How to build a habit that sticks: a practical guide

Most advice on building habits is a mix of one real finding, a lot of folklore and a few slogans. This guide tries to keep the real finding, label the folklore, and give you a method you can start today. I make a habit tracker, so I have a bias toward tracking; I will say where tracking helps and where it does not. Where something is my opinion, I say so.

What the research actually says

The study most people cite is from University College London. Phillippa Lally and colleagues, from the Cancer Research UK Health Behaviour Research Centre, studied how long it takes a new behaviour to become automatic, and published the paper in the European Journal of Social Psychology in 2009. UCL's own news item describes two findings worth remembering:

• On average it took 66 days to reach the plateau of self-reported automaticity. That average applies to the participants for whom the researchers' model fitted well, and UCL does not give a range in the news item, so treat 66 as a rough midpoint and not a deadline.
• Missing a single opportunity did not significantly affect habit formation. People who were very inconsistent did not form habits, and the researchers said they did not yet know what level of consistency is needed.

Two practical conclusions follow. First, plan for roughly two months of effort, not three weeks. Second, one missed day is not the end of the habit; a pattern of misses is. The popular "21 days" figure does not appear in that news item, so do not plan around it.

Step 1: pick one habit, and shrink it

The most common failure is starting with a habit that is too big, such as "exercise for an hour every day". My rule of thumb (opinion, not research) is to choose a version so small that it feels almost silly: two minutes of stretching, one page of a book, one glass of water after waking up. You are not trying to get results this month. You are trying to make the behaviour automatic, and the Lally finding is about automaticity. Size can grow later, once showing up is easy.

Pick one habit at a time. Several new habits at once compete for the same limited willpower, and if one fails the others tend to wobble. Add the next one when the first feels automatic, whatever that takes you: two months is a reasonable guess.

Step 2: anchor it to something you already do

A habit needs a cue. The simplest cue is an existing routine: "after I pour my morning coffee, I do two minutes of stretching". The existing routine is the trigger, so you do not have to remember. Write the sentence down in exactly that form: after [existing action], I will [tiny habit] in [place]. Vague intentions ("I'll exercise more") rarely survive a busy day, while concrete if-then plans give you something to follow.

If you cannot find an anchor, use time and a reminder instead, and make sure the reminder actually fires. Our Android reminder troubleshooting guide covers the settings that silently kill notifications.

Step 3: make it easy, make the alternative harder

Environment beats willpower. Put the book on the pillow, the running shoes by the door, the water bottle on the desk. For habits you want to quit, add friction: delete the app, move the snacks to a high shelf, leave the cigarettes at the shop. Each small barrier is cheap, and together they change how often the "wrong" action is the easy one.

Step 4: track it, and track it honestly

Tracking has one job: to make the behaviour visible so you notice when it slips. A good tracker takes five seconds a day, shows your history as a whole rather than only the current streak, and lets you correct mistakes. In Daymark you can tick a habit from the list, see the whole year as a grid, and edit any past day if you forgot, so the history reflects what happened. That last point matters. If you cannot fix yesterday, you will either lie or give up, and either breaks the point of the data.

Decide up front what counts as "done". Write it next to the habit: "stretch for at least two minutes" is clear; "stretch" invites negotiating with yourself at 11 p.m. Count the tiny version as a full success. If you do more, great, but the minimum is the standard.

Step 5: plan for the missed day

You will miss days. Travel, illness, a bad week. Decide in advance what to do. The research says one miss does not matter much, so the plan is simple: resume the next day, without making up for it and without a speech to yourself. If you have a tool that protects a streak, such as Daymark's streak shield, use it deliberately for genuine disruptions, not as a licence to skip. Our guide on what to do after missing a day goes into this in more detail.

If you miss two or three days in a row, treat that as information, not failure. Ask what changed: Was the anchor gone? Was the habit too big? Did a reminder stop working? Adjust the design, not your character.

Step 6: review weekly, not daily

Once a week, spend two minutes looking at the last seven days. How many days did you do it? Which days did you miss, and is there a pattern (Mondays? Weekends?). A weekly review smooths out noise. A year view, such as Daymark's heat-grid, shows patterns like seasonal dips that no daily check can reveal; see our guide to reading a year heat-grid.

Step 7: raise the bar slowly, or leave it

When the habit feels automatic (for most people, roughly two months, but it will vary) you can make it a little bigger: stretch for five minutes instead of two. Make the increase small. If it breaks the habit, drop back. There is no rule that a habit must grow. A two-minute stretch done every day is better than an hour-long workout done twice and abandoned.

Building a habit versus quitting one

Quitting works the same way with a twist: you need a replacement behaviour for the cue, and you need to count something. "Days since the last cigarette" and "money not spent" are powerful because they make the absence visible. Daymark treats quit counters as first-class: they sit in the same list as your daily habits and can show money saved. Our money-saved calculator shows the arithmetic. If you are trying to stop smoking, drinking or another addictive behaviour, talk to a doctor or a qualified support service as well; some substances should not be stopped suddenly without medical advice.

A starter plan you can use today

1. Choose one habit and write the two-minute version.
2. Write the anchor sentence: after [existing action], I will [tiny habit].
3. Set a reminder if you have no natural anchor, and test it.
4. Decide what "done" means and track it in a tool you will actually open.
5. Agree with yourself that one missed day is fine and two is a signal.
6. Review once a week; grow the habit only after it feels automatic.

That is the whole method. The only part that needs patience is time: expect to be doing this for about two months before it stops feeling like effort, and do not be surprised if it takes longer.

Daymark and this website are not medical or mental-health services. For health-related habits, follow the advice of your doctor.

Frequently asked questions

How long does it take to build a habit?

A well-known UCL study (Lally et al., 2009) found an average of 66 days to reach the plateau of self-reported automaticity, for participants whose data fitted the model. It is an average, not a deadline.

What happens if I miss a day?

The same UCL study reported that missing a single opportunity did not significantly affect habit formation, although very inconsistent people did not form habits. Resume the next day.

Is it better to build one habit at a time?

Most people find it easier, because each habit needs attention until it is automatic. This is practical advice rather than a research claim.

Do I need an app to build a habit?

No. A calendar and a pen work. An app helps with reminders, history and seeing patterns over a whole year.