- Doubles results
- Consistent link
- 66 days
- Misses forgiven
How to start.
-
01
Track one habit
Start with a single behavior. Franklin worked one virtue at a time and merely logged the rest, cycling through all thirteen four times a year.
-
02
Mark it immediately
Record it right after you do it, never at bedtime from memory. Recording frequency is what predicted results in the trials, so the mark has to be honest.
-
03
Use whatever you'll mark
The medium barely matters. Paper was the most common method across Burke's 22 studies, and Franklin ran his on ruled paper in the 1720s.
-
04
Plan for 66 days
Do not stop the tracker at 21 or 30. Automaticity took 66 days on average, so three weeks ends the log right when the behavior still needs the prompt.
Why it works.
-
Doubles results
Across 1,685 adults in a six month trial, participants who kept daily food records lost roughly twice as much weight as those who kept none, with mean loss of 5.8 kg overall.
Hollis et al., 2008, American Journal of Preventive Medicine
-
Consistent link
Across 22 studies published between 1993 and 2009 (15 on dietary self-monitoring, six on self-weighing), a significant association between self-monitoring and weight loss showed up consistently.
Burke, Wang and Sevick, 2011, Journal of the American Dietetic Association, 111(1), 92 to 102
-
66 days
Tracking a daily behavior in a fixed context, volunteers took an average of 66 days to reach automaticity, so a tracker sets a realistic horizon instead of the 21 day myth.
Lally et al., 2010, European Journal of Social Psychology (UCL summary)
-
Misses forgiven
In the same study, missing one opportunity did not significantly impact the habit formation process, though people who were very inconsistent did not succeed in forming habits.
Lally et al., 2010, European Journal of Social Psychology (UCL summary)
Who swears by it.
Benjamin Franklin
Ruled a book into seven columns and thirteen virtues, marking a black spot for each daily fault.
Phillippa Lally
UCL researcher whose daily tracking study produced the 66 day habit formation figure.
Lora Burke
Pittsburgh nursing researcher who called self-monitoring the centerpiece of behavioral weight loss programs.
John's take.
A habit tracker sitting inside a habit library is about as circular as this site gets, so let me be blunt about why it earns a page. This is the one that makes the others land. Not because the marks are magic, but because a tracker is the only thing standing between what you actually did and the story you tell yourself afterwards about how consistent you have been. Memory is a very generous editor. Every other habit here assumes you can tell the truth about your own frequency, and most people cannot without a record.
The 21 day rule needs to die. Lally’s volunteers averaged 66 days to reach automaticity, and that was an average, not a finish line: simple things hooked to an existing cue landed faster, effortful ones took considerably longer. A tracker that throws confetti at day 21 is congratulating you in the middle of the job. The more useful finding from the same study is that missing one day did not meaningfully dent the process, while chronic inconsistency killed it outright. High frequency, not perfection. Anyone shaming you over a broken streak has the science backwards.
The honest tradeoff: almost every number I can point to here comes from food logging and weight loss, and Burke, who has read more of this literature than nearly anyone, still rates the evidence weak. Samples were mostly white and mostly female. Adherence was self-reported. The required dose has never been worked out. It is entirely possible that consistent trackers are simply motivated people, and the tracker is a symptom rather than a cause. Stretching the doubling effect from food records to reading or meditating is an extrapolation nobody has tested. I still track. I just track it as a good bet, not a law.
Common questions.
How long does it take to form a habit?
66 days on average, per Lally's UCL study, where volunteers tracked a daily behavior in a fixed context until it felt automatic. That is an average, not a deadline. Simple behaviors attached to an existing cue landed faster, and effortful ones took considerably longer. Set your tracker's horizon well past three weeks.
Is the 21 day rule true?
No. It is the most repeated number in self improvement and the habit formation data does not support it. Lally's volunteers averaged 66 days to automaticity. A tracker that stops at 21 or 30 days quits in the middle of the process, right when the behavior still depends on the prompt to happen at all.
Does missing one day break the habit?
No. In Lally's data, missing a single opportunity did not significantly affect the habit formation process, which makes streak shaming counterproductive. The real caveat sits elsewhere: people who were very inconsistent never reached automaticity at all. The target is high frequency, not perfection. Miss one, mark it honestly, carry on tomorrow.
Do habit trackers actually work?
The strongest evidence is self-monitoring in weight loss: across 1,685 adults, daily food records were associated with roughly double the weight loss, and the number of records kept per week predicted the result. But Burke's review of 22 studies rates that evidence weak. Consistent trackers may partly just be motivated people.
Which habit tracker app is best?
The medium barely matters. Burke's review of 22 studies found the paper diary was the most commonly used method, and Franklin ran his on ruled paper in the 1720s. What predicted results in Hollis's trial was the number of records kept per week. Pick whatever you will actually mark, then stop shopping.
How many habits should I track at once?
One to begin with. Franklin focused on a single virtue per week while merely logging the rest, cycling through all thirteen four times a year. Tracking everything is the most common way this fails, because the log turns into admin and admin gets skipped. Add a second habit once the first mark happens without thinking.