Habits small enough that you cannot fail them: why to aim at your worst day, and how long it really takes for a habit to become automatic

BJ Fogg, a behaviour researcher at Stanford, likes to tell people about the first habit he built with his own method: flossing one tooth. Not the whole mouth. One tooth.

It sounds like a joke, and he makes it sound funny on purpose. But the reasoning is serious: there is no day too tiring, too busy or too dull to floss one tooth. And on almost every day, once the floss was in his hand for one tooth, he did the lot.

Aim at your worst day, not your best

A resolution to run thirty minutes every morning usually survives about nine days. Not because thirty minutes is a lot, but because it only works on ordinary days, and life is not all ordinary days. Then comes a morning with a sick child, or a night working late, and the chain breaks.

The central rule is to choose a level so small you can still do it when ill, busy or short of sleep. Not thirty minutes of running but putting on your shoes and stepping out of the door. Not twenty pages but one. It sounds pointless because the bar is too low to produce results. But results are not the goal yet. Keeping the rhythm is.

Attach it to something you already do

A new habit needs an anchor. The most effective is to attach it right after something you already do daily without thinking. Fogg teaches a template: "After I [existing routine], I will [tiny new behaviour]." After I make coffee, I take five breaths. After I brush my teeth at night, I read one page. The old routine becomes the reminder, so you do not have to remember.

The evidence here is fairly clear. In 2002 three British psychologists split people who wanted to exercise into groups. Among those who only read material on the benefits of exercise, about a third exercised at least once a week afterwards. Among those also asked to write one specific sentence, "I will exercise on this day, at this time, in this place", the figure rose above nine in ten.

A habit without an anchor is not yet a habit. It is an intention waiting to be forgotten.

Sixty-six days, and one missed day

You may have heard that it takes 21 days to form a habit. That number does not come from any research on habits. In 2009 Phillippa Lally's team at University College London followed 96 people doing one new thing daily for twelve weeks, such as drinking a bottle of water after lunch or running for fifteen minutes before dinner, and measured when it became automatic.

The average was 66 days, but the range was huge, from 18 days to an estimated 250-plus, depending on the person and the behaviour. Simpler behaviours got there faster. And the most encouraging finding: missing a single day barely slowed things down. What kills a habit is not one slip but treating one slip as failure and giving up.

Three rules for the minimum

• Define the minimum clearly and count it as done. One page still counts as reading.
• Do not make up for missed days. Skip one and do the normal amount the next day, not double. Catching up is the fastest way to feel weighed down and quit.
• One missed day is fine. Two in a row is worth noticing. That is when a habit starts to slide.

When to scale up

Only increase once the current level is automatic, meaning you do it without deciding to. By Lally's findings that usually takes weeks to months, not days. Scaling up too early is the commonest reason a habit that was going well suddenly collapses.

And often you do not need to scale up at all. Like Fogg's floss, once you have your shoes on and are out of the door, on most mornings you will run a bit. The minimum is not a ceiling. It is a door, and the lighter the door, the easier it is to open every day.