How to build habits that stick, according to the research

What a habit actually is

A habit is a behaviour that has gone automatic: you do it in response to a cue without deciding to. Brushing your teeth after breakfast is a habit. "Going to the gym more" is a goal, and goals are lovely but they don't do anything on their own. The whole trick is turning a goal into a specific behaviour that a specific cue can trigger, then repeating the pair until the deciding quietly falls out.

The 66-day finding

In a study at University College London, 96 people each picked one new daily behaviour, such as eating a piece of fruit with lunch or running for 15 minutes after dinner, and reported every day for 12 weeks how automatic it felt. On average it took 66 days for the behaviour to feel close to fully automatic, and the range across individuals was 18 to 254 days (Lally et al., European Journal of Social Psychology, 2010).

Two things from that study matter more than the headline number. First, simpler habits formed faster: a glass of water became automatic far sooner than a 50-sit-up routine, which will surprise nobody who has attempted a sit-up. Second, missing a single day made no measurable difference to the final outcome. Perfection was never the point. The cue-plus-repeat pairing was.

Missing once is noise. Missing twice in a row is the start of a new habit: not doing it.

The single most effective trick: an if-then plan

Psychologists call them implementation intentions, which is a lot of syllables for a very small idea. Instead of "I'll exercise more," you write "If it's 7 a.m. on a weekday, then I put on my running shoes." A meta-analysis of 94 studies covering more than 8,000 people found this simple rewording had a medium-to-large effect on whether people actually followed through (Gollwitzer & Sheeran, 2006). It works because it moves the decision from the moment of temptation to a calm moment beforehand, and hands it to a cue you're guaranteed to bump into.

Make the habit smaller than feels sensible

The habits that stuck fastest in the UCL study were the easy ones, and every practitioner from BJ Fogg to James Clear says the same thing: shrink the behaviour until it's almost embarrassingly small. Not "read for an hour," but "read one page." Not "meditate," but "sit down and take three breaths." The first two months are not about results. They're about showing up when the cue says so. You can scale up once it's automatic, and you will, because one page has a habit of turning into ten.

Anchor it to something you already do

The best cues are existing routines, not clock times. "After I pour my morning coffee" beats "at 7:30," because coffee happens every single day whether the alarm did or not. Pick a routine that already happens at the right time and place, and slot the new behaviour straight after it. If you do want a time-based cue, a reminder on your phone at that time is the next best thing to a real one.

Track it, but keep the tracking cheap

A tick in a box is genuinely useful: it's a tiny reward, and a row of ticks makes the next one feel worth protecting. What you want to avoid is a tracking system that's more work than the habit itself. One tap a day is the right amount of effort. If the habit lives in the same place as the rest of your day, you'll see it without going looking.

Plan for the miss

You'll miss days. The research says that's fine, but how people react to a miss is not always fine: the day after a slip is when most habits quietly expire. Decide in advance what a miss means (nothing, you just do it tomorrow) and what the tiny version looks like for a rotten day (one breath, one page, one minute). A habit you can do badly on a terrible day is a habit that survives.

A one-week setup

1. Pick one habit. Not three. One. We can see you reaching for three.
2. Write the if-then. "After [existing routine], I will [tiny behaviour]."
3. Set the cue. A repeating reminder at the time the routine happens.
4. Tick it daily for two months, and shrug at the odd miss.
5. Then scale it, or start the next one.