got a thin coat. Day 90 you've got a thicker one. Day 300 you've got a habit that survives stress, illness, travel, and the death of your motivation.
This is why the difference between 66 and 254 days matters. A habit at 66 days is workable but fragile. A habit at 200 days is just how you live.
How to use the number honestly
One: assume 66 days is the floor, not the ceiling. If your habit is complex or you have ambivalent feelings about it, expect 90 to 120 days. If it's simple and you actually want it, you might be done in 30.
Two: design for the variance. Pick one habit. Just one. Run it daily — daily, not "most days" — for at least three months. The biggest cause of failure is starting six habits on Monday and abandoning four by Thursday.
Three: do not skip two days in a row. Lally's data shows that one missed day does not derail the habit. Two in a row starts to. The body interprets a two-day gap as new information: maybe we're not doing this any more.
Four: track the right thing. Don't track outcomes. Track inputs. If your habit is journalling, the tracker is "did I open the journal." Not "did I have a profound insight." Insights are weather. The habit is climate.
A habit isn't an event. It's an erosion. Drop matters. Day matters more.
Why Selfism uses the number
The Selfism app, when it ships, is designed around a 9-week arc — 63 days. The 9 weeks isn't a marketing number. It's the conservative end of the Lally range for simple inner-state habits like the morning check-in or the say-no practice.
Nine weeks is also long enough to survive at least one bad week, which is the test that distinguishes a real habit from an enthusiasm. The first month most people are fine. Week five tends to be where the enthusiasm dies. If you can keep going through week five, the practice usually carries you the rest of the way.
The honest deadline
You will not be a fundamentally different person in 66 days. Anyone who promises you that is a salesperson.
You will have a single new default in 66 days, if you actually ran the practice daily, and that default will pay for itself for the rest of your life. Compounded across ten years of one-new-default-every-quarter, that's a different life. But it doesn't look like a transformation. It looks like a Tuesday in which one thing went slightly better than it used to.
Before starting a new habit, knowing your baseline helps you know what to track. The Selfist Score gives you eight dimensions in five minutes.
Take the Selfist Score →A short reading list, if you want the real thing
The original Lally paper is open-access in most university libraries; search "How are habits formed: Modelling habit formation in the real world". It's short and readable. Wendy Wood's Good Habits, Bad Habits is the longer academic treatment. James Clear's Atomic Habits is the practical version, with the caveat that it sometimes treats the Lally number as more definitive than the paper does.
Read the paper before the books. The paper is the actual thing.